electroconvulsive therapy.9780195148206.29298

Authors: Abrams, Richard

Title: Electroconvulsive Therapy, 4th Edition

Copyright Â©2002 Oxford University Press

> Front of Book > Authors

Author

Richard Abrams M.D.




> Front of Book > Dedication

Dedication

For Trudy, again and always




> Front of Book > Disclaimer

Disclaimer

The science of medicine is a rapidly changing field. As newresearch and clinical experience broaden our knowledge,changes in treatment and drug therapy do occur. The authorand publisher of this work have checked with sourcesbelieved to be reliable in their efforts to provide informationthat is accurate and complete, and in accordance with thestandards accepted at the time of publication. However, inlight of the possibility of human error or changes in thepractice of medicine, neither the author, nor the publisher,nor any other party who has been involved in thepreparation or publication of this work warrants that theinformation contained herein is in every respect accurate orcomplete. Readers are encouraged to confirm theinformation contained herein with other reliable sources, andare strongly advised to check the product information sheetprovided by the pharmaceutical company for each drug theyplan to administer.




> Front of Book > Preface

Preface

Electroconvulsive therapy (ECT) entered the new millenniumas the oldest surviving biological treatment in psychiatryâ!”it is now 54 years old and still going strong. Thirty-fouryears after the introduction of imipramine, the firstantidepressant drug, no medication has been found thatequals ECT in antidepressant potency, which is doubtlesswhy more than 100,000 patients will receive ECT this year inthe United States alone. Of these, the over whelmingmajority will be major depressives, of whom a remarkable85%â!“90% will respond with marked improvement or fullrecovery when the treatment is properly administered.

For the first time since 1988 (the year this book firstappeared) I have added an entirely new chapter, onnonconvulsive transcranial magnetic stimulation (TMS), anew technique for electrical stimulation of the brain.Although this treatment method has not yet been approvedfor use in the United States, it has been the subject ofnumerous carefully-controlled studies, and when itultimately receives approval it will be the only otherbiological treatment in psychiatry besides lithium to have itsefficacy demonstrated by double-blind, placebo-controlledstudies prior to its introduction for general use.

In preparation for this 4th edition, I reviewed all of themore than 1000 articles on ECT published since the previousedition, as well as almost 500 articles on the psychiatric useof TMS since this method was first introduced. Of theseapproximately 1500 articles the 200 most important oneshave been analyzed in detail in the present volume.

It is far too early to specify the relative indications,advantages, and disadvantages of the two electricalstimulation methodsâ!”one convulsive, the other

nonconvulsiveâ!”as years of general clinical experience withnonconvulsive TMS will be needed before such judgments canreliably be made. Although the history of biological therapiesin psychiatry is strewn with the graves of nonconvulsiveelectrical stimulation methods, my reading of the literatureleads me to believe that nonconvulsive TMS will not sharethis fate.

Like Man, ECT is at the end of an evolutionary line, but,also like Man, rather than facing imminent extinction it isflourishing. I do not see this millennium bringing anyexciting new advances in ECT instrumentation or techniqueâ!”indeed, it is hard to see how the treatment might befurther improved at this point other than throughrefinements in patient selection, prediction of response, andmore effective dissemination of knowledge. I view this as asatisfactory state of events for patients everywhere.

R. A.Chicago, Illinois

Â†‘

Â†‘

Â†‘

FRONT OF BOOK

[+] Authors

- Disclaimer

- Dedication

- Preface

TABLE OF CONTENTS

[+] Chapter 1 - History of Electroconvulsive Therapy

[+] Chapter 2 - Efficacy of Electroconvulsive Therapy

[+] Chapter 3 - Prediction of Response to Electroconvulsive Therapy

[+] Chapter 4 - The Medical Physiology of Electroconvulsive Therapy

[+] Chapter 5 - Electroconvulsive Therapy in the High-Risk Patient

[+] Chapter 6 - The Electroconvulsive Therapy Stimulus, Seizure Induction,and Seizure Quality

[+] Chapter 7 - Treatment Electrode Placement: Bitemporal, Unilateral,Bifrontal

[+] Chapter 8 - Technique of Electroconvulsive Therapy: Theory

- Chapter 9 - Technique of Electroconvulsive Therapy: Praxis

[+] Chapter 10 - Memory and Cognitive Functioning after ElectroconvulsiveTherapy

[+] Chapter 11 - Neurobiological Correlates and Mechanisms

[+] Chapter 12 - Patients' Attitudes, Medicolegal Considerations, and InformedConsent

[+] Chapter 13 - Transcranial Magnetic Stimulation Therapy (TMS)

BACK OF BOOK

- References




> Table of Contents > Chapter 1 - History of Electroconvulsive

Therapy

Chapter 1

History of Electroconvulsive

Therapy

The traditional litany on the history of the medical uses ofelectricity, beginning with the Roman use of electric fish totreat headaches (Harms, 1956; Sandford, 1966; Brandon,1981), is simply beside the point; electroconvulsive therapy(ECT) evolved solely as a result of Ladislaus von Meduna'soriginal investigations on the effects of camphor-inducedconvulsions in schizophrenic patients. It is the chronology ofthe medical (and specifically, psychiatric) uses ofconvulsions that provides the appropriate historical perspective to his work.

This chapter draws extensively, and often without specificattribution, from the excellent historical reviews of thesubject by Mowbray (1959), Sandford (1966), Fink (1979,1984), Brandon (1981), Kalinowsky (1982, 1986), Endler(1988), and Endler and Persad (1988); from Cerletti's(1950) personal recollections; from the English translationsof the autobiography of Meduna (1985); from Accornero's(1988) eyewitness account of the discovery of ECT; andfrom my own numerous conversations over 25 years and mypublished interview with Lothar Kalinowsky (Abrams, 1988a).

According to Mowbray (1959), Paracelsus, the 16th-centurySwiss physician and alchemist, â!œâ!" gave camphor bymouth to produce convulsions and to cure lunacy.â! ! Thefirst published citation, however, is generally attributed toLeopold von Auenbrugger, the originator of the percussionmethod of examining the heart and lungs, who, in 1764,treated â!œmania vivorumâ! ! with camphor every 2 hours to

P.4

the point of convulsions (Mowbray, 1959; Sandford, 1966).The next publication (and the first in English) was by oneDr. Oliver, whose case report in 1785 in the London Medical

Journal described the successful use of camphor in a patientwho had been â!œseized with mania with few intervals ofreasonâ! ! (Kalinowsky, 1982). Fifteen minutes after a singledose of camphor, the patient had a grand mal seizure andawakened in a rational state. The case was later cited byBurrows in his 1828 textbook, Commentaries on Insanity:

In a case of insanity, where two scruples ofcamphor were exhibited, it produced a fitand a perfect cure followed. When given tothe same gentleman two years afterwards,upon a relapse, i.e., a recurrence, it hadthe same effect, even to an alarmingdegree; but the patient did not, as before,

progressively recover from a single dose,for it was repeated afterwards in smallerdoses of ten grains.

Next came Weickhardt, a councilor of the Russian ImperialCollege, who reported in a Viennese textbook in 1798 thathe had obtained cures in 8 out of 10 cases of mania withcamphor-induced seizures (Mowbray, 1959; Sandford, 1966;Meduna, 1985). The last citation given, before the methodfell into obscurity for almost a century, is from anunpublished 1851 man uscript in Hungarian by a Dr.Szekeres, who described the technique for treating maniarecommended by a Dr. Pauliczky, who gave

â!" camphor, beginning with a dose of 10grains and increasing the dosage by fivegrains daily up to 60 grains a day. Afterthis the patient will have dizziness andepileptic attacks. When he awakes fromthese, his reasoning will return. (Sandford,1966).

An English translation of Meduna's autobiography (1985)reveals that none of this work was known to Meduna until a

year after he had published his first report on inducedseizure therapy in schizophrenia, at which time a Hungarianpsychiatrist accused him of plagiarizing Weickhardt's 18thcentury ideas. Stung by the unfairness of the accusation,which was subse quently published in a Hungarian medicaljournal, Meduna says

â!"I began to read old manuscripts andfound that the convulsive method had beenused 20 years before Weickhardt byAuenbrugger â!"I found other reports:Simmon, whose nationality I could notascertain, used camphor to produceepileptic attacks to cure insanity; as didPauliczky, a Polish scientist of the 18thcentury, and a Dr. Laroze of Paris, probablyat the beginning of the 19th century.

Meduna's decision to treat schizophrenic patients by inducingepileptic seizures stemmed directly from the results ofneuropathologic studies (Meduna, 1932) in which heobserved an â!œoverwhelming and almost crushing growthof the glial cellsâ! ! in the brains of epileptic patientscompared with an equally evident lack of glial-cell growth inthe brains of schizophrenic patients. He thought theseobservations to be evidence of a â!œbiological antagonismâ! !and decided to pursue this line of inquiry further. He wasencouraged in this approach by a friend and colleague, Dr.Julius Nyiro, who had observed that epileptic patients had amuch better prognosis if they were also diagnosed as havingschizophrenia; Dr. Nyiro actually had attempted(unsuccessfully) to treat epileptic patients with injections ofblood from schizophrenic patients (Nyiro and Jablonszky,1929). Not mentioned by Meduna in his autobiography or inFink's (1984) historical review is Mowbray's (1959) assertionthat these earlier authors also had reported using pentylenetetrazol to produce convulsions in their schizophrenicpatients.

After unsatisfactory animal trials of strychnine, thebaine,nikethamide, caffeine, brucine, and absinthe(!), Medunalearned from the International League Against Epilepsy thatone of its officers had written a monograph

P.5about producing artificial convulsions with camphormonobromide. Choosing the less toxic simple camphor,Meduna successfully produced experimental epilepsy inguinea pigs (Meduna, 1934). Two months later, on January23, 1934, Meduna injected camphor in oil into aschizophrenic patient who had been in a catatonic stupor for4 years, never moving, never eating, being incontinent, andrequiring tube feeding.

After 45 minutes of anxious and fearfulwaiting the patient suddenly had a classicalepileptic attack that lasted 60 seconds.During the period of observation I was ableto maintain my composure and to make thenecessary examinations with apparent calmand detached manner. I examined hisreflexes, the pupils of his eyes, and wasable to dictate my observations to thedoctors and nurses around me; but whenthe attack was over and the patientrecovered his consciousness, my legssuddenly gave out. My body began totremble, a profuse sweat drenched me, and,as I later heard, my face was ashen gray.

Thus, convulsive therapy was born. The patient went on tofull recovery after a short series of seizures, as did the next5 patients treated; by the end of a year, Meduna hadcollected results, which he then published, from a sample of26 schizophrenic patients: 10 who recovered, 3 who enjoyedgood results, and 13 who did not change (Fink, 1984).Meduna soon replaced camphor with the chemically relatedpentylenetetrazol (Cardiazol, Metrazol), which he preferredbecause of its solubility and rapid onset of action.

Pentylenetetrazol convulsive therapy spread rapidlythroughout Europe; however, the extremely unpleasantsensations induced in conscious patients during the preictal(or myoclonic) phase of the treatment soon led investigatorsin Rome to seek alternative methods of induction (Cerletti,1956). Von Fritsch and Hitzig had already demonstrated thatepileptic seizures could be produced in dogs by electrical

P.6

stimulation of the exposed brain, and von Schilf hadsuggested the feasibility of producing convulsions in humanswith extracerebral electrodes (Mowbray, 1959; Sandford,1966).

In 1934, Chiauzzi, working in Cerletti's laboratory, producedseizures in animals by passing a 50-Hz, 220-V stimulus for0.25 seconds across electrodes placed in the mouth andrectum; in May of 1937, Bini, another of Cerletti's assistants(and himself a fine clinician who later wrote a leading Italiantextbook on psychiatry), reported similar animal studies atan international meeting in Munsingen, Switzerland, on newtherapies for schizophrenia. About 50 of the dogs thusstimulated died, and, according to Kalinowsky (1986), it wasBini who first realized the danger of passing current throughthe heart with oral-rectal electrodes and who demonstratedthe safety of applying both electrodes to the temples of thedogs he was studying. Bini confirmed this during a visit withanother of Cerletti's assistants, Fernando Accornero, to theRome slaughterhouse where, they had been told, pigs werekilled by electricity. In actuality, the pigs were firstconvulsed by an electrical stimulus to the head and thendispatched while they were comatose. The fact that suchtranscerebral electrical stimulation did not actually kill the

pigs provided encouragement for continued attempts byCerletti and Bini to define the electrical stimulus parametersthat might be safe and effective for application to humans(Cerletti, 1950; Accornero, 1988).

This goal was soon accomplished, and the first patient toreceive electroconvulsive therapy was a 39-year-oldunidentified man found wandering about the train stationwithout a ticket. He was delusional, hallucinating, andgesticulating, and alternated between periods of mutism andincompre hensible, neologistic speech (Cerletti, 1940, 1956).

After he was observed for several weeks, he was diagnosedas having schizophrenia; he received his first treatment on11 April 1938. Present were Cerletti, Bini, and only one ortwo others. An initial stimulus of 80 V for 0.25 seconds wassubconvulsive. Two subsequent stimuli of the same voltage,but with durations of 0.5 and 0.75 seconds each, wereadministered several minutes apart (Bini, 1938), despite thestatement of the patient that he did not want a third

stimulus. No effect was observed on the patient, and nofurther attempts to induce a seizure were made that day.

A few days later, a second attempt was made, this time withthe entire research team in attendance. Again the initialstimulus was unintentionally subconvulsive (80 V for 0.2seconds): The patient exhibited a brief myoclonic reactionwithout loss of consciousness and began to sing loudly. Helapsed into silence while those in attendance discussed whatto do next, and then solemnly intoned clearly and withoutjargon, â!œNot again, it's murderous!â! ! Despite thisominous warning, which understandably caused someapprehension among those present, the patient wasrestimulated at 110 V for 0.2 seconds and a grand malseizure ensued. After awakening,

The patient sat up of his own accord,looked about him calmly with a vaguesmile, as though asking what was expectedof him. I asked him: â!œWhat has beenhappening to you?â! ! He answered, with nomore gibber ish: â!œI don't know; perhapsI have been asleep.â! !

The patient's eventual full recovery with a course of 11 ECTswas dramatic, but not the important contribution made bythe Italian investigators â!”the striking effectiveness ofinduced convulsions had already been shown many timessince 1934â!”rather, it was the demonstration that suchconvulsions could be induced safely, reliably, andinexpensively by electrical means, that constituted thetechnical advance for which Cerletti and Bini justly achievedfame, and that stimulated the rapid spread of this uniquelyeffective therapeutic modality.

Cerletti and Bini (1938) published their results a few monthslater in an Italian journal, but Bini (1938) enjoyed the firstEnglish-language publication on the topic when his paper onâ!œExperimental Researches on Epileptic Seizures Inducedby the Electric Currentâ! ! was published in a supplement tothe American Journal of Psychiatry . (The topic was hisresearch in dogs, but he alluded to the first use of ECT inman in the cryptic sentence: â!œThese experiments have sofar been conducted almost exclusively in ani mals.â! !)

P.7

Electroconvulsive Therapy in theUnited StatesPresent during the second ECT administered several daysafter the first was Lothar B. Kalinowsky, a young Germanpsychiatrist who had left Berlin for Rome in 1933 whenHitler came to power. Along with Bini, Accornero, andseveral other associates, Kalinowsky was a member of aresearch team that investigated the multiform effects of ECTon the organism and eventually published its results in aspecial issue of an Italian journal of experimental psychiatry(Cerletti, 1940). Kalinowsky left Rome with his wife in 1939â!” one jump ahead of the Nazis (his mother was Jewish)â!”and traveled extensively in Switzerland, France, Holland,and England before emigrating to the United States in 1940,where he received an appointment at the New York StatePsychiatric Institute. While in England, he and Dr. J.Sanderson McGregor treated some patients at the NetherneHospital at Coulsdon with a device constructed according toplans Kalinowsky brought with him from Rome; the resultsof this work provided the basis for the first English-languagepublications on the clinical use of ECT (Kalinowsky, 1939;Shepley and McGregor, 1939).

Kalinowsky was not the first to give ECT in the UnitedStates as all of the possessions that he had shipped,including his ECT device, were delayed for 10 years by thewar. That honor belongs to Drs. Renato Almansi and DavidImpastato, who administered the first treatment at ColumbusHospital in New York City in early 1940, with a deviceAlmansi had obtained in Rome (Almansi and Impastato,1940). A few months later, Dr. Douglas Goldmanâ!”whosubsequently invented nondominant unilateral ECTâ!”demonstrated ECT at the annual meeting of the AmericanPsychiatric Association (Fink, 1987). Later that same year,Kalinowskyâ!”who by then had had another device builtâ!”started giving ECT at the Psychiatric Institute, which,because of its academic reputation as a research center,soon became a focal point for the spread of the newtreatment method in this country.

Among the postwar generation of physicians who becameinterested in studying the ECT process from a scientific point

P.8

of view, none was more influential than Dr. Max Fink, aneurologist by training, whose rapidly-developing interest inbrain-behavior relationships subsequently led him to obtainresidency training in psychiatry and certification inpsychoanalysis.

During his medical student and internship days, Dr. Finkactively participated in two research trials of note. In thefirst, as a medical student at Bellevue Hospital, headministered intravenous infusions of the dye trypan red toHuddie (Leadbelly) Ledbetter in an unsuccesful experimental

attempt to treat the amytrophic lateral sclerosis (ALS)1 thateventually killed the noted folksinger (who, ironically, hadalready survived two death sentences for murder, pardonedin each instance after singing for the prison warden). In thesecond trial, as an intern at Morrissania Hospital, Dr. Finkparticipated in one of the earliest comparisons ofsulfadiazine with the new antibiotic, penicillin, for thetreatment of empyema secondary to pneumonia. It soon

became apparent that penicillin was indeed a miracle drug,and when a severely ill young mother of two chanced to beassigned to receive sulfadiazine, Dr. Fink gave her penicillin

instead, resulting in a rapid cure. When Dr. Eli Rubin,2 thechest surgeon conducting the study, detected the switch, hethrew Dr. Fink off his service.

After medical school and internship, Dr. Fink was called toactive duty in the Army, where, after 4 months' training atthe School of Military Neuro-Psychiatry, he spent theremainder of his military career as Chief of Psychiatry at amilitary hospital.

By the time he had completed his residency in Neurology atBellevue in 1951, Dr. Fink had performed the first carotidangiogram ever done at that institution, confirming adiagnosis of subdural hematoma (Fink and Green, 1950),submitted an article on the results of 102 consecutivecarotid angiograms (Fink and Stein, 1952), and published,with his teacher, Dr. Morris Bender, a seminal article on theface-hand test (Fink, Bender, and Green 1951), aneurological â!œsoft signâ! ! assessment that was to becomea standard part of every neurological examination fordecades.

P.9

He then took a residency in Psychiatry at Hillside Hospitalfrom 1951 to 1952, and the following year did a fellowshipwith Dr. Bender, obtained certification in psychoanalysis atthe William Allanson White Institute (which he had attendedsince 1948), published his first psychoanalytic paper â!”oneof the earliest applications of statistical methods topsychoanalytic hypothesesâ!”(Tarachow and Fink, 1953),and opened an office in Great Neck, New York, for thepractice of neurology. His career in private practice (whichincluded occasionally administering ECT, unassisted, topatients in their homes) lasted only a few years, however,as he was appointed Chief of the ECT Service at HillsideHospital in 1954, and Director of the Department ofExperimental Psychiatry there in 1958, a research divisioncre ated expressly for him. At this point he enteredacademic medicine full-time to more effectively pursue hisseveral ongoing research grants.

His first papers in electroencephalography (EEG) appeared in1957, describing the EEG effects of the CNS stimulantmegimide (Green and Fink, 1957), and the relation of ECT-induced EEG delta activity to treatment response in ECT(Fink and Kahn, 1957), the latter a classic in the field, itsresults abundantly confirmed 40 years later (Sackeim et al.,1996). Among the first to recognize the importance of EEGas a research tool for the burgeoning fields of bothpsychopharmacology and ECT, Dr. Fink also pi oneered theapplication of computer -analytic quantitative EEG methods inhis studies of ECT and psychopharmacology.

Because a simple listing of his research accomplishmentswould occupy the rest of this chapter, and would, in anyevent, be incapable of conveying the essence of Max Fink'simportance to the field of ECT, I am reproducing here aneditorial I wrote a number of years ago for a specialFestschrift issue of Convulsive Therapy honoring him asfounding editor of the journal (Abrams, 1994b):

This, the first Festschrift issue of Convulsive Therapy ,honors Max Fink, founder and Chief Editor of the journalfrom its inception in 1985 through his retirement as editorin 1993.

Max Fink's firstâ!”and, for me, foremostâ!”contribution to

the field was his introduction of the scientific method intoECT research in the US in the late 1950's, during his tenureat Hillside Hospital. At a time when many of the leading ECTpractitioners in this country were purveying their anecdotaland often self-serving claims for one or another particulartreatment method, Max was conducting and publishingcarefully-controlled studies on virtually every aspect of ECT:Clinical, electrophysiological, pharmacological,neuropsychological, biochemical, psychosocial, and, ofcourse, theoretical. For example, the 1956 paper of Korin,Fink and Kwalwasser on the relation of changes in memoryand learning to the clinical efficacy of ECT was arguably thefirst neuropsy chological study of ECT to be conducted withmodern methodology, and remains a classic in the field.

When ECT fell into desuetude following the introduction ofpsychopharmacological agents, Max indefatigably stalked thecorridors of power in the American Psychiatric Association,National Institutes of Mental Health, and CongressInternationale Neuropsycho-Pharmacologium, collaring themovers and shakers of psychiatry, shaming them intoincluding a section on ECT in their programs. When theparvenu geniuses of psychopharmacology tolled the death -knell of ECT, who but Max (himself a leadingpsychopharmacologist) was there to remind them that theirreport of its demise was premature? When the firstAmerican Psychiatric Association Task Force on ECT wasconvened in 1978, who but Max was capable of initiating itsChairman into the scientific basis of this apparently arcanetherapy?

Probably the determining event for the eventual healthysurvival of ECT was not so much the publication of the 1978American Psychiatric Association Task Force Report as theappearance a year later of Max' magnum opus: Convulsive

Therapy: Theory and Practice , the first U.S. textbookdevoted entirely to ECT. Until that time, the several editionsof Lothar Kalinowsky's textbook on somatic treatments hadbeen the standard in the field, but that text's inclusion ofconsiderable material on psychosurgery, insulin coma, andmiscellaneous other somatic treatments (many of themalready long-defunct)â!”as well as a long and distractinglyEuropeanized section on psychopharmacologyâ!”had dilutedthe impact of the chapters on ECT, which, were in any case

P.10

largely anecdotal and tended to cite uncritically theconclusions of most of the papers published on the subject.

In contrast, Max' volume (which essentially constituted thebasis for his later receipt of the Anna Monika award) wasmainstream and data-oriented, presenting in full scientificdetail his more than 20 years of studies on the nature of theECT process, as well as extensive critical evaluations of theworks of others. The book remained in print for over adecade, during which time it became the undisputed bible ofECT, in fluencing practice in virtually every corner of theglobe.

Now that the number of papers published on ECT once againswells annuallyâ!”due in no small measure to Max'stewardship of Convulsive Therapyâ!”and the media hasperceptibly toned down its strident attacks on any physiciancallous enough to subject his patients to the barbarictorture of ECT, it is easy to forget the time, not so longago, when Max virtually single-handedly nursed ECT back tolife while the rest of the psychiatric community looked theother way.

The Anti-Electroconvulsive TherapyLobby: ScientologyAbsent Scientology there would hardly be an organized anti-ECT movement in the United States or anywhere else.According to Burton (1991), Scientology's vitriolic attacks onpsychiatry, psychopharmacology, and ECT are financiallymotivated:

Scientologists' central belief is that humanbeings have a soul-like entity called a â!œthetanâ! ! that is perfect and travels fromgalaxy to galaxy. Their goal is to help theirthetans get rid of something calledengramsâ!”essentially bad memories. Tothis end, Scientology developed a lie-detector -like device called an E-meter,which is used to treatmental problems oftenat hundreds of dollars per session.Psychiatrists consider these â!œtreatmentsâ

! ! quackery.

Founded in the late 1940s by science-fiction writer L. RonHubbard (who reportedly died in hiding in 1986 after 5 yearsof successfully evading an Internal Revenue Serviceindictment for tax fraud), Scientology portrayed itself as areligion despite an Internal Revenue Service ruling that

stripped the mother â!œchurchâ! ! of its tax-exempt status3

by arguing that it was more a business than a church(Behar, 1991; Burton, 1991). A Time magazine cover storydescribing the self-styled church as â!œa hugely profitableglobal racket that survives by intimidating members andcritics in a Mafia-like manner,â! ! further noted that â!œinthe early 1980's, eleven top Scientologists, includingHubbard's wife, were sent to prison for infiltrating,burglarizing, and wiretapping more than 100 private andgovernment agencies in attempts to block theirinvestigationsâ! ! (Behar, 1991).

The California Legislative ExperienceThe disproportionate effectiveness of the anti-ECT lobby isamply demonstrated by the history of the introduction andpassage of the highly restrictive Assembly Bill 4481 forlegislating ECT use in California (Moore, 1977), as well asthe saga described below of the US Food and DrugAdministration's unsuccessful effort to reclassify ECT devices(Isaac, 1990; Abrams, 1991b).

California Assembly Bill 4481 was written by a member ofthe Network Against Psychiatric Assault and presented byCalifornia Assemblyman Vasconcellos. It passed with onlyone dissenting vote and was signed into law by then

Governor Ronald Reagan.4 A successful challenge of AB 4481was mounted by the International Psychiatric Association forthe Advancement of Electrotherapy (now the Association forConvulsive Therapy), leading to replacement of AB 4481 bythe somewhat less restrictive AB 1032, which continues in

force at this time.5

In 1991, the San Francisco Board of Supervisors passed aresolution against the use or financing of ECT (Peterson,1991). Although the resolution had little practical effect, itprovided impetus for the anti-ECT forces to sponsor

P.11

California Assembly Bill 1817 that, in addition to broadeningpatients'

rights advocacy, permitted local restrictions on the use ofECT and a ban on its use for patients under 16 years of age.The California Alliance for the Mentally 111 opposed the bill,citing FDA and Alcohol, Drug Abuse, and Mental HealthAdministration support for the safety and efficacy of ECT.After a public hearing of the Committee in which patients,their families, and the California Psychiatric Association,testified enthusiastically for ECT, the bill was withdrawn forreconsideration the following year, and subse quently died incommittee.

Interestingly, although the availability of ECT steadilydeclined during the 7 years after the enactment of AB 1032,there was little year-to -year variation in its use inCalifornia: Approximately 1.1 persons per 10,000 populationper year received ECT during 1977-1983 (Kramer, 1985), afigure that is just below the range of the national average of1.3 to 4.6 per 10,000 when sampled in 1978 (Fink, 1979),but less than the reported 2.42 patients per 10,000population who received ECT in Massachusetts from 1977 to1980 (Kramer, 1985).

A subsequent review of the use of ECT in California (Kramer,1999) covering the decade from 1984 to 1994 found theaverage annual rate of administration of ECT to be 0.9 per10,000 population, only marginally below the average forthe years 1977-1983. However, the rate recorded for 1994,the last year of the decade studied, was only 0.8 per 10,000population, suggesting a significant decline since the earlierstudy. There was no change from the previous period in thenumber of counties where ECT was avail able, and a slightincrease in the number of facilities offering ECT.

Regulation of ElectroconvulsiveTherapy Devices in the United StatesWhen the Medical Device Amendment to the Food, Drug, and

Cosmetic Act6 gave the Food and Drug Administration (FDA)regulatory responsibility for medical devices in 1976, FDAplaced ECT devices in Class III, which requiresmanufacturers to provide data demonstrating safety and

P.12

efficacy of new devices they intend to market. However,because existing ECT devices, as well as those subsequentlyintroduced as â!œsubstantially equivalentâ! ! to pre-1976devices were exempted from such premarket approvalprocedures under a grandfather clause, there was nopractical significance to the FDA's action.

In 1978 the FDA recommended reclassifying ECT devices intoClass II, which assumes ECT to be both safe and efficaciousand requires only that devices meet a performance standardfor safety of construction and instructions for use. Underfire from Scientologists and other antipsychiatry activists,the FDA quickly reversed its opinion, placing ECT devicesback into Class III in 1979, while simultaneously inviting theAmerican Psychi atric Association (APA) to submit a petitionrequesting that ECT devices be reclassified to Class II.

The APA procrastinated 3 years before finally filing the

petition7 in 1982, following which the FDA againrecommended that ECT devices be reclassified into Class II,but this time contingent on the development of a

performance standard.8 A year later, the FDA publishednotice of its intent to reclassify ECT devices into Class II,and there the matter stood for the next 7 years while theFDA waffled to the tune of the Scientologists and other anti-ECT activists, despite the fact that various national andinterna tional performance standards (e.g., IEC 601) hadalready achieved world wide acceptance.

Finally, in 1990, the FDA proposed a definitive rule9 to placeECT devices in Class II (still, however, contingent on thedevelopment of a performance standard) but this time onlyfor devices labeled as intended solely for use in patientswith â!œsevere depression,â! ! which the FDA defined asDSM-III major depressive disorder with melancholia. (ECTdevices intended for use in other conditions, includingmania, catatonia, and schizophrenia, were to remain in ClassIII, requiring manufacturers to undertake enor mouslyexpensive controlled trials of safety and efficacy and to seekFDA approval separately for each condition.)

Although the FDA claimed that its decision to severely limitthe clinical indication for ECT was â!œbased on its review of

P.13

new, publicly available, valid scientific evidence,â! ! the taskforce conducting the review included one identifiedphysician, three individuals with non-medical doctorates,and one person without a stated degree, none of whom werepsychiatrists or had published on ECT. The 200 articlesreferenced in their report omitted numerous controlledstudies of the use of ECT in nonmelancholic depression,mania, catatonia, and schizophrenia that had appeared inmajor world journals during the years reviewed. Mostegregiously, the FDA report relied mostly on long-outdatedstudies in support of its decision to exlude non-melancholicmajor depression from the approved incidations for ECT (4of the 5 studies cited were from 1953-1965), while ignoringevery single random-assignment, double-blind, sham ECT-controlled study of the modern era that had demonstratedthe efficacy of ECT in such patients. Equally flawed was theFDA's justification for excluding from approval for ECT otherdiagnoses, including especially mania and catatonia.

About this time, in 1990, the House of Representatives,apparently fed up with the FDA's dawdling incompetence,introduced a proposed new section of the Safe Medical

Devices Act10 calling for judicial review of any of thepromulgated regulations that were contestedâ!”effectivelytaking the final decision out of the FDA's hands. The Senatefollowed with its own committee reportâ! ! stating that it wasnot its intention to require that a performance standardmust exist before such Class III devices could be placed inClass II.

Another Senate Report issued in late 1994 required the FDAto complete its reclassification of all pre-amendment ClassIII devices into classes II or I, or retain them in Class III,by December, 1995, while at the same time authorizinganyone who felt adversley affected by the regulation topetition the US Court of Appeals for a judicial review. Amonth later, the

FDA announced it would issue an immediate call to allmanufacturers of ECT devices to submit a summary of andcitation to any data known to them respecting safety andefficacy of their products. The FDA would then be requiredby law to publish a proposed regulation to reclassify ECT devices to Class II or I or retain them in Class III, providing

up to 90 days for public comment before taking final effect.

The call was issued and the responses duly submitted, butnothing further was ever heard from the FDA on thesubjectâ!”the status of ECT devices in the United States atthe time of this writing, 8 years after the 1994 Senatereport, remains in the same limbo in which it was placed aquarter -century ago.

Fortunately, this has prevented neither US ECT devicemanufacturers from introducing new models, nor USpsychiatrists from administering the latest forms of ECTaccording to their best clinical judgment, with one importantexception: the FDA has steadfastly refused to allow USdoctors to administer the higher ECT dosages required todeliver clinically effective treatmentâ!”or even to obtain aseizureâ!”in a significant number of patients (Sackeim,1991b; Lisanby et al., 1996; Abrams, 2000; Krystal, Dean,and Weiner, 2000), dosage levels that have long beenavailable to psychiatrists in many other countries. Indeed,during the late 1980s, the Royal College of Psychiatrists(1989) issued the requirement that all ECT devices offeredin the United Kingdom be able to deliver twice the dosagelevel presently permitted for ECT devices sold in the UnitedStates.

Electroconvulsive Therapy Use in theUnited StatesAn analysis of the National Institutes of Mental Healthnational survey data for the years 1975, 1980, and 1986showed that the declining use of ECT ended in the 1980s(Thompson, Weiner, and Myers, 1994). In 1986, 36,558patients received ECT, which represented a decrease fromthe 58,667 who received ECT in 1975, but an increase over1980, when 31,514 patients were so treated. Strikingly,recipients of ECT were primarily older white patients inprivate institutions; patients in state hospital facilities rarelyreceived ECT. The figures presented are doubtlessunderestimates of the true use, primarily because ofsampling error such as chance omission from the sample ofa few large-volume ECT centers. The authors estimated thatthe 36,558 patients treated in 1986 received approximately300,000 ECTsâ!” equivalent to the number of proceduresperformed for coronary bypass, tonsillectomy, inguinal

P.14

hernia, or appendectomyâ!”thus making ECT one of themost common procedures carried out in patients givengeneral anesthesia.

Analyzing data from the American Psychiatric AssociationProfessional Activities survey of 1988-1989, Hermann et al.(1995) found that 1102 psychiatrists reported treating 4398patients during the previous month. Extrapolating from theseresults, the authors estimated that 4.9 patients per 10,000population received ECT annuallyâ!”a modest increase overthe 1978 American Psychiatric Association estimate of 4.4per 10,000 populationâ!”

yielding an estimate of 100,000 patients treated in theUnited States during 1995. Because ECT is given in virtuallyevery other country of the world â!”and not infrequently atmuch higher rates of use than in the United States â!”it islikely that between 1 and 2 million patients per year receiveECT worldwide.

The aging of the US population has resulted in an increasingnumber of geriatric patients who receive ECT. Rosenbach,Hermann, and Dorwart (1997) studied a 5% sample(representing about 4000 individuals) of all Medicare part Bclaims for 1987-1992 and found a 30% increase in receiptof ECT during the period (equal to an extrapolated totalnational increase from 12,000 to 15,500 Medicare patientstreated). The rate of use of ECT in this group increased from4.2 to 5.1 per 10,000 population, which is consistent withthe overall use of ECT reported in the previous paragraph.During the study interval, outpatient use of ECT more thandoubled: from 7% to 16% of all treatments administered.This study is the first to show a clear increase in ECT use inthe United States, after 30 years of generally declining, and,more recently, stable, use.

Olfson et al. (1998) analyzed inpatient ECT data from the1993 Health-care Cost and Utilization Project of the Agencyfor Health Care Policy and Research, which comprises asample of 6.5 million case records from 913 communityhospitals in 17 states (approximately 20% of US communityhospitals). Close to 10% of the 22,761 general hospitalpatients admitted with a principal diagnosis of recurrentmajor depression received ECT during the survey year. Thehighest rates of ECT use were found in older, white,

P.15

privately-insured, more affluent patients, with sharply loweruse among black, Hispanic, and low-income patients. Aftercontrolling for patient se lection bias, prompt administrationof ECT was found to be associated with shorter and lesscostly hospital stays.

It is apparent from the above surveys that ECT is alive andwell in the United States. It is also apparent that ECT ismarkedly underutilized in economically disadvantaged, andstate-hospital, populationsâ!”presumably for the sameeconomic reasons that these populations receive the lowest-echelon treatments available in every other branch ofmedicine as well. When health insurance becomes availableto members of these populations, as it has in recent decadesthrough Medicare, ECT utilization rates increase sharply;those patients entering state facilities because they couldnot afford a higher level of care, find themselves candidatesfor admission or transfer to community hospitals, where, forthe first time, they become eligible to receive more effectivetreatments, including ECT. If and when a national healthinsurance program is introduced in the United States, therewill doubt lessly be a corresponding substantial increase inECT use.

The Future of ElectroconvulsiveTherapyAs Fink (1979) pointed out, convulsive therapy burst on thescene during an era of unprecedented therapeutic optimismin psychiatry, following hard

on the heels of Wagner-Jauregg's malarial fever therapy forgeneral paresis of the insane (1917) and Klaesi's prolongedsleep therapy (1922), and virtually coeval with Sakel'sinsulin coma therapy (1933) and Moniz' psychosurgery(1935). One by one, the other treatments flourished brieflyand then fell into disuse, to be replaced by less complex andmore definitive methods. Only ECT flourished and remainswidely used to this day, doubtless because of itsdemonstrable efficacy, safety, and relative ease ofadministration, all due in large measure, to the advances intechnique (e.g., succinylcholine muscle relaxation,barbiturate anesthesia, oxygenation, unilateral and bifron talelectrode application, seizure monitoring, brief-and

ultrabrief-pulse stim ulation) that have been introduced overthe years.

Will ECT ultimately be replaced by a less intrusive,pharmacologic therapy that alters brain function in thedesired direction (e.g., via a hypothalamic neuropeptide) butwithout the auxiliary convulsion and its attendant risks anddrama? Perhaps, but not in the foreseeable future. The rateof accumulation of new techniques and discoveries in theapplication of neurotransmitter pharmacodynamics to thetreatment of major depression has been excruciatingly slow.Despite manufacturers' claims, no significant improvement inthe therapeutic potency of antidepressant drugs has materialized since the introduction of imipramine and amitriptylinenearly a halfcentury ago (Barbui and Hotopf, 2001).

Moreover, incremental advances in the technique of ECThave refined the treatment to the point that, with high-doseright unilateral ECT, or moderate-dose bifrontal orbitemporal ECT, administered with brief-or ultrabrief-pulsetechnique, many patients can now enjoy the full therapeuticbenefit of ECT without the prominent cognitive side effectsthat were so common with sine-wave bitemporal ECT. Mostimportantly, those patients requiring bitemporal ECT cannow receive it in a more physiological form than before,using the shorter pulse-widths and longer stimulus trainsthat are more consistent with the parameters of neuronaldepolarization and recovery (see Chapter 6), and thereforeless likely to impair memory and cognition.

In the previous edition, I described the possibility that ECTmight someday be replaced by magnetic convulsive therapy.My view now is that the economic realities detailed inChapter 13 make this unlikely to occur for many years, ifever. Improvements in instrumentation and stimulationparameters for administering nonconvulsive repetitivetranscranial magnetic stimulation (r TMS) might elevate its

efficacy in major depression to the level of ECT, although itwould be hard to improve significantly on the 87%-95%remission rate recently achieved in a 4-hospital collaborativestudy of brief pulse bitemporal ECT (Petrides et al., 2001).Certainly, it would be an important advantage ifnonconvulsive rTMS were at the same time to have fewer

side-effects than ECT on memory and cognition, and

P.16

definitive infor mation on these points should be availablewithin a few years.

Most likely brief pulse ECT will continue to benefit fromfurther fine-tuning of its stimulus parametersâ!”thereintroduction of the ultrabrief stimulus

providing a typical exampleâ!”and advances in techniquesfor ensuring the maximum possible benefit from eachtreatment session. Given the facts that recent research hasclearly demonstrated how to administer ECT with a degree ofefficacy that equals or surpasses any other treatment inmedicine, yet with a morbid and mortal risk below that ofmany drugs, and virtually all other procedures carried outunder general anesthesia, the long-term sur vival of ECTseems assured.

Notes1An autotoxic theory of ALS was then in vogue, which thetrypan red was intended to combat by preventing transfer ofautotoxins across the blood-brain barrier.

2Dr. Rubin subsequently published the extremely favorablestudy results in his 1947 volume, Diseases of the Chest.

3The tax-exempt status of the Church of Scientology wassubsequently meekly restored by the IRS after the Churchthreatened to bring IRS operations to a halt throughlawsuits and various other actions.

4When later criticized for this action, Reagan typicallydisclaimed responsibility, stating that he â!œhad no respectfor the type of people who had supported the Vasconcelloslawâ! ! and had signed the bill at the end of the legislativesession when he had had more than 1000 legislative actionsto consider (Bennett, 1983).

5A few years later, in November 1982, the citizens ofBerkeley, California, approved by referendum a Board ofSupervisors ordinance that made administering ECT in cityhospitals a crime punishable by a fine of $500 or 6 monthsin prison, or both (Bennett, 1983). This ordinance wassubsequently reversed by the Alameda County SuperiorCourt on a technical point of law.

6May 28, 1976: Congress enacts Medical Device Amendmentto Federal Food, Drug & Cosmetics Act (FFDCA).

7August 13, 1982: APA submits petition 82P-0316/F820007to FDA under section 513(e) of FFDCA (21 USC 360c(e)) toreclassify the ECT device to Class II.

8November 4, 1982: FDA Advisory Panel agrees to APA'srequest, contingent on development of mandatory Safety &Performance standard under section 514 of the FFDCA (21USC 360d).

9September 5, 1990: FDA proposes rule (55 FR 56378-90).

10October 5, 1990: House Report #94-583, p. 53.

11October 9, 1990: Senate Report #101-513, p. 17.




> Table of Contents > Chapter 2 - Efficacy of Electroconvulsive

Therapy

Chapter 2

Efficacy of ElectroconvulsiveTherapy

Experimental DataIt is axiomatic that rigorous experimental methods arerequired to demonstrate the efficacy of a medical treatment.Whether the comparison is with placebo (sham treatment) orwith an alternative active therapy, a prospective design withrandom assignment of consecutive patients to treatmentgroups and blind assessment of outcome using objectivemeasures are absolute requirements. Both the diagnosticcriteria and the precise treatment parameters must bespecified, and appropriate statistical analyses must beemployed (or the data presented in sufficient detail forreaders to perform their own calculations). Scrupulousadherence to these rules is especially crucial when studyingan emotionally charged and physiologically active treatmentsuch as ECT, for it is often used for illnesses (depression,mania) with a high spontaneous remission rate.

The first part of this chapter assesses the efficacy of ECT byreviewing the evidence from controlled trials in the threedisorders for which such data are available: depression,schizophrenia, and mania. The results of uncon trolled orotherwise methodologically weak studies, anecdotal reports,and case history studies are referred to in the second part.

Depressive Illness

Sham Electroconvulsive Therapy StudiesThe studies of genuine versus sham ECT published through

P.18

1966 and reviewed by Barton (1977), Fink (1979), andTaylor (1982) generally support the efficacy of ECT intreating severe depression, although each suffers frominadequate methods of varying degree (Crow and Johnstone,1986). The following review concentrates on the randomassignment studies published since then, each of whichsatisfies the methodological requirements outlined earlier.

Freeman, Basson, and Crighton (1978) treated 40 primarydepressives with either 2 genuine (bilateral, partial sine-wave) or 2 simulated ECTs during their first week oftreatment, after which, for ethical reasons, all patients

received genuine bitemporal ECT for the remainder of thecourse. Anesthesia was identical for both groups andincluded atropine, barbiturate, and muscle relaxant. Meanscores on the Hamilton, the Wakefield, and the VisualAnalogue depression scales after the first 2 treatments weresignificantly lower after genuine than after simulated ECT,and patients in the simulated ECT group ultimately receivedsignificantly more treatments prescribed by clinicians whowere blind to group assignment. (The Beck self-ratingdepression scale did not reveal any significant between-group differences, perhaps because depressed patients,particularly those with retardation, have diffi cultycompleting it.)

Lambourn and Gill (1978) assigned 32 patients withpsychotic depression to receive either 6 brief-pulse, low-dose (10 joules [J]), unilateral ECTs, or an equal number ofidentical anesthesia inductions without the passage ofelectricity. Mean Hamilton rating-scale scores obtained 24hours after the sixth treatment did not differ significantlyfor the 2 groups.

In the Northwick Park ECT trial, Johnstone et al. (1985)gave 70 endogenous depressives a 4-week course of 8partial sine-wave bitemporal ECTs or 8 anesthesia inductionswithout electrical stimulation. Mean Hamilton depressionscale scores after 4 weeks were significantly lower in thegenuine ECT group by about 26, a difference that was nolonger present at 1-and 6-month follow-up intervals, duringwhich additional treatment (including ECT) had been givenad libitum. The advantage of genuine over sham ECT in thisstudy was most marked in the subgroup of delusional

P.19

depressives (Clinical Research Centre, 1984).

West (1981) treated 22 primary depressives with courses of6 genuine or sham ECTs. The patients then completed theBeck self-rating scale for depression, were blindly rated onboth doctors' and nurses' rating scales, and were thenswitched to the alternate treatment if indicated. There was ahighly statistically significant and clinically importantimprovement in the genuine compared with the sham ECTgroup, and 10 out of 11 sham ECT patients (but no genuineECT patients) were switched to the alternate method, fromwhich they derived the expected degree of improvement.

In the Leicestershire trial, Brandon et al. (1984) studied 95major depressives who were allocated to up to 8 genuine(bitemporal, partial sine-wave) or sham ECT, administeredtwice weekly. A significantly greater improvement inHamilton depression scale scores was seen in the genuine(compared with the sham) ECT group at 2 and 4 weeks, butnot at 12 and 28 weeks. As in the Northwick Park trial, thelargest between-group differ ences occurred in the subgroupof delusional depressives.

In the Nottingham ECT study, Gregory, Shawcross, and Gill(1985) randomly assigned 60 depressives to partial sine-wave ECT with bitemporal or unilateral placment, or to shamECT. Both genuine methods were superior to sham ECT after2, 4, and 6 treatments, as measured by the Hamilton andthe Montgomery and Asberg depression scales, which wereadministered blindly.

Thus, 5 out of 6 methodologically impeccable studies ofsimulated compared with real ECT in the treatment ofdepressive illness show both a statistically significant andclinically substantial advantage for the genuine article inreducing depression scale scores during and immediatelyfollowing the treatment course. It is not surprising thatevaluations done later in the maintenance phase of thetreatment course or at follow-up generally fail to show suchan advantage; during the intervening weeks patientstypically received a variety of â!œdoctor's choiceâ! !treatments, including both ECT and drugs, administeredunsystematically.

The single study (Lambourn and Gill, 1978) that failed to

show an advantage for real compared with sham ECT alsodiffers from all the others in having used brief-pulse, low-dose (10 J) unilateral ECT as the active treatment. A similarlow-dose technique using an even higher stimulus energy(mean = 18 J) was shown by Sackeim et al. (1987a) to beclinically ineffective for right unilateral ECT. Recent evidencedemonstrates that this method must be administered withhigh stimulus dosing to maximize efficacy (Abrams, Swartz,and Vedak, 1991).

Electroconvulsive Therapy Compared withAntidepressant DrugsThe case for a therapeutic advantage of ECT overantidepressant drugs rests primarily on three studies:Greenblatt, Grosser, and Wechsler (1964), the MedicalResearch Council trial (MRC, 1965), and Gangadhar, Kapur,and Kalyanasundaram (1982). Although many studies haveprovided interesting and useful insights into special aspectsof the relative efficacy of the 2 treatment methods, nonehas the scientific rigor necessary for an unequivocaldemonstration of the superiority of EC T. Abrams (1982b)and Rifkin (1988) have detailed the methodological flaws ofthe published comparisons of ECT and antidepressant drugsin the treatment of depressive illness. Half of the studieshave to be excluded from consideration because ofretrospective design; nonblind evaluation and faulty dataanalyses account for most of the remainder. These are by nomeans trivial points. In a retrospective study, for example(they are all chart-reviews), patients have not beenassigned randomly to treatments; there is no sure way toequate the groups for psychopathology or illness severity;the reasons why physicians or patients chose one or theother treatment constitute a major source of bias; there isno control over drug dosage or numbers of ECTsadministered; and outcome assessment (even if done by â!œblindedâ! ! reviewers) is necessarily based on thenonsystematic observations recorded at the time by nonblindclinicians with unknown biases.

Even studies that apparently follow a rigorous method mayfade into insubstantiality on closer scrutiny. A case in pointis the previously mentioned study by Greenblatt, Grosser,and Wechsler (1964) that is widely cited as a demonstration

P.20

of the therapeutic superiority of ECT over imipramine in thetreatment of depressive illness. In this trial, 281 patientswere randomly

assigned to receive either ECT, a maximum obligatory doseof 200 mg/day of imipramine, phenelzine, isocarboxazid, orplacebo and evaluated blindly. The authors indeed found ECTto be superior to imipramine across the total samplestudied, but diagnoses were heterogeneous and includedpsychoneurotic depression, schizophrenia, and a largenumber categorized only as â!œother,â! ! in addition to thediagnostically relevant categories of manic -depressive,depressed, and involutional psychotic reaction. A combinedanalysis is clearly noninformative with such diagnosticheterogeneity. Although a table provides separatepercentages for each diagnostic subgroup of patients whowere markedly improved with each treatment, the actualnumbers of patients receiving each method are not given,nor are chi-square values or significance levels provided.The authors nevertheless affirm that their analyses showECT to be significantly more effective than imipramine forthe treatment of involutional psychotic reaction (85 versus42 markedly improved) but not for the depressed phase ofmanic -depressive illness (78 versus 59 markedly improved);these groups were not combined for analysis.

In the multihospital Medical Research Council trial (MRC,1965), 269 patients with endogenous depression wererandomly assigned to 4 different treatment groups, 2 ofwhich comprised 4 to 8 ECTs (65 patients) and imipramine,100 to 200 mg/day (mean =19 3 mg/day; 63 patients).Fifty-eight patients in each group completed the first 4weeks of treatment, at which time physicians' blind globalassessments showed 71 of the ECT group to have no orslight symptoms, compared with 52 of the imipramine group

(X2 = 8.75; p = 0.0005).

Gangadhar, Kapur, and Kalyanasundaram (1982) studied 24primary endogenous depressives who were randomlyassigned to receive a course of genuine bilateral or shamECT given over a 12-week trial in conjunction with eitherplacebo capsules or imipramine, 150 mg/day. The first 6treatments were given over 2 weeks, followed by 1treatment per week for 2 additional weeks and then 1 â

P.21

!œmaintenanceâ! ! treatment at the 6th, 8th, and 12th weeksof the trial (total, 11 treatments). Genuine ECT plus placebocapsules was significantly superior to sham ECT plusimipramine in lowering Hamilton depression scale scoresafter 6 treatments; no significant between-group differenceson this scale were observed at subsequent assessmentintervals. Assuming that imipramine does not antagonize theantidepressant effects of ECTâ!”Price et al. (1978) suggestthat this may not be the caseâ!”this study alsodemonstrates the efficacy of genuine versus sham ECT.Although the sample size is small and the dose ofimipramine used is low, this is the only study to employ thecritical format of genuine ECT plus placebo compared withsham ECT plus active drug in conjunction with all of theother methodologic requirements.

All three studies, however, can be criticized for the low drugdosages employed. Although there is little doubt thatimipramine, 100 to 200 mg/ day, is an effective treatmentfor some patients, most psychopharmacologists today wouldpeg the therapeutic range of this antidepressant at 200 to300 mg/day and would also require plasma-level monitoring.The two-phase

study of Wilson et al. (1963) addresses the question ofdosage, albeit in a very small sample. In the initial phase,depressives were randomly assigned to 4 treatment groups,of which 2 (6 patients each) were ECT plus placebo andsham ECT plus imipramine at a dose of 150 to 220 mg/day(mean = 180 mg/day). Assessment on the Hamilton Scaleafter 5 weeks showed a large and highly significantadvantage for ECT. In the second phase of the study, 14new patients were treatedâ!”4 with ECT and 10 withimipramine aloneâ!”at a higher dosage: 215 to 270 mg/day.After 5 weeks on this regimen, the high-dose imipraminegroup showed significantly more improvement than the first-and second-phase ECT groups combined (although theauthors erroneously describe the 2 methods as identical).The rating procedure, however, presents an importantproblem in this study: Different numbers of raters were usedat different assessment periods; 1 of the raters was neverblind; and 1 was not a psychiatrist. Moreover, the authorsdo not say which raters participated in the second-phaseassessments or how the Ham ilton scores were derived when

more than 1 rater was used.

Another aspect of the imipramine dose-response relation wasstudied by Glassman et al. (1977), who treated 42nondelusional psychotic depressives with a fixed milligramper kilogram dose of imipramine and examined the relationof plasma level to clinical outcome. The proportion ofimipramine-responders increased directly with plasma levels:29 for plasma levels of 150 ng/mL, 64 at 150 to 225 ng/mL,and 93 for levels of 225 ng/ mL. The study is renderedmeaningless, however, by the authors' failure to specifytheir criteria for defining treatment response.

Thus, although ECT is clearly more effective than moderatedoses of imipramine in treating several subtypes ofendogenous depression, it is less obvious that this differencewould obtain under the optimal conditions of higher drugdosages, perhaps with plasma-level monitoring. To be sure,most practitioners neither administer high-doseantidepressant drugs nor routinely monitor plasma levels; inthis sense, ECT can justifiably be considered superior to themediocre antidepressant therapy that is generallyprescribed.

In a different paradigm, Dinan and Barry (1989) randomlyassigned 30 severely depressed patients who did notrespond to treatment with tricyclics â!”of whom 23 metcriteria for melancholiaâ!”to receive either 6 bilateral ECTsor the addition of lithium to the tricyclic. There was nodifference between groups in blindly obtained, depression-scale score reductions at the end of 3 weeks, althoughpatients receiving the lithium-tricyclic combination improvedfaster. This is the most favorable outcome obtained to datefor drug therapy of depression when compared with ECT andsupports the rapid antidepressant efficacy previouslyreported for the lithium-tricyclic combination (De Montignyet al., 1983; Heninger, Charney, and Sternberg, 1983).

Electroconvulsive Therapy versus Drugsin Depressive Illness: Other Studies ofInterestThe 1964 study of DeCarolis et al., as reviewed by Averyand Lubrano (1979), provides unique information on animportant clinical question: What

P.22is the response to ECT in depressives who have failed high-dose antidepressant drug therapy? These authors intiallytreated a diagnostically heterogeneous sample of 437depressives with imipramine, 200 to 350 mg/day. Allpatients who failed to improve after 30 days on this regimenwere then given a course of 8 to 10 ECTs. Endogenousdepressives constituted the largest diagnostic subgroup (n =282), of which 172 (61) responded to imipramine. Of theremaining 109 patients (1 patient dropped out), 93 (85)then responded to a course of ECT. In the subgroup of 181delusional depressives, only 72 (40) responded toimipramine, compared with 91 (83) of the 109 imipraminenonresponders who went on to receive ECT. Althoughassessment of outcome was not blind in this study, thisseems at least partially counterbalanced by the powerfulbias against ECT response introduced by withholding thistreatment until patients had first failed high-dose antidepressant drug therapy.

A paper by Coryell (1978) considers a different question:What is the response of patients who had received ECT andantidepressants during different depressive episodes? In thisstudy, hospital charts were reviewed and blindly rated for allpatients who received ECT for depression in the pre-antidepressant era (1920-1959) and who later receivedtricyclic antidepressants from 1961 to 1975 for a differentepisode. Complete recovery occurred in 94 of the episodestreated with ECT compared with 53 of those treated withantidepressants. Drug dosages were low by presentstandards, however, and no data are provided on therelative efficacy of the 2 methods within patients (e.g., howoften the ECT response was superior to the tricyclicresponse).

Electroconvulsive Therapy versusIsofluraneBecause of a superficial analogy between ECT-inducedpostictal suppression and the total suppression of cerebralactivity that can occur with deep isoflurane anesthesia,Langer et al. (1985) conducted an open clinical trial of thisprocedure in 11 treatment -resistant depressed patients whopreferred not to undergo ECT, and claimed thereby to have

P.23

achieved a very rapid antidepressant effect that wascomparable to ECT. Stimulated by that article, Greenberg etal. (1987) conducted an open replication trial in 6 patientswith recurrent depressive disorder, 5 of whom had recoveredwith ECT from prior episodes. No clinical antidepressantactivity of deep isoflurane anesthesia was observed duringthe study, and 5 of the 6 patients went on to recover withECT.

An open study of isoflurane anesthesia and ECT in depressedpatients (Carl et al., 1988) is simply incomprehensible asinsufficient methodology is provided to determine what wasdone to whom, or why, and what the results were. Thissame group published a subsequent open clinical trial(Engelhardt, Carl, and Hartung, 1993) in which 12treatment -resistant depressed patients were first given 6isoflurane inductions, and those who failed to respondâ!”orimproved only temporarilyâ!”were then given ECT.

In a nonblind assessment, the authors rated 7 of the 12patients as markedly improved after isoflurane anesthesia,but only 3 of them could be discharged from the hospital.The remaining 9 patients went on to receive ECT, but theauthors unaccountably omit mention of whether any of themwere subsequently discharged from the hospital. As theauthors point out, however, isoflurane anesthesia iscontraindicated in patients with coronary or cerebralvascular disease, a fact that would certainly prevent asubstantial number of patients who receive ECT from everbecoming candidates for this procedure.

Most recently, Langer et al. (1995) reported an open,nonrandom, clinical comparison of ECT with isofluraneanesthesia in depression, purporting to find isofluraneanesthesia the more effective therapy. Remarkably, all patients continued to receive antidepressant drug therapythroughout the study period.

However, their study is invalidated by inadequate ECTtechnique. Using a Siemens Konvulsator partial sine-wavedevice set to the intermittent stimulus mode, with peakcurrent of 500 mA, these authors delivered a stimulus only 2seconds long. (If they didn't induce a seizure, theyimmediately restimulated at 600 mA.) Langer et al. (1995)do not provide figures for the mean charge they used, but it

P.24

is easy to calculate using the correction factor of 0.64provided by the manufacturer of the Siemens device. Eachsecond of stimulation in the intermittent mode provides0.125 second (25 pulses per second of 0.005 second each)of current flow, so:

For the 600-mA setting, the dose would have been 96 mC.Such a low dosage range is doubtless responsible for theincredibly poor results these authors obtained with bilateralECT: just 49% improvement in depression scores after 6treatments.

In comparison, Lamy, Bergsholm, and d'Elia (1994), whoalso used an old Siemens Konvulsator, delivered an 800-mApeak current for an average of 6 seconds' stimulation (range4 to 10 seconds), yielding a mean stimulus charge of 384mC and achieving recovery in most patients with bilateralECT. Similarly, Abrams, Swartz, and Vedak (1991) reported79% improve ment with 6 bilateral ECTs, using a meanstimulus charge of 378 mC de livered via a brief-pulseinstrument.

Efficacy in ManiaOnly one prospective controlled trial of ECT in mania hasbeen published at the time of this writing (Small et al.,1988): A sample of 34 newly admitted manic patients werediagnosed as bipolar I according to the Research DiagnosticCriteria and were randomly assigned to receive a course ofbrief-pulse ECT (Â« = 17) or lithium therapy (Â« = 17). Themean number of ECTs administered was 9.3, and lithiumdosages were adjusted to yield serum

lithium levels between 0.6 and 1.2 mmol/L. Concomitantneuroleptic drug therapy was permitted ad libitum. Aftercompletion of the ECT course, pa tients in this group wereplaced on maintenance lithium therapy.

Ratings by nonblind observers as well as blind evaluations ofvideotaped interviews were done at weekly intervals for thefirst 8 weeks, using a variety of rating instruments,including the Brief Psychiatric Rating Scale, Clinical GlobalAssessment scale, and Bech-Rafaelson Manic Rating Scale.At all rating intervals after the first week, and for each of

the 3 previous measures, ECT induced greater improvementthan lithium, a difference that reached statistical significanceon most measures at weeks 6, 7, and 8. The results of theblind and nonblind ratings were in general agreementthroughout the study. The mean daily dose of neurolepticagents received during the trial was similar for both groups.It is notable that a significant advantage for ECT emerged inthis study despite the fact that the first group of ECTpatients treated initially received unilateral electrodeplacement, failed to respond (or got worse), and thenresponded to bilateral placement, thereby confirming anearlier retrospective study that demonstrated a therapeuticadvantage for the latter method in manic patients (Small etal., 1985; Milstein et al., 1987). Had all ECT patientsreceived bilateral ECT from the outset, the observedadvantage over lithium might well have been larger.

Efficacy in SchizophreniaEvaluating the efficacy of ECT in schizophrenia is moredifficult than in depression because of the greater variabilityin the diagnostic criteria for the former disorder. Mostinvestigators who were quite specific in their description ofthe signs and symptoms of endogenous depression wereunaccountably satisfied with merely proclaiming theirschizophrenic patients to be either chronic or acute, with anoccasional subtype thrown in. Moreover, no attempt wasmade until the 1980s to exclude schizophrenic patients whohad prominent affective symptoms from ECT studies, raisingthe spectre of misdiagnosis because many patients with amixture of affective and schizophrenic symptoms actuallysuffer from affective disorder (Abrams and Taylor, 1976c,1981; Pope and Lipinski, 1978; Pope et al., 1980).

Moreover, just as in the studies in depressive illness alreadyreviewed, most of the older ECT studies in schizophrenia aremethodologically defective. The few acceptable studies fromthis era are briefly reviewed here, followed by a moredetailed examination of the few methodologically ac ceptablestudies conducted in recent years.

Miller, Clancy, and Cummings (1953) assigned 30 chroniccatatonic schizophrenics to genuine ECT or to pentothalanesthesia with or without the addition of nonconvulsiveelectrical stimulation. Partially blind assessment (two of the

P.25

four interviewers were blind) after 3 to 4 weeks oftreatment showed no differences among the three groups forreduction of psychotic symptoms or for improvement insocial performance. Brill et al. (1959)

compared 20 genuine ECTs with an equal number ofthiopental or nitrous oxide anesthesia inductions in 67 malechronic schizophrenics and found no significant differenceamong the methods, as blindly assessed on 3 separateoutcome measures 1 month after treatment. Heath, Adams,and Wakelin (1964) gave short courses of 4 or 8 genuine orsham (thiopental anesthesia) ECTs to 45 chronicschizophrenics and found no significant changes orintergroup differences 1 month after treatment on a blindlyad ministered nurses' behavior rating scale.

Langsley, Enterline, and Hickerson (1959) randomly assigned106 acutely schizophrenic or manic patients to a course of12 to 20 ECTs or 200 to 2000 mg/day chlorpromazine (CPZ)(mean = 800 mg/day). Blind evaluation at 8 and 12 weeksrevealed no between-group differences on either apsychiatrist's or nurse's rating scale. King (1960) randomlyassigned 84 newly admitted female schizophrenics to acourse of 20 ECTs or 900 to 1200 mg/day CPZ for onemonth. Hospital discharge rates were the same for bothgroups, as were the subsequent relapse rates while onmainte nance CPZ.

It is reasonable to conclude from these data that ECT is nobetter than sham ECT in the treatment of chronicschizophrenia and no better than neuroleptic agents in thetreatment of nonchronic schizophrenia.

Modern Studies Including a ShamElectroconvulsive Therapy GroupTaylor and Fleminger (1980) studied 20 paranoidschizophrenic patients diagnosed according to the PresentState Examination and referred for ECT after having failed atleast 2 weeks of low-dose therapy with neuroleptic agents(e.g., 300 mg/day CPZ, 15 mg/day trifluoperazine).Chronically ill patients were excluded from this study, aswere those who had a short (6 months) history of psychiatricproblems. Patients were randomly assigned to a course of 8to 12 genuine or sham ECTs (10 in each group),

P.26

administered thrice-weekly, during which neuroleptic drugswere continued at the same low pre-ECT dosages in bothgroups. Blind ratings on the Comprehensive PsychiatricRating Scale revealed lower scores for the genuine ECTgroup at 2 weeks, 4 weeks, and 8 weeks, but not 1 monthafter the treatment course ended. Half of the patients ineach group had pretreatment Beck Depression Inventoryscores of 20, indicative of clinical depression; althoughgenuine ECT caused a greater reduction in these scorescompared with sham ECT at 2, 4, and 8 weeks, thedifferences were not quite signif icant.

In a separately published part of the Leicester ECT trialdescribed earlier (Brandon et al., 1984), Brandon et al.(1985) randomly assigned 17 patients who were diagnosedby the PSE-based Catego program as having schizophreniato receive 8 genuine (n = 9) or sham (n = 8) ECTsadministered over a 4-week course. Those patients alreadyon stable doses

of neuroleptic drugs were continued on them; there was nodifference in mean dosage between the groups. Blindpsychiatric evaluations on the Montgomery-AsbergSchizophrenia Scale at 2 and 4 weeks showed significantlygreater improvement with genuine compared with sham ECT;this was no longer the case at the 12-and 28-week follow-up examinations. No effort was made in this trial to excludepatients with affective symptoms: Mean Hamilton Depressionscale scores were 26 and 37 for the genuine and sham ECTgroups, respectively, well within the range of most studiesof ECT in major depressive disorders.

The small sample sizes and failure to exclude patients withprominent affective symptoms limits the conclusions that canbe drawn from these 2 studies: They both clearlydemonstrate a therapeutic effect of ECT in nonchronicschizophrenic patients receiving neuroleptic drugs anddiagnosed according to modern British criteria.

Bagadia et al. (1983) randomly assigned 38 predominantlynonchronic schizophrenic patients diagnosed according to theResearch Diagnostic Criteria (RDC) to receive either 6genuine bilateral ECTs plus placebo (n = 20) or 6 sham ECTplus 600 mg/day CPZ (n = 18). Blind evaluation on the Brief

P.27

Psychiatric (BPRS) and Clinical Global Impressions ratingscales after 7 and 20 days of treatment revealed nosignificant between-group differences. This study is notablefor its larger sample size and for excluding patients who hadexhibited depressive or manic symptoms sufficient for adiagnosis of schizoaffective or affective illness. Although thenumber of ECTs given was small by any standard, the studydesign demonstrates that a short course of ECT is no moreor less effective a treatment for schizophrenia than anequally short course of moderate-dose neuroleptic treatment.

Studies Without a ShamElectroconvulsive Therapy ControlGroupJanakiramaiah, Channabasavanna, and Murthy (1982)randomly assigned groups of 15 schizophrenic patients eachto receive 6 weeks of treatment with 1 of 4 methods: ECTplus 500 mg/day CPZ; ECT plus 300 mg/day CPZ; 500mg/day CPZ; or 300 mg/day CPZ. Diagnoses were madeaccording to the RDC. Eight to 15 bilateral sine-wave ECTswere administered three times per week. Blind ratings onthe BPRS at weekly intervals were significantly differentamong the four methods by analysis of variance at the firstthrough fifth weeks of treatment: the ECT plus 500 mg/dayCPZ group was the most effective at the end of the firstweek, and the 300 mg/day CPZ group fared worse than anyof the others at the 2-to 5-week assessments. In the maineffects analysis, ECT always resulted in lower BPRS scorescompared with no ECT (regardless of CPZ dose), but thesedifferences only reached significance at the end of thesecond and third weeks of treatment. The interactive effectof CPZ and ECT, that is, the efficacy of combined treatmentcompared with either treatment given separately, wassignificant

at the end of the third through fifth treatment weeks. (Thestudy is marred by the fact that 5 patients in the groupreceiving ECT plus 300 mg/day CPZ revealed to the â!œblindâ! ! examiner that they were receiving ECT.) In sum,this study shows that although at different times during thetreatment course ECT was better than no ECT, and ECT plusCPZ was better than either treatment given alone, by theend of 6 weeks, schizophrenic patients fared equally well

with or without ECT. Essentially, ECT served to acceleratethe treatment response to low-dose CPZ.

The literature is best summarized by the followingstatements:

1. ECT is no better than sham ECT in the treatment ofchronic schizo phrenia.

2. ECT is better than sham ECT in the treatment ofnonchronic schizo phrenic patients who have manyaffective symptoms.

Efficacy in Other DisordersNo data exist from controlled trials to support the use ofECT in disorders other than those described earlier; suchusage belongs to the art rather than to the science ofpsychiatry. Of course, such art has an important place: 111patients must be treated, emergencies responded to, andfamilies assured that no reasonable treatment alternativeshave been overlooked. In truth, most suffering patients areprimarily interested in seeking out a physician with greatexperience and reputed success in the treatment of theirillness, correctly believing that if the physician is alsohonest, and learned, he will choose the most effectivemethod of cure. How much does it really matter to thepatient or doctor that data from controlled trials do not existto support the use of ECT in the particular disorder underconsideration? If a clinician has successfully used ECT undersimilar clinical circumstances in the past, he has amplejustification for yet another therapeutic trial, especially whenother treatments have already failed.

Clinical ConsiderationsThis section is a personal interpretation and elaboration ofthe anecdotal clinical lore that has evolved during more thana half -century of ECT practice. Such uncontrolledobservations are useful for the following reasons:

1. Controlled trials of ECT do not exist for all diagnoses.Some psychiatric syndromes for which ECT is oftenprescribed (e.g., mania) have not yet been been the

P.28

subject of flawless controlled trials, and others (e.g.,catatonic stupor, depressive pseudodementia) occurinfrequently enough to make it doubtful that they everwill be. Yet the universal clinical experience has beenthat manics, catatonics, and depressive pseudodements

respond to ECT at least as well as to other treatments,even when they have failed to respond to intensivepharmacotherapy, and occasionally achieve dramaticremissions that surprise even ardent foes of ECT.

2. The results of controlled trials may be conflicting: Forexample, the carefully controlled and sharplycontradictory study of Wilson and Gottlieb (1967),which clearly demonstrates that right-unilateral ECTcauses more verbal impairment than bilateral ECT, orLambourn and Gill's (1978) methodologicallyimpeccable demonstration that sham ECT is just aseffective in relieving depressions as the genuine article.After attempting in vain to incorporate these and othersimilarly contradictory studies into a comprehensiveview of the ECT process, most writers have simplychosen to ignore them, albeit without scientific groundsfor doing so. The point is that practicing clinicians, aswell as their more rigorously scientific colleagues, pickand choose daily from among the available data thoseresults that best support their personal biases andexperience and reject those that do not.

3. Controlled trials do not have a monopoly on truth. Thissentiment may seem strange in view of the openingsentence of this chapter, but only because truth is herebeing considered at a different level of discourse.Double-blind, random-assignment methodology was notrequired to demonstrate the efficacy of penicillin inmeningococcal meningitis and will not be needed toprove the efficacy of the first drug that cures a fewpatients with acquired immune deficiency syndrome.When a drowned boy is restored to life through phasedwarming and the intravenous administration ofcomplexly balanced electrolyte solutions, no onesuggests the need for a placebo-controlled study toconfirm the results. Likewise, the dramatic response ofa mute, stuporous, rigid, incontinent, and drooling

P.29

catatonic patient to one or two induced seizures alsopartakes of the truth, a truth that has generally beenmore difficult to accept with regard to ECT than withmost other treatments.

4. Data from controlled trials may be incomplete. Somepsychiatric disorders (e.g., melancholia), althoughsubjected to controlled study for responsiveness toECT, have protean manifestations that have not alwaysbeen examined with enough thoroughness to drawdefinitive conclusions.

Finally, although much of this volume is devoted to a criticalreview of the supporting research data for each topiccovered, it is nevertheless intended as a clinical guide toECT, and clinicians can never allow themselves to be boundsolely by the narrow confines of data from controlled trials.This is simply because many, if not most, psychiatricpatients present with syndromes that do not fit the nicelydefined categories of research studies that have the luxuryof specifying inclusion and exclusion criteria and minimalscores on standardized rating scales. For these patients,there simply are no definitive research data to guide theclinician's choice of treatment, yet treat he must. As is truefor anyone who publishes frequently on the topic of ECT, Ireceive numerous queries from clinicians requesting

advice on the management of a particularly difficult patient.The question is never â!œWhat objective data fromcontrolled trials are there to support the use of ECT in mypatient,â! ! but always â!œIn your clinical experience, howshould my patient be treated?â! ! The present chapteraddresses the latter class of questions, relying not only onan extensive clinical case-report literature but also on thecumulative wisdom of teachers and colleagues, all filteredthrough the residue of personal clinical experience.

Choice and Timing ofElectroconvulsive TherapyIt is no secret that patients with affective disorder, unipolaror bipolar types, are prime candidates for ECT; however,despite the chastening lessons of the cross-national study(Kendell et al., 1971) and several articles documenting the

P.30

pitfalls of misdiagnosing affective disorder as schizophrenia(Lipkin, Dyrud, and Meyer, 1970; Carlson and Goodwin,1973; Taylor and Abrams, 1975; Abrams and Taylor, 1974a,1976c, 1981; Pope and Lipinsky, 1978; Pope et al., 1980),many affectively disordered patients who might otherwisehave fully recovered or substantially benefited from ECT (orlithium, for that matter) still receive neuroleptic drugs in themistaken belief that their psychotic symptoms indicate adiagnosis of schizophrenia. Melancholia often presents with astereotypic syndrome that is easy to recognize: Agitation orretardation, weight loss, early waking, self-reproach,anhedonia, impaired concentration, low self-esteem, andruminations of guilt, worthlessness, hopelessness, or suicidepresent an unmistakable clinical picture that augurs well forfull remission with ECT. The presence of such psychoticfeatures as delusions (e.g., of guilt, sin, poverty) orhallucinations (e.g., a rotting odor, a voice counselingsuicide), far from suggesting a diagnosis of schizophrenia,only further cements the diagnosis of melancholia. If adelusional mood is also present or the patient appearsdazed, perplexed, or clouded, the effects of ECT can bedramatic, with the patient awakening from the first orsecond treatment as from a dream, thoroughly astounded tolearn of his whereabouts and recent strange behaviors.

Although 6 to 8 ECTs (mean = 6.5) constitute the modalrange for a treatment course in melancholia (Fink, 1979), anoccasional patient may require considerably more. Iremember well 1 woman in her early 70s whose severeunipolar depression did not respond at all to the first 10ECTs. Treatment was continued despite these disappointingresults because of her classical presentation and a history ofan excellent response of similar symptoms to ECT 30 yearsearlier. Improvement became evident by the 12th ECT andwas complete after the 15th, illustrating the general rulethat as long as the clinical syndrome is prognosticallyfavorable, ECT should be continued until the expecteddegree of improvement is obtained. Although there isprobably a maximum number of ECTs that should not beexceeded in a single treatment course, I know of no way todetermine it. (I have seen only 1 patient with classicmelancholia who did not respond to ECT, a late-onsetunipolar

depressive whose presentation was so typical and whosesymptoms so severe that he received 22 ECTs withoutsignificant improvement before the treatment was finallydeclared a failure.)

Conversely, there is no reason to adhere slavishly to aminimum number of treatments. I can think of severalinstances, especially in older patients, where markedimprovement occurred after the third or fourth ECT, only tofade and then disappear with additional treatments. Whetherthis course of events represents a dose-response curveanalagous to that reported for the tricyclic antidepressantnortriptyline or simply the deleterious effects of developingcognitive dysfunction, the advisable course of action when apatient has shown marked improvement early in thetreatment course is to withhold further ECT pending thereturn of symptoms. The rationale for the common clinicalpractice of giving two additional ECTs to prevent relapseafter full recovery has been achieved was not confirmed in acontrolled follow-up study by Barton, Mehta, and Smith(1973). Finally, some depressed patients who do not respondto the usual rate of administration of 3 times a week maynonetheless recover if double ECTs are given each session(Swartz and Mehta, 1986).

Of all the possible behavioral responses to ECT, theeuphoric -hypomanic pattern (Fink and Kahn, 1961) is best.In my view, it always indicates that enough ECT has beengiven. Its occurrence after right-unilateral ECT as well asafter bilateral ECT suggests that it represents not merely anonspecific organic frontal lobe response, but rather afocused limbic effect of ECT (a direct hit), producing anaffective overshoot with gradual return to euthymia severaldays after treatment has stopped. The relation of thissyndrome to mania is discussed elsewhere in this volume.

Antidepressants should be discontinued during ECT. Noadditive or synergistic effect of combining ECT andantidepressants has been demonstrated (Abrams, 1975;Siris, Glassman, and Stetner, 1982), and Price et al. (1978)actually found a statistically significant reduction in affectiveimprovement in patients in their ECT sample who hadreceived concomitant tricyclic antidepressants compared withthose who had not. Lithium, as noted in Chapter 8, shouldnot be coadministered with ECT because it may cause severe

P.31

organic confusional states and prolong the apnea induced bysucci nylcholine.

Two clinical variants of melancholia are exquisitelyresponsive to ECT: catatonic stupor and depressivepseudodementia.

1. Catatonic stupor: Known in the older literature asmelancholia attonita, this syndrome of stupor, mutism,negativism, catalepsy, and incontinence of saliva,urine, and feces may linger for weeks or months untilabruptly dissolved by a few (sometimes, just one)induced seizures. Again, a diagnosis of schizophrenia isnot inherent to this syndrome, which is diagnosticallynonspecific and occurs most frequently in patients whosatisfy research criteria for affective disorders (Abramsand Taylor, 1976c).

Other individual catatonic features (e.g., echolalia,echopraxia, manner isms, stereotypies) are notassociated with a particularly good response to ECT,perhaps because they are more often seen in patientswith chronic schizophrenia. A positive, albeit transient,response of catatonic stupor to intravenous sodiumamobarbital (the â!œamytal interviewâ! !) often predictsa favorable outcome with ECT. Indeed, while awaitingcompletion of the pretreatment workup before startingECT in a catatonic patient, 250 mg of sodium amytalgiven intramuscularly 30 minutes before meals willoften enable him to eat and drink. Parenteralbenzodiazepinesâ!”e.g., lorazepamâ!”are also effectivefor this purpose.

2. Depressive pseudodementia (dementia syndrome of

depression): If this syndrome of disorientation andimpaired memory accompanying depressive symptomsin an older person is misdiagnosed as senile dementiaand the luckless patient is placed in a nursing home,he may not have long to live. A few ECTs, however,can rapidly restore such an apparently deterioratedindividual to rosy health, an occurrence rendered allthe more remarkable by the family's frequent commentthat â!œhasn't looked this well in years.â! ! One of themost striking examples of this phenomenon, dubbed â

P.32

!œthe Rip Van Winkle syndrome,â! ! occurred in a 72-year-old man (Fisman, 1988), who recovered fully froma â!œpresenile dementiaâ! ! with a course of ECTs given14 years after the diagnosis was made and the patientinterred in a nursing home. In their literature review,Price and McAllister (1989) found 10 reports describinga total of 22 cases of depressive pseudodementiatreated with ECT (mean age = 64.2 years), all of whomimproved, with only 23% showing significant ECT-induced cognitive impairment.

There is little risk in inadvertently treating a patient whosedepressive symptoms are only an early manifestation ofAlzheimer's dementia. ECT has been used successfully undersuch circumstances without worsening the cognitivesymptoms or accelerating the progression of the underlyingdis order (see Chapter 5), so there is nothing to be lost inquestionable cases by a therapeutic trial.

A melancholic patient should receive ECT in preference toany other treatment under circumstances of increasedclinical urgency or intolerance to psychotropic drugs:

1. The presence of delusions or hallucinations: Psychoticdepression is notoriously resistant to antidepressantdrugs but very responsive to ECT (Hordern et al, 1963;Glassman, Kantor, and Shostak, 1975; Davidson et al.,1978; Crow and Johnstone, 1986; Petrides et al., inpress). Although tricyclic-neuroleptic combinations arereported to be more effective in delusional depressionthan tricyclics alone (Spiker et al., 1986), it seemsunwarranted to expose a depressed patient to the risksof tardive dyski

nesia and neuroleptic malignant syndrome when thesafe and rapidly-effective alternative of ECT isavailable.

2. The presence of stupor: Neither full-blown catatonicstupor nor the more severe degrees of psychomotorretardation respond well to antidepressant drugs, yetboth are rapidly responsive to ECT. Moreover, there isoften some urgency involved, because such patients donot eat well, or at all, and have often lost significant

body mass by the time they are first seen.

3. The presence of suicidal ruminations or behavior:

Completed suicide presents the single greatest risk inmelancholia, yet abundant clinical observations suggest

it rarely occurs acutely once ECT has been initiated.1

No similar observations obtain for antidepressantdrugs, which are traditionally held to increase the riskof completed suicide early in the treatment course ifpsychomotor retardation is relieved before despondency.

4. Coexisting severe medical disease: The pronouncedcardiovascular effects of some tricyclic antidepressantshave been reported to increase the risk of suddendeath in cardiac patients (Coull et al., 1970; Moir etal., 1972), making ECT often seem to be the moreconservative treatment choice. Hepatic and renaldisease also increase the risks of drug therapy, due toimpaired metabolism and excretion, but not the riskwith ECT.

5. Depressive pseudodementia: Antidepressants seem onlyto aggravate this syndrome, presumably because oftheir anticholinergic effects.

6. Old age: Elderly melancholies also seem particularlyintolerant to the anticholinergic effects of manytricyclics, which often cause severe constipation oreven precipitate anticholinergic delirium. If anything,the response to ECT improves with the age of thepatient, and fewer treatments are generally required toachieve remission.

7. Pregnancy: All psychotropic drugs, including lithium,cross the placental barrier and exert unknown, butdoubtless protean and unfavorable effects on the fetusboth during and after development, continuing into thepostnatal period if the mother nurses. ECT is notknown to exert any such adverse fetal effects (Chapter5) and should be used in preference to drugs inpregnant women and during the nursing period.

8. Drug therapy failure: Any patient who has failed acourse of adequate antidepressant therapy should beoffered ECT in preference to another trial with a

P.33

different compound. In practice, this covers manydepressives who are admitted to the hospital afterfailing to respond to outpatient pharmacotherapy. Ofcourse, any patient with a history of previousunresponsiveness to antidepressants should receive ECTas the initial treatment.

9. Patient preference: The 2001 APA Task Force on ECTreport (American Psychiatric Association, 2001) alsothoughtfully includes patient preference among theindications for the primary use of ECT, somethingpsychiatrists have paid surprisingly little attention towhen administering biologic treatments.

ManiaMania, the obverse of melancholia, responds so well to ECTthat it is difficult to account for the absence until recentlyof any prospective controlled trials of induced seizures inthis disorder. Perhaps the rapidly spreading use (andremarkable efficacy) of lithium therapy at a time when theimportance of controlled trials of ECT was also beingrecognized had an inhibiting effect; probably the generaldisinclination of manics to cooperate with any form oftreatment, let alone random assignment to ECT or drugs,also played a role. At the time of this writing, the study ofSmall et al. (1988) cited earlier remains the only publishedcontrolled trial of ECT in mania.

Several retrospective studies shed light on the efficacy ofECT in mania. In a chart-review study, McCabe (1976)compared a sample of manic patients who received ECT inthe predrug era (1945-1949) with an age-and sex-matchedcontrol sample of manics who were hospitalized at the IowaPsychopathic Hospital before ECT had been introduced(1935-1941). All subjects were selected on the basis of thesame research criteria, and the resultant groups wereremarkably similar on most of 27 clinical psychopathologicvariables studied. On all outcome measures (duration ofhospitalization, condition at discharge, percent dischargedhome, degree of social recovery), the ECT-treated groupfared substantially and significantly better than theuntreated control sample, with the most striking difference

P.34

being that 96 of the ECT patients were discharged to theirhomes, compared with only 44 of the untreated patients.

A second chart-review study from the same hospital (McCabeand Norris, 1977) examined the question from a differentvantage point, comparing the outcome in the same twogroups already described with that obtained in a third age-and sex-matched sample of manics who received CPZtherapy from 1958 to 1964. Not surprisingly, ECT and CPZwere both superior to no treatment and were about equaloverall in their beneficial effects in mania; however, 10patients who did not respond well to CPZ therapy went on torecover with ECT. In another retrospective chart-reviewstudy (Thomas and Reddy, 1982), ECT, CPZ, and lithiumwere reported to be equally effective; however, the threegroups were not as well matched compared with groups inthe studies from Iowa, and the sample sizes (10 in eachgroup) were much smaller.

Black, Winokur, and Nasrallah (1987) reviewed the charts of438 patients hospitalized for mania over a 12-year periodand found that a significantly greater percentage whoreceived ECT could be classed as having â!œmarkedimprovementâ! ! compared with those who received adequateor in adequate lithium therapy or no therapy at all.Unilateral and bilateral ECT were equally effective.

In a review of all patients receiving ECT at McLean Hospitalfrom 1973 to 1986, Alexander et al. (1988) found that 10 of18 manic patients (56%) were significantly improved (n = 9)or recovered (n = 1) with ECT. A similar review at AarhusHospital in 1984 (Stromgren, 1988) revealed a

far more salutary response: ECT induced a moderate (n = 6)or satisfactory (n = 10) effect in mania in 16 out of 17series administered, virtually all with right unilateral ECT.

Most manic patients come to ECT only after treatment withlithium or neuroleptic drugs has failed and the patient, whohas typically been in a state of relentless excitement forseveral days, is on the verge of exhaustion. Even underthese unfavorable circumstances, ECT works. Other thanpsychomotor excitement, a particularly favorable featurethat is usually present in patients with acute mania, noindividual manic symptom or symptom-cluster is especially

predictive of a good response to ECT. Conversely, thepresence of psychotic symptoms, however outlandish orbizarre, in no way reduces the likelihood of recovery as longas the full manic syndrome is also present (Taylor andAbrams, 1975).

It has long been standard clinical practice to administerdouble bilateral ECTs on consecutive treatment days duringthe first session or two of a manic patient's treatmentcourse, perhaps reflecting the clinical urgency often felt bythe time such patients are finally referred for ECT. Recentevidence (Small et al., 1988), however, suggests thatmanics are responsive to conventional single bilateral ECTswhen administered at the usual rate (al beit with concurrentneuroleptic therapy).

SchizophreniaThere is little doubt that many patients diagnosed as havingacute or schizo-affective schizophrenia respond remarkablywell to ECT; there is also little doubt that most of thesepatients are misdiagnosed manics (Abrams and Taylor1974a, 1976c, 1981; Taylor and Abrams, 1975). When thediagnosis of schizophrenia is made by first excludingpatients with prominent affective syndromes (Taylor andAbrams, 1978), most of the ECT-responsive clinical varianceis thereby also excluded. This should not be taken to meanthat patients with an early, insidious onset of emotionalblunting, avolition, first-rank symptoms, and formal thoughtdisorder should never be offered ECT. On the contrary,every such patient deserves one full trial of ECT (preferablyearlier rather than later in their illness course) to insurethat no treatment will be overlooked that has a chance,however slim, of halting the otherwise relentless progressionof this devastating illness (Abrams, 1987). Two uncontrolledstudies (Friedel, 1986; Gujavarty, Greenberg, and Fink,1987) suggest that such a trial, in conjunction withneuroleptic drug therapy, may yield unexpectedly favorableresults. Controlled studies, with and withoutcoadministration of neuroleptic drugs, should now beundertaken to dem onstrate the efficacy of such treatment.

In those rare instances when the catatonic syndrome ofnegativistic stupor is a manifestation of schizophrenia, ECTworks just as well to remove the stupor as it does in

P.35

patients with affective disorder, but often reveals a coreschizophrenic syndrome that is resistant to further ECT.

Symptomatic PsychosesNumerous individual case reports over the last half -centurytestify to the effectiveness of ECT in psychotic statessecondary to a wide variety of toxic, metabolic, infectious,traumatic, neoplastic, epileptic, and endocrine disorders(Taylor, 1982). Particularly striking are the results that areoften achieved with ECT in drug-induced states, especiallyamphetamine psychosis, and in the acute epileptiformpsychoses. A favorable response is less likely in cases inwhich the underlying disorder has been chronically in placeor remains uncorrected at the time of treatment. In general,it is reasonable to reserve ECT for those patients withsymptomatic psychoses that have neither responded tomedical treatment of the underlying condition nor to aweek's trial of neuroleptic drugs.

Neuroses and Personality DisordersRegardless of the current terminology used to classify thesedisorders, they are rarely responsive to (and oftenaggravated by) ECT. This is particularly true for anxiety-related syndromes such as panic disorder.

DeliriumFor decades, European psychiatrists have used ECT to treatacute delirious states secondary to alcohol and sedativewithdrawal, toxic and febrile states of various etiologies(Roberts, 1963; Kramp and Bolwig, 1981; Taylor, 1982),and those occurring in the context of severe mental illness,such as delirious mania and malignant catatonia (Stromgren,1997). Unfortunately, although the clinical experience(including my own) with this use of ECT can be compelling(Fink, 1999a), it remains unsupported by prospective clinicaltrials. Nevertheless, a systematically presented retrospectiveseries of 14 cases of delirious mania successfully treatedwith ECT over a 6-year period illustrates the dramaticresults that can be obtained in such patients, who typicallypresent with â!œthe acute onset of the excitement,

P.36

grandiosity, emotional lability, delusions, and insomniacharacteristic of mania, and the disorientation and alteredconsciousness characteristic of deliriumâ! ! (Fink, 1999b).

Use of Electroconvulsive Therapy inChildren and AdolescentsNeither the first APA ECT Task Force Report (AmericanPsychiatric Association, 1978) nor the Consensus Conferenceon ECT (National Institutes of Health, 1985) address theindications for ECT in children or adolescents, and the thirdAPA ECT Task Force Report (American PsychiatricAssociation, 2001)

is not very helpful either. Excluding fully postpubertaladolescents (e.g., age 16 years and above) fromconsideration as not different from adults in any medicallyrelevant way leaves an extremely sparse literature remainingon the subject. Further excluding the older literature on theuse of induced seizures in autism, childhood schizophrenia,and other childhood encephalopathic disorders leaves but ahandful of cases, only 3 of which were prepubertal, toprovide guidance.

Warnecke (1975) reported the successful use of ECT to treata severe depression that emerged in a 14-year-old boy afterhe had an acute manic episode that had responded well toneuroleptic drugs.

Carr et al. (1983) used right unilateral ECT to treat a 12-year-old prepubertal girl who met criteria for bipolaraffective disorder, manic type. Seven ECTs induced fullremission of her pronounced and chronic manic psychosis,despite a recent history of asymmetric, left -sided EEGslowing and enlargement of the frontal horns and bodies ofthe lateral ventricles on computed tomographic (CT) scan.

Black and colleagues (1985) described an excellent responseto unilateral ECT in an 11-year-old prepubertal boy withmajor depression, and Powell, Silveira, and Lindsay (1988)obtained similarly favorable results with unilateral ECTduring three separate episodes of recurrent familialdepressive stupor in a 13-year-old prepubertal boy whoremained well over a 4year follow-up interval. Bertagnoliand Borchardt (1990) also successfully used unilateral ECT

P.37

to treat a 15-year-old adolescent girl suffering from bipolaraffective disorder; she remained symptom-free onmaintenance lithium therapy during a 9-month follow-upinterval.

Guttmacher and Cretella's (1988) survey of 10 years ofexperience with ECT at Strong Memorial Hospital is puzzling,both for the generally unfavorable outcome and theastoundingly high incidence (75%) of prolonged seizuresrecorded in the 4 patients, aged 12 to 15 years, whoreceived ECT during that time. The 3 patients who failed torespond were hardly prime candidates for ECT: One hadflattened affect and poverty of speech; another was sociallyisolated, emotionally blunted, and unresponsive totherapeutic levels of desipramine; and the third had ahistory of delayed language onset, a facial seizure,incontinence and echolalia, made guttural noises, had failedto respond to successive trials of imipramine, haloperidol,clonidine, clonazepam, and monoamine oxidase inhibitors,and was considered a candidate for psychosurgery. Even thesole designated responder had his ECT discontinued after 16treatments because â!œthe treatment team felt it to beineffectiveâ! ! (improvement began only 7 days after his lasttreatment). Two of the prolonged seizures occurred inpatients who were also receiving medications known to lowerthe seizure threshold: desipramine and trifluoperazine; thethird was observed in the brain-damaged patient withepilepsy.

The survey methodology may have elicited this unusualoutcome by asking physicians â!œif they had ever treated apatient aged 15 years or younger and if they had ever had apatient who experienced a seizure lasting more than 4minutes,â! ! thus potentially biasing respondents to recallselectively

those children and adolescents with prolonged seizures, inwhom preexisting brain pathologic conditions may have beenresponsible both for treatment resistance and prolongedseizures.

One of the most striking cases was reported by Cizlado andWheaton (1995), who administered a course of 19 ECTs (15bilateral, 4 unilateral) to an 8.5-year-old girl who developeda severe syndrome of catatonic stupor while undergoing

antidepressant drug therapy for major depression. Clinicalimprovement was not observed until the 8th treatment,following which a neuroleptic was added to the treatmentregimen (without any apparent medical basis). After the llthECT, dramatic improvement was observed with eachadditional seizure through the 16th, when full recovery wasobtained. At 6-month follow-up, she remained well onmaintenance fluoxetine therapy.

In an open, prospective study of all adolescents treated withECT during a 12-year period at a University hospital,Strober, Rao, and De Antonio (1998) reported the resultsobtained in 10 treatment -refractory, psychotically depressedadolescents, half of whom were under age 16 at the time oftreatment. All but one demonstrated dramatic improvementthat was maintained at 1-and 6-month post-ECT assessmentintervals, at which times they were also receiving a varietyof psychopharmacologic maintenance treatments.

From the available data, then, it is apparent that when theclinical indications are straightforward and the prognosisfavorable, as in unipolar or bipolar affective disorder, theresponse to ECT in children and adolescents is no differentfrom that obtained in adults: excellent. Likewise, littlebenefit is to be expected from ECT in patients, regardless ofage, whose chronic nonaffective psychoses are characterizedby treatment resistance, emotional blunting, social isolation,and the stigmata of coarse brain disease.

Nonpsychiatric Indications forElectroconvulsive Therapy

Parkinson's DiseaseParkinson's disease is a neurodegenerative disorder thataffects 1% of the population over age 50â!”approximatelyhalf a million Americans. Pharmacologic treatments includedopaminergic agents such as levodopa (usually combinedwith the peripheral decarboxylase inhibitor carbidopa),amantadine, bromocriptine, pergolide, and selegiline, as wellas anticholinergic agents such as benztropine andtrihexyphenidyl. When they are first administered, theseagents all tend to be effective. However, they are associatedwith numerous side effects (including psychosis and mooddisorders) and typically lose efficacy as the disease

P.38

progresses. With levodopaâ!”the standard pharmacologictherapy for Parkinson's diseaseâ!”severe, abrupt, and oftenincapacitating fluctuations in extrapyramidal symptoms (theâ!œon-offâ! !

syndrome) ultimately develop in virtually all patients takingthe medication who survive long enough.

In addition to these pharmacologic agents, autologousadrenal medullary and fetal brain tissues have beentransplanted into the caudate nucleus at considerable riskand expense, but with limited success (Abrams, 1989b;Goetz et al., 1989; Ahlskog et al., 1990).

For over 30 years, ECT has been reported to be helpful inmany dozens of patients with Parkinson's disease whoreceived this therapy for a variety of psychiatric indications,including depression, mania, and drug-induced psychosis(Rasmussen and Abrams, 1991, 1992). There is alsosubstantial evidence for the antiparkinsonian effect of ECT innonpsychiatric patients, leading recent editorials torecommend the use of ECT in selected Parkinson's diseasepatients (Fink, 1988; Abrams, 1989b).

The first clinical report was from Fromm (1959), who treated8 patients with Parkinson's disease and unspecifiedpsychopathologic symptoms, who had received variousantiparkinsonian medications and showed no sustainedimprovement. His patients, who ranged in age from 40 to 69years with a duration of symptoms of 5 to 20 years, allexhibited prominent rigidity, bradykinesia, and bradyphrenia.After 5 to 6 ECT treatments, 5 patients improved markedly,and some previously bedridden patients were able to walk.Improvement generally began after the first treatment, wasmaximally observed after the second or third, and wassustained for 2 to 3 months before relapse. Prominenttremor was a poor prognostic sign in this sample.

Virtually all subsequent reports have been equally sanguineand have been reviewed in detail elsewhere (Rasmussen andAbrams, 1991, 1992).

In the most systematic clinical presentation to date, Douyonet al. (1989) provided extensive quantitative data for 7patients treated for depression with ECT. All had received

P.39

levodopa/carbidopa before ECT and were continued on thisregimen during the course of treatment. Ages ranged from61 to 73 years, with an average duration of Parkinson'sdisease of about 9 years. Bilateral ECT given at just aboveseizure threshold yielded substantial improvements in all offive subscales of an extrapyramidal symptoms rating scalethat included postural stability, gait, tremor, bradykinesia,and rigidity. These effects appeared after only twotreatments and generally continued for up to five moretreatments, although one patient had only minimalimprovement.

Depressed mood improved in all. Two patients developedtreatment -emergent dyskinesias that resolved when theirlevodopa doses were halved. While patients were onmaintenance levodopa/carbidopa, relapses occurred at 4weeks in 1 patient, 6 weeks in another, and not at all in 4patients during a 6-month follow-up. Interestingly, therewas a high positive association (r = 0.917) between the ageof the patient and the scored degree of ECT-inducedimprovement.

In addition to the numerous case reports and small series, agroup of Swedish investigators has used ECT as the primarytreatment for Parkinson's disease without concomitantpsychopathologic symptoms. Balldin et al. (1980b, 1981)

used bilateral ECT to treat 9 such patients, age 52 to 70years, with a duration of Parkinson's disease of 6 to 20years. All patients had been on levodopa and had developedthe on-off syndrome. The investigators quantified thepercent of â!œonâ! ! and â!œoffâ! ! time before and after thecourses of ECT. After 5 to 6 treatments, a marked reductionin â!œoffâ! ! time was observed for 4 to 41 weeks in 5patients; 4 patients experienced little if any benefit. As inthe report by Douyon et al. (1989), there was a highpositive correlation between age and degree of improvement.Duration and amount of levodopa treatment also correlatedsubstantially with ECT response.

In the methodologically most compelling study to date, alsofrom the Swedish group (Andersen et al., 1987), 11nondepressed, nondemented Parkinson's disease patientswere randomly assigned to receive either genuine or shamECT, the latter consisting of anesthesia induction without

subsequent electrical stimulation. Patient ages ranged from54 to 81 years, with a 6-to 32-year history of Parkinson'sdisease. All patients had been on levodopa and haddeveloped the on-off phenomenon. Patients treated withgenuine ECT enjoyed a greater reduction in â!œoffâ! ! timethan those receiving sham ECT. Most patients were treatedwith bilateral ECT; of the 2 who received unilateral ECT,only 1 experienced substantial reduction in â!œoffâ! ! time.For the 9 patients who experienced a marked reduction inextrapyramidal symptoms with genuine ECT, improvementlasted from 2 to 6 weeks.

Thus, numerous case reports of psychiatric patients withParkinson's disease demonstrate a substantial andoccasionally long-lived antiparkinsonian effect of ECT.Additionally, three prospective trials of ECTâ!”one of themplacebo-controlledâ!”have confirmed this effect innondepressed, non-demented Parkinson's disease patients.Although systematic clinical data have not been extensivelycollected, it appears that advanced age and severe disability(bedridden, on-off syndrome) may actually be favorableprognostic features for a response to ECT. Althoughtreatment -emergent dyskinesias may occur during ECT inpatients receiving concomitant levodopa therapy, theyrespond to levodopa dosage reduction (prophylactic halvingof levodopa dosage before starting ECT may prevent thistroubling side effect).

Improvement in Parkinson's disease symptoms is reportedwith both bilateral and unilateral ECT and tends to occurafter the first few treatments (Douyon et al., 1989). Fewinvestigators have recorded the stimulus parameters used.Roth, Mukherjee, and Sackeim (1988) achieved success withright unilateral placement and stimuli that were 150% abovethreshold, and Douyon et al. (1989) used just-above-threshold bilateral ECT. Although there are virtually no dataconcerning optimal seizure length in the treatment ofParkinson's disease, it is notable that one of the patients ofDouyon et al. (1989) improved substantially after twoseizures of less than 15 seconds each.

Although sustained improvement in extrapyramidal symptomshas been described after ECT (Birkett, 1988), patientsgenerally enjoy only a few days' to several months'remission, even with maintenance antiparkinsonian

P.40

medication (Balldin et al., 1981; Andersen et al., 1987). Theonly published

allusion to maintenance ECT is from Holcomb, Sternberg,and Heninger (1983), who briefly described a several-monthextended antiparkinsonian benefit of monthly ECTtreatments.

Practical ConsiderationsThe primary indication for ECT in Parkinson's disease isdemonstrated refractoriness to antiparkinsonian medicationsor intolerance to their side effects. Thus, patients withsevere disability (e.g., bedridden, on-off status) will be theusual candidates for this form of therapy. Patients should bethoroughly informed of the tentative nature of the use ofECT for their conditionâ!”the possibilities of limited benefitand rapid relapseâ!”and explicit, written consent should beobtained in each instance. Levodopa doses should bereduced by half to prevent emergent dyskinesias, andadjunctive agents (e.g., anticholinergics, amantidine) shouldbe discontinued before starting ECT.

As is done in the treatment of patients with associateddementia, brief-pulse right unilateral ECT should be usedintially, because it has been reported effective in Parkinson'sdisease and is generally free from significant cognitive sideeffects. If the patient does not respond to the first threetreatments, he should be switched to bilateral (bitemporal orbifrontal) electrode placement. An electrical dosage at least1.5Ã— threshold should be given, and seizure length, as inECT used for psychiatric disorders, should be at least 30seconds of EEG activity.

Optimal antiparkinsonian medication should be reinstitutedas soon as ECT is terminated. This includes return of thelevodopa dosages to previous levels (assuming nointolerable adverse effects) and restarting any previouslyhelpful adjunctive agents. Serious consideration should begiven to initiating maintenance ECT at regular intervals toprevent or delay return of extrapyramidal symptoms. Thiswill require a trial-and-error approach of progressivelyincreasing the intertreatment interval from, say, once aweek in itially, to the longest interval that will effectivelysustain improvement (e.g., monthly).

P.41

Neuroleptic -induced parkinsonism also responds to ECT(Shapiro and Goldberg, 1957; Ananth, Samra, and Kolivakis,1979; Goswami et al., 1989; Chacko and Root, 1983), evenin patients who remain on neuroleptic drugs.

Neuroleptic MalignantSyndrome/Lethal CatatoniaThe neuroleptic malignant syndrome, a potentially fataldisorder, is characterized by the development of fever,rigidity, dysautonomia, stupor, and elevated creatinephosphokinase levels in a patient receiving neurolepticdrugs. The rationale for treating neuroleptic malignantsyndrome with ECT is obscure, but may derive from itsresemblance to â!œfebrileâ! ! or â!œlethalâ! ! catatonia(a.k.a. pernicious catatonia, delirium acutum, Bell's mania,manicdepressive

exhaustion syndrome), a rare and reportedly highly ECT-responsive syndrome (Lotstra, Linkowksi, and Mendlewicz,1983). Indeed, Mann et al. (1990) conceive neurolepticmalignant syndrome to be a neuroleptic-induced, iatrogenicform of lethal catatonia.

Several reviews during the past 20 years have amplydemonstrated the often striking efficacy of ECT in alleviatingneuroleptic malignant syndrome (Greenberg and Gujavarty,1985; Casey, 1987; Mann et al., 1990; Pearlman, 1990;Davis et al., 1991): About 80% of cases respondsignificantly, reducing the untreated mortality rate for thissyndrome by about half, a result similar to that obtainedwith dantrolene, bromocriptine, levodopa, or amantidine(Davis et al., 1991). Several authors report, however, thatthe procedure is not without risk. Regestein et al. (1971)report a 22-year-old man diagnosed as having catatonicstuporâ!”but who clearly met criteria for neurolepticmalignant syndromeâ!”complicated by thrombophlebitis,atrial tachycardia, pneumonitis, and massive pulmonaryembolism that required inferior vena cava clipping.Immediately after his sixth ECT he developed ventricularfibrillation and lapsed into a coma with decerebrateposturing in which he remained at the time the report waswritten 7 months later. This disastrous outcome was

P.42

doubtless aggravated (if not caused) by 5 days of parenteraladministration of chlorpromazine just prior to the course ofECT in an unsuccessful attempt to treat the â!œcatatonia.â! !Hughes (1986) reported a 33-year-old woman withneuroleptic malignant syndrome who developed cardiacarrest during a session of multiple bilateral ECT, from whichshe was successfully resuscitated, and Grigg (1988)described a patient who received ECT 2 days after anepisode of neuroleptic malignant syndrome with coma andwho developed fever, tachycardia, and marked elevation inserum creatine phosphokinase several hours later.Interestingly, none of these 3 patients experiencedsignificant relief of their neuroleptic malignant syndromewith ECT. Three deaths occurred despite intervention withECTâ!”2 of them in patients who continued to receive high-potency neuroleptic drugs â!”whereas ECT benefited thesyndrome in all cases where concomitant neuroleptics werenot used (Davis et al., 1991). Incredibly, a major reviewarticle on the pathogenesis and treatment of neurolepticmalignant syndrome (Ebadi, Pfeiffer, and Murrin, 1990)contains but a single reference on the use of ECT for thisdisorder.

The frequently dramatic efficacy of ECT in patientsdiagnosed as having lethal catatonia (differing fromneuroleptic malignant syndrome only in its temporalproximity to neuroleptic drug administration) has also beenthoroughly documented by Mann et al. (1986), although thecase they reportâ!” as well as the one described by Nolenand Zwaan (1990)â!”is unconvincing because the patienthad developed the initial syndrome after receivinghaloperidol. The fact that each patient then enjoyed aprolonged period of improvementâ!”without furtheradministration of neurolepticsâ!”before again developing thesyndrome hardly seems an adequate basis for claiming theexistence of two separate disorders. More convincing is therecent report of Rummans and Bassingthwaigte (1991), inwhich 8 bilateral ECTs rapidly

reversed a syndrome of severe hyperthermia, bilateralextensor toe reflexes, and decerebrate posturing in a 49-year-old catatonic woman who had not received neurolepticdrugs for over a year.

Lethal catatonia and its equivalents were described andreported to be responsive to ECT long before theintroduction of neuroleptic drug therapy. Because itscharacteristic symptomsâ!”with or without a recent historyof neuroleptic therapyâ!”are so often resistant to oraggravated by neuroleptic therapy and are rapidly sensitiveto the effects of a few induced seizures, ECT is doubtlessthe conservative treatment of choice for this syndrome aswell as its neuroleptic-induced twin.

Pain SyndromesThe only prospective trial of ECT for pain in nonpsychiatricpatients is that of Salmon et al. (1988), who gave 6 rightunilateral ECTs to 4 dextral patients with intractablethalamic pain secondary to right hemisphere strokes, noneof whom experienced any relief.

Notes1The over-all reduction in mortality rate reported forpatients who have received ECT cannot, however, beattributed to a reduction in suicide rate (Prudic andSackeim, 1999).




> Table of Contents > Chapter 3 - Prediction of Response to

Electroconvulsive Therapy

Chapter 3

Prediction of Response to


Prediction in medicine is more aptly termed prognosticationand usually constitutes an assessment of variables thatdetermine the likelihood of developing a given illness (e.g.,myocardial infarction risk factors), surviving one (e.g., thestaging of cancers), or enjoying a favorable outcomefollowing an intervention.

Earlier prognostic methods are now mainly of historicalinterest. These include physiologic measures, such as themethacholine and sedation threshold tests (Funkenstein,Greenblatt, and Solomon, 1952; Shagass and Jones, 1958),and personality variables as assessed by the MinnesotaMultiphasic Personality Inventory (Feldman, 1951),Rorschach test (Kahn and Fink, 1960), and California F scale(Kahn, Pollack, and Fink, 1959). These have been reviewedin detail elsewhere (Fink, 1979; Hamilton, 1982) and are notincluded here because they provide no practical guide to thepresent-day administration of ECT. Somewhat more relevantto modern practice are the prognostic scales derived fromthe clinical and psychopathologic features of the depressedstate, although, as we shall see, they are also of limitedutility.

Hobson (1953) was the first to construct such a predictiveindex, recording the presence or absence of 121 clinical andhistorical features in a sample of 127 patients (â!œalmostallâ! ! of them depressed) before they received ECT. Twoweeks after the treatment course, patients were classified ashaving had a â!œgoodâ! ! or â!œpoorâ! ! response, and eachof the 121 features was examined for its correlation withthis measure. Thirteen features correlated significantly withthe outcome. Five were favorable (sudden onset, goodinsight, obsessional personality, self-reproach, duration of

P.44

illness less than 1 year) and 8 were unfavorable (mild-to -moderate hypochondriasis, depersonalization, emotionallability, adult neurotic traits, hysterical attitude, above-average intelligence, childhood neurotic traits). By addingthree â!œvery suggestiveâ! ! features (a favorable one ofpronounced retardation and two unfavorable ones of ill-adjusted or hysterical personality) to this list, a checklist of16 features was provided and scored by assigning one pointfor the absence of a favorable feature or the presence of anunfavorable one. Scores ranged from 1 to 14, the lower thebetter. A mean score of 7.5 was found to provide the fewestmisclassifications into the good and poor out comegroupings, yielding a â!œhit rateâ! ! of 79.7%.

Roberts (1959a) found that the Hobson index successfullypredicted the outcome in 80% of depressives, and Mendels(1965a) in 78%, whereas Hamilton and White (1960) andAbrams, Volavka, and Fink (1973) found it to be of novalue. Mendels (1965b) also employed an item analysis(similar to Hobson's method) to construct his own index ofall variables associated with a 50% or greater reduction indepression scale score 1 month after ECT. Four favorableitems (family history of depression, early waking, delusions,retardation) and four unfavorable ones (neurotic traits,inadequate personality, precipitating event, emotionallability) correctly predicted the outcome in 90 of the cases;a subsequent study with an enlarged sample and slightlydifferent index (Mendels, 1967) yielded a predictive accuracyof 86%. In this study, personality traits (e.g., histrionic)that were associated with reactive or neurotic depressionwere better predictors of outcome than were the clinicalvariables (e.g., early waking) that were classicallyassociated with a diagnosis of endogenous depression.Abrams, Volavka, and Fink (1973) were unable to confirmthe predictive value of Mendels' (1965b) index in a sampleof 76 primary depressives studied before and after a courseof 4 bilateral or right unilateral ECTs.

Multiple regression analysis is a more sophisticatedprocedure for weighting the prognostic value of variables. Itwas first used for this purpose by Hamilton and White(1960). These authors included five pretreatment variablesin their regression analysis that were selected for theirsignificant correlation with the Hamilton (1960) depressionscale score after ECT: duration of illness, body index,postmethacholine fall of systolic blood pressure, andpretreatment Hamilton depression scale score. The multiplecorrelation achieved was a modest 0.62, somewhat lower,

P.45

correlation achieved was a modest 0.62, somewhat lower,but in the same general range, as that reported bysubsequent investigators using similar techniques. Nystrom(1964) selected 24 variables that were correlated withoutcome and calculated individual prognostic values for eachpatient based on the partial regression coefficients of theitems. Favorable features included early waking, retardation,and a profoundly depressed mood; unfavorable ones includedseclusiveness, ideas of reference, depersonalization,obsessionality, and histrionic behavior. Outcome wascorrectly predicted in 76 of the cases.

Carney, Roth, and Garside (1965) constructed theirNewcastle Scale predictive index by applying multipleregression analysis to determine the weights for each of 35clinical variables that best predicted the outcome 3 monthsafter ECT. Five favorable features (weight loss, pyknicphysique, early waking, somatic delusions, paranoiddelusions) and 5 unfavorable ones (anxiety, worsening ofmood in evening, self-pity, hypochondriasis, hystericaltraits) were selected and yielded a score that had a multiplecorrelation with outcome of 0.67. Each item that correlatedsignificantly with outcome also did so with a diagnosis ofendogenous depression. The Newcastle scale wassubsequently tested on a new sample of depressives byCarney and Sheffield (1972) and found to have a predictiveaccuracy of 76%; Katona et al. (1987)

used the scale to predict both immediate and 6-monthoutcome in response to ECT.

It is also instructive to compare the accuracy of thepredictive indices with the results that would have beenachieved by simply giving ECT to every depressive referredfor treatment (i.e., predicting a favorable outcome in everypatient). Such a procedure would have yielded a 62%predictive accuracy in the sample of Hobson (1953), and60% in the sample of Mendels (1967). Giving ECT to everypatient with prominent melancholic symptoms would haveincreased the predictive yield to 72% (Mendels, 1965b).

Inflation of the prognostic accuracy reported for thesevarious indices is inherent in the statistical procedures usedto derive them (Abrams, Volavka, and Fink, 1973; Hamilton,1982; Abrams, 1982a): Selecting those items significantlycorrelated with outcome and combining them in a coefficientof multiple correlation that â!œpredictsâ! ! outcome in thesample from which they were derived. Because the item setwas derived from the partic ular sample studied, the samevariables and weights are never as effective at predicting

P.46

outcome in other samples.

One research group consistently found little prognostic valuefor melancholic or endogenous features other than thepresence of psychosis. Coryell and Zimmerman (1984) testedthe predictive value of a DSM-III diagnosis of melancholia,along with a variety of other response predictors, in asample of primary unipolar depressives receiving ECT. Thepresence or absence of melancholia was the only variable of7 tested that actually bore an inverse relation to any of the3 outcome measures employed (Hamilton scale score atdischarge, global rating at discharge, and mean weeklyfollow-up symptom score): melancholies had slightly higherfollow-up symptom scores and poorer global ratings atdischarge. Nevertheless, the presence of delusions was themost favorable predictor variable, followed by increasing ageand female sex. The familial subtype of depressive spectrumdisease was unfa vorable, as was a longer duration of illnessprior to admission.

In a subsequent analysis of the same data set, Coryell,Pfohl, and Zimmerman (1985) also found secondarydepressives to have a worse outcome than primarydepressives. Depressed patients with a diagnosablepersonality disorder, however, did not fare any worse withECT (Zimmerman et al., 1986). Another study from the samegroup failed to find any predictive value of a Newcastle scalediagnosis of endogenous compared with reactive de pressionin predicting ECT respone (Zimmerman et al., 1986).

Although Rich et al. (1984) also failed to find any predictivevalue of RDC and DSM-III subtypes of major depressionâ!”reactive, secondary, nonpsychoticâ!”in 48 depressivesreceiving bilateral ECT, their sample con sisted mostly ofpatients with endogenous or melancholic depressives, leaving little variance for their predictors to explain.

Abrams (1982a) retrospectively analyzed the data from 97endogenous depressives who had participated in a series ofstudies of unilateral and bilateral ECT conducted over a 10-year period (Abrams and Taylor, 1973, 1974b, 1976b;Abrams et al., 1983) in which Hamilton depression scale

scores were obtained blindly before and after a course ofECT. Hierarchical multiple regression analysis was used toexamine the relation between each of the depression scaleitems at baseline and the post-ECT depression score, whichwas adjusted to account for the variances of thepretreatment depression score, treatment electrode

placement, sex, and the total number of ECTs received.After applying Bonferroni's correction for multiple statisticaltests, no significant correlations were obtained.

Similar results were obtained in subsequent prospectivestudies attempting to predict ECT response in endogenousdepression/melancholia. Andrade and associates (1988b)used four different indices (Hobson Index, Mendels Index,Newcastle Prognostic Index, and Newcastle Diagnostic Index)to predict ECT response in 29 endogenously depressedpatients and found that, although the Newcastle PrognosticIndex identified with high specificity (86.7) ECT respondersamong endogenous depressives, it inaccurately classified toomany patients as nonrespenders, thus gaining specificity atthe expense of sensitivity. The response rates of the otherindices ranged from 75% to 76.9%, and their predictiveaccuracies were likewise unimpressive. Similarly, if oneapplies Bonferroni's correction to the data of Pande et al.(1988), no significant predictors of ECT response emergefrom among 17 Hamilton depression scale items in a sampleof Research Diagnostic Criteria major depressives. Prudic etal. (1989) also found that, among a group of patients whomet research criteria for endogenous depression,endogenous and nonendogenous symptom clusters derivedfrom the Hamilton depression scale were equally responsiveto low-dose ECT.

Most recently, we attempted to predict the response of 47melancholic male patients to 6 unilateral or bilateral ECTs byusing individual rating items from the Hamilton depressionscale obtained at baseline, as well as 5 depression factorsderived therefrom (Abrams and Vedak, 1991). After applyingBonferroni's correction for multiple independent statisticaltests, none of the 15 individual Hamilton depression scaleitems or depression factors significantly predicted theadjusted post-treatment Hamilton depression score,confirming that excluding patients with nonmelancholicsyndromes from ECT samples attenuates the predictive valueof individual or grouped clinical psychopathologic features ofdepression, presumably by truncating the variance to beexplained, as in Abrams (1982a).

Diverse individual clinical and demographic items have beenreported to exhibit an association with ECT reponse. Afavorable effect of increasing age on outcome with ECT wasfound by several earlier investigators (Roberts, 1959b;Mendels, 1965a; Carney, Roth, and Garside, 1965), as was ashorter duration of illness (Hamilton and White, 1960;Hobson, 1953; Carney, Roth, and Garside, 1965; Herrington,

P.47

Bruce, and Johnstone, 1974). Although there are disparateresults (Hamilton, 1982), two studies in addition to that ofCoryell and Zimmerman (1984) also report better results inwomen than in men (Herrington, Bruce, and Johnstone,1974; Medical Research Council, 1965). Other variablespredicting a favorable response to ECT include cyclothymicpersonality (Ottosson, 1962a; Abrams and Taylor, 1974),

pyknic physique (Abrams and Taylor 1974), diminishedsalivary flow (Weckowicz et al., 1971), and psychomotorretardation (Hickie et al., 1990). The report of a better ECTreponse in bipolar than in unipolar depressives (Ferris andd'Elia, 1966) was not confirmed by Abrams and Taylor(1974a) or by Heshe, Roeder, and Theilgaard (1978).Although there are contradictory data, the presence ofprecipitating social stressors does not appear to predict theresponse of depressed patients to ECT (Zimmerman et al.,1987).

Delusional (Psychotic) DepressionThe presence of delusional or psychotic features has longbeen considered a useful clinical predictor of a favorableresponse to ECT. Indeed, the largest between-groupdifferences in the controlled, random-assignment prospectivestudies of genuine compared with sham ECT have occurredin the subgroup of patients with delusional depression(Clinical Research Centre, 1984; Brandon et al., 1984), andthe presence of delusions was the best predictor of outcomein the generally negative study of Coryell and Zimmerman(1984) described above.

The strong predictive value of the presence of psychosis wasrecently confirmed in a 4-hospital, prospective, random-assignment, double-blind, controlled collaborative study ofbitemporal ECT in major depression by the CORE studygroup (Petrides et al., in press). These investigatorscompared the treatment response of 177 nonpsychoticdepressives with that of 77 psychotic depressives. Theoverall treatment response for the entire sample of 253patients was 87%: 95% of the psychotic depressivessatisfied strict remission criteria, compared with 83% of thenonpsychotic depressives.

The fact that all patients received a clinically robust form ofECT (bitemporal placement with 1.5Ã— threshold dosing)may account for the sharply different results theseinvestigators obtained from those reported by Sobin et al.(1996), who found no predictive value of the presence of

P.48

psychosis in a sample of depressed patients who receivedeither unilateral or bilateral ECT at varying dosing levels. Itis not possible in their report to separately examine thepredictive effects of psychosis relative to response tobitemporal ECT alone, because the only comparisons madewere between what the authors at the time termed â!œineffectiveâ! ! and â!œeffectiveâ! ! forms of ECT, and thelatter goup included many patients whom we now knowactually received an ineffective treatment method: 2.5 Ã—threshold unilateral ECT. Patient population differences mayhave also contributed to the discordant study results: theCORE study drew its patients from 4 community hospitals,whereas the Sobin et al. (1996) study was conducted at astate hospital, where ECT response rates may be limited byan insufficient population of patients capable of the usualand expected response to ECT (Swartz, 2001a; Fink, 2001).

SchizophreniaTo my knowledge, Dodwell and Goldberg (1989) havereported the only prospective study of clinical factorsassociated with the response of schizophrenic patients toECT. In 17 patients who satisfied Research DiagnosticCriteria for schizophrenia or schizoaffective disorder, shorttotal and present illness durations, a paucity of premorbidschizoid or paranoid personality traits, and the presence ofperplexity, were all associated with a favorable treatmentresponseâ!”a finding entirely in keeping with the knownprog nostically favorable features in untreated schizophrenia.

Dexamethasone Suppression TestYears ago, the dexamethasone suppression test receivedspecial attention as a possible predictor of treatmentresponse in depressed patients receiving biologicaltreatments (Fink, 1986a). In an analysis of pooled data fromstudies that assessed the value of this test in predictingtreatment response of major depressives to adequate dosesof an antidepressant or to ECT, Arana, Baldessarini, andOrnsteen (1985) found that more than 70 of responders hada positive (nonsuppression) response compared with fewerthan 50 of nonrespenders, a modest but highly significantdifference. ECT response was not examined separately inthis analysis, and several investigators have now conductedprospective studies specifically relating suppressor status toECT response.

Coryell (1982) obtained dexamethasone suppression tests

P.49

within 1 week of admission in 42 DSM-III major depressiveswho subsequently received ECT for clinical reasons. Blindlyassigned global ratings of improvement based on a review ofnursing and progress notes for the final 3 days ofhospitalization were greater for nonsuppressors than forsuppressors, a difference not found for Hamilton DepressionScale scores. At follow-up 6 months later (Coryell andZimmerman, 1983), no between-group differences werefound on any measures. In a subsequent and differentsample of unipolar depressives receiving ECT, these authors(Coryell and Zimmerman, 1984) again found a significantcorrelation in the predicted direction between the suppressorstatus and the globally rated outcome, but not in theoutcome as determined by Hamilton Scale scores.

Subsequent studies generally failed to confirm these initialsanguine results (Ames et al., 1984; Katona and Aldridge,1984; Coppen et al., 1985; Lipman et al., 1986a, 1986b;Devanand et al., 1987; Katona et al., 1987; Fink, Gujavarty,and Greenberg, 1987), and one can only agree with Scott(1989) that â!œthere are no physiological measures or testswhich are superior to clinical criteria in the selection ofdepressed patients for whom ECT would be an effectivetreatment.â! ! The dexamethasone suppression test is littleused today for either prediction or monitoring of ECTtreatment response.

Some individual clinical features may, however, lackpredictive utility. The observation that a given depressedpatient is female, or older, or has a stocky build, or isagitated, is unlikely to influence the decision to give ECT.The fact is that favorable and unfavorable features do notexist independently of ill patients but tend to group togethernaturally into depressive syndromes, unfortunately oftenwith substantial overlap. Favorable features generallycharacterize patients with endogenous or melancholicsyndromes, and these are the patients who recover rapidlywith ECT.

The unfavorable features characterize patients with long-standing anxiety, hypochondriacal, and somatizationsymptomsâ!”often dating from childhoodâ!”and personalitytraits of hysteroid dysphoria, dependency, and inadequacy.Such patients typically respond briefly or not at all to ECTand may even be made worse by the treatment. In practice,it is not so difficult to identify patients who are likely tobenefit from ECT. For example, it is rare for apsychomotorically retarded patient with guilty delusions and

suicidal intent not to respond favorably to ECT. Thedifficulty usually arises in patients who would not generallybe considered prime candidates for ECT but whose atypicaldepressive features resist adequate treatment withantidepressant drugs. In such instances there is little help tobe derived from the predictors, and the administration of atrial course of ECT results in improvement with a frequencythat is sufficient to maintain the strategy in clinical practice.




> Table of Contents > Chapter 4 - The Medical Physiology of


Chapter 4

The Medical Physiology ofElectroconvulsive Therapy

Electroconvulsive therapy exerts its greatest physiologicalimpact on the brain and cariovascular system.

Cerebral Physiology and MetabolismThe combined physiological effects of the electrical stimulusfor ECT and the resultant generalized seizure discharge areimmediate and powerful and may be detectable days orweeks after the treatment course terminates.

Cerebral Electrographic EventsPrevious editions of this book have included extensivedescriptions of the effects of sine-wave ECT on the ictal andinterictal EEG. Because this method of ECT has essentiallybeen replaced by brief-pulse ECT, the sinewave results willonly be summarized here; the main focus will be on thebrief-pulse effects.

If an electrical stimulus depolarizes a sufficient number ofneurons, a generalized, paroxysmal, cerebral seizure ensues,the threshold for which is defined as the electrical dose (inmillicoulombs, mC) that produced it. Sub-convulsive stimulielicit only an electroencephalographic (EEG) â!œarousalâ! !response of low-voltage fast activity that is indistinguishablein appearance from that seen in the earliest phases of ECT-induced seizures (Penfield and Jasper, 1954; Chatrian andPetersen, 1960; Staton, Enderle, and Gerst, 1981) and thatWeiner (1982) has dubbed the â!œepileptic recruitingâ! !stage. With substantially suprathreshold stimuli, this initiallow-voltage, 18-to 22-Hz activity is rapidly replaced by acrescendo of high-voltage 10-to 20-Hz hyper-synchronouspolyspikes occurring simultaneously throughout the brain

P.51

and corresponding to the tonic phase of the motor seizure.This discharge gradually decreases in frequency as theseizure progresses, evolving into the characteristic polyspikeand slow-wave complexes of the clonic motor phase, whichslow to 1 to 3 Hz just before seizure termination, and areoften abruptly replaced by EEG flattening (â!œpostictalsuppressionâ! !).

The degree of postictal suppression immediately followingtermination of the clonic phase of the seizure varies.Although it occurs abruptly in about one-third of the casesand more often after bitemporal than after unilateral ECT(Abrams, Volavka, and Fink, 1973; Daniel et al., 1985),many records show a less precise end point, with polyspikeand slow-wave activity apparently stopping for a second ortwo and then resuming (Daniel et al., 1985; Swartz andAbrams, 1986) or gradually and imperceptibly blending intoa mixture of alpha and beta activity (Small et al., 1970;Abrams, Volavka, Fink, 1973). In about 10% of instances,however, the seizure end point is indeterminate (Larson,Swartz, and Abrams, 1984). When it is clearly observed, thephase of postictal suppression lasts up to about 90 seconds,when high-voltage, irregular delta waves of 1 to 3 Hzgradually appear, followed by increasingly rhythmic thetawaves that progressively merge into the pre-seizure rhythmsby about 20 to 30 minutes after seizure termination. Asdiscussed in Chapter 6, the degree of EEG postictalsuppression reflects both the extent of intracerebralgeneralization and the therapeutic quality or im pact of theECT-induced seizure.

At the level of the scalp -recorded EEG, the seizure appearsas an all-or-none phenomenon. The titration proceduresdeveloped by Sackeim et al. (1987a) show that it only takesa small increment in electrical dosage to convert an entirelysubconvulsive stimulus into one that yields a generalizedseizure. Moreover, substantial increases in stimulus intensityabove threshold do not further lengthen the seizure (e.g.,Sackeim et al., 2000), and often shorten it. In the brain,however, seizure threshold, frequency, amplitude, andduration varies according to the structures involved.Chatrian and Petersen's (1960) depth electrode recordings inpatients receiving inhalant-or chemically induced seizuresalso clearly demonstrate the polyspike phase starting atvariable postinduction intervals: â!œA high-voltage,rhythmic discharge already may be well developed in onearea of the brain, while in other areas a similar discharge is

P.52

only at the outset or the recording is still apparently flat.â! !The induced seizure activity described often terminatedseparately in different parts of the brain.

Seizure duration is variable in relation to stimulus wave-form and treatment electrode placement. Using a sine-wavestimulus, Abrams, Volavka, and Fink (1973) foundsignificantly shorter seizures with unilateral than withbitemporal ECT, but the same method in a youngerpopulation (Abrams et al., 1983) revealed slightly longerseizures with unilateral ECT. Moreover, when brief-pulse ECTwas studied in the latter population, seizure length wasequal for both methods (Swartz and Abrams, 1984).

In a comparison that systematically varied both electrodeplacement and stimulus wave-form, Weiner (1980) reportedsignificantly longer seizures with bitemporal than withunilateral ECT, a difference seen most clearly with sine-wavestimuli. Most brief-pulse ECT studies (Weiner et al., 1986a;Sackeim et al., 1987a, 1993 , 2000) find no differences inseizure length for the two methods when they areadministered with just-or moderately above-thresholdstimulation, although contradictory data exist (Delva et al.,2001).

Seizure monitoring during ECT has also provided theinteresting observation that paroxysmal EEG seizure activityvariably continues for about 10 to 15 seconds after allvisible motor activity ends, a discrepancy that may reflectincomplete intracerebral seizure generalization.

Because seizures induce physiologic effects related to theactivation of different regions of the brain, the duration ofthese effects should correlate highly for well-generalizedseizures. Swartz and Larson (1986) calculated thecorrelation coefficients between pairs of four differentmeasures of seizure duration (motor activity, EEG spikeactivity, total paroxysmal EEG activity, and tachycardiaduration) and found them to be significantly higher withbilateral compared with unilateral ECT, suggesting greatergeneralization throughout the brain with the formertechnique. Using the same method, Larson and Swartz(1986) also demonstrated greater intracerebralgeneralization with the first than with the second of 2seizures given consecutively in a single treatment session.Because less well-generalized seizures may be expected tospread more slowly and therefore last longer (indeed, theduration of total paroxysmal EEG activity was significantly

P.53

longer for the second ECT), the authors' finding may helpexplain the increased occurrence of prolonged seizures withmultiple-monitored ECT, and the shorter seizure durationsobtained with higher-dose stimuli.

With brief-pulse stimulation intraseizure EEG patterns areusually symmetrical across the hemispheres for bitemporalECT but not for right unilateral ECT, which is associatedwith a more intense right hemisphere re sponse that may befrontally accentuated (Weiner, Coffey, and Krystal, 1991;Krystal et al., 1992, 1993; Luber, Nobler, and Moeller,2000).

Following a single ECT, very little EEG change persists afterthe seizure patterns have terminated and been graduallyreplaced by the pretreatment rhythms. As the numbers oftreatments increase, however, the EEG slowing persists intothe postconvulsive period, accumulating as a function of thetotal number of ECTs and their rate of administration(Volavka et al., 1972; Fink, 1979). This EEG activityincreases in amplitude and duration and decreases infrequency with each additional treatment as long as the rateof administration remains above 1 per week. These changesare accompanied by a decreased mean frequency and totalbeta activity and an increased mean EEG amplitude, totalpower, and total paroxysmal activity (Fink, 1979; Kolbeinsonand Petursson, 1988).

With the usual three treatments per week, the EEG obtained24 to 48 hours after 6 to 8 seizures given with sine-wavebitemporal ECT is often dominated by theta/delta activity(Figure 4-1) with a marked reduction in the abundance ofnormal alpha/beta rhythms. This postconvulsive (interictal)EEG slowing is also related to the pretreatment EEG, age,and method of seizure induction (Volavka et al., 1972).

Following the final treatment of a course of ECT, thecumulative EEG slowing typically diminishes gradually overtime and eventually disappears (Weiner, 1980). Most sine-wave studies show a return to baseline by 30 days post-ECT(Moriarity and Siemens, 1947; Roth, 1951; Bergman et al.,1952).

Total EEG power reached its peak 1 week following a courseof partial sine-wave ECT (intermediate in intensity betweenfull sine-wave and brief-pulse ECT), and returned to baselinewithin 4 weeks (Kolbeinsson and Petursson, 1988). A recentstudy of brief-pulse ECT (Sackeim et al., 1996) found noevidence of EEG slowing 8 weeks after unilateral or

bitemporal ECT.

Figure 4-1 EEG delta activity 24 hours after ECT.

With sine-wave stimulation, bitemporal ECT inducesinterictal EEG slowing that is either symmetric oraccentuated over the left hemisphere, whereas the interictalEEG slowing of unilateral ECT is typically accentuated overthe stimulated hemisphere (Abrams et al., 1970; Volavka etal., 1972; Abrams, Taylor, and Volavka, 1987). The resultswith brief-pulse ECT are more variable. Abrams, Volavka,and Schrift (1992) found no tendency for any treatmentelectrode placement, bitemporal or left or right unilateral, toinduce lateralized interictal EEG patterns. Sackeim et al.(1996), however, reported similar lateralization effects tothose for sine-wave ECT: left -hemisphere accentuation withbitemporal placement, and right-hemisphere accentuationwith right unilateral placement.

Interictal EEG Change and TreatmentResponseFink and Kahn (1957) used objective methods to measurethe ECT-induced delta activity in a large sample of patientsand related this activity to clinical response as evaluatedduring at least 8 weeks of follow-up. They found thegreatest clinical improvement in patients with the most EEGslowing and also reported that the early appearance of suchslowing was related to as sessments of global improvement.

In a subsequent study of sine-wave ECT, Volavka et al.

P.54

(1972), computer -derived frequency and power spectralanalyses were used to measure EEG delta activity after sine-wave unilateral and bitemporal ECT. No relation was foundbetween these measures and clinical improvement in

depression as assessed by reduction in depression scalescores. The authors suggested that global EEG slowing andtherapeutic response were both a function of the number ofECTs given, but were not themselves causally related.

Sackeim et al. (1996) confirmed the original suggestion ofFink and Kahn (1957) that increased frontal delta activity isrelated to improvement with ECT. In a sample of 62inpatients with major depressive disorder who wererandomized to receive brief-pulse right unilateral orbitemporal ECT given at high or low dosages, effective formsof ECT (i.e., both forms of bitemporal ECT, plus high-doseunilateral ECT) increased interictal delta power in prefrontalbrain regions, an increase that correlated significantly withthe degree of clinical improvement. Confirming earlierstudies (Abrams et al., 1970; Volavka et al., 1972; Abrams,Taylor, and Volavka, 1987) Sackeim et al. (1996) founddifferential lateralization of ECT-induced delta power forbitemporal and right unilateral ECT, with an accentuationover the right hemisphere with right unilateral ECT, and theleft hemisphere with bitem poral ECT.

The most recent study is that of Heikman et al. (2001), whoused magnetoencephalography (MEG) to examine therelation between ECT-induced frontal slowing in the 0.5 Hz-7Hz bandwidth, and treatment response to unilateral orbifrontal ECT, administered at 5 Ã— the titrated seizurethreshold. In 7 patients with major depression who hadnever before received ECT, although ECT-induced increasesin frontal delta and theta activity reached variable levels ofsignificance according to the hemisphere studied, only theincrease in left frontal MEG activity in the theta bandwidthcorrelated with reduction in depression scale score after the4th treatment, but not after 8â!” 9 ECTs. Although thesedata are consistent with the EEG data reviewed above, theymust be interpreted with caution as no statistical correctionwas made for the multiple independent f-tests andcorrelational measures employed.

The Sleep ElectroencephalogramAbnormalities in the all-night sleep EEG have been describedin depressed patients, the most consistently reported of

P.55

which is a reduction in the speed of onset (latency) of rapid-eye-movement (REM) sleep, which also correlates with theseverity of the depression (Kupfer, 1986). Although ECT isgenerally reported to increase mean REM latency in samplesof drug-free depressives (Hoffman et al., 1985; Linkowski etal., 1987; Coffey et al., 1988b), the data have beencharacterized by marked individual variability, perhaps as aresult of the small samples studied and their diagnosticheterogeneity (Coffey et al., 1988a). Even more problematicis the relation between ECT-induced REM latency increaseand relief from depression, because several instances of fullrecovery without restoration of REM latency to the normalrange, and vice versa, have been reported. The report ofGrunhaus et al. (1997)

that the presence of sleep-onset REM was associated with apoor response to ECT is methodologically flawed andtherefore unhelpful. Characterization of the specific effectsof ECT on EEG sleep stages and the resultant effects, if any,on clinical symptoms, awaits the prospective, double-blindstudy of larger samples of drug-free melancholic patients.

Cerebral Metabolism: Blood Flow,Oxygen Consumption, and GlucoseUptakeThe induced cerebral electrical discharges during ECT andthe subsequent changes in electrical rhythms areaccompanied by changes in cerebral blood flow, oxygenconsumption, and glucose utilization. Because these 3aspects of cerebral function are highly correlated (increasedfunctional brain activity is associated with increases ofsimilar magnitude in all 3 variables), they are often usedinterchangably when discussing the effects of ECT oncerebral metabolism.

Early Studies

Blood FlowBrodersen et al. (1973) reported that cerebral blood flowdoubled during sine-wave bitemporal ECT, a result that wasconfirmed by Lovett-Doust and Raschka (1975) usingimpedance plethysmography. The increased cerebral bloodflow reported during the induced seizure contrasts withreports of its reduction postictally (Kety et al., 1948; Wilsonet al., 1952).

P.56

Oxygen ConsumptionPosner, Plum, and Van Poznak (1969) found the femoralarterial pO2 always remained above 100 mm Hg during sine-

wave bitemporal ECT, and that jugular venous pO2 remained

steady around 60 mm Hg throughout the seizure, neverapproaching the 20 mm Hg level indicating cerebral hypoxia.Brodersen et al. (1973) found that, like blood flow, oxygenconsumption doubled during seizures with sine-wavebitemporal ECT, with modest increases in both cerebralvenous pO2 and pCO2. Szirmai, Boldizsar, and Fischer

(1975) found no change in cerebral venous oxygensaturation during any phase of the EEG seizure; saturationremained above 90% at all times.

Glucose UtilizationJust as for blood flow and oxygen consumption, Brodersen etal. (1973) found a doubling of glucose utilization duringseizures with sine-wave bi temporal ECT.

Modern StudiesInvestigators using a variety of radioisotope uptake imagingtechniques have examined the effects of ECT on cerebralblood flow and glucose uptake.

Blood FlowBolwig, Hertz, and Holm-Jensen (1977) used a xenon 133(Xe) inhalation technique to demonstrate increased cerebralblood flow during partial sine-wave bitemporal ECT andpostulated that a simultaneously observed increasedpermeability of the blood-brain barrier was secondary tocapillary hyperperfusion consequent to the increased flow.Silverskiold et al. (1986) studied depressed patients with theXe inhalation technique before and after partial sine-waveunilateral or bitemporal ECT. They found cerebral blood flowsignificantly reduced 1-2 hours after ECT with bothtreatment methods, more with bitemporal than withunilateral ECT. Moreover, suppression of flow was mostpronounced in the stimulated (right) hemisphere withunilateral ECT. All differences were greater earlier than laterin the treatment course. These results were confirmed byProhovnik et al. (1986), who examined depressed patients25 minutes before and 50 minutes after inductions with low-dose, brief-pulse, unilateral or bitemporal ECT. Bitemporal

P.57

ECT resulted in symmetric frontal flow reductions of about15%, and right unilateral ECT resulted in similar reductionsover the stimulated hemisphere but only about 5%reductions over the nonstimulated hemisphere. Thesereductions in blood flow were greatly attenuated after thefirst ECT in a series.

Rosenberg et al. (1988) obtained similar results usingsingle-photon emission computer -tomography (SPECT) withXe inhalation, reporting that cerebral blood flow dropped 8%following the third brief-pulse ECT and a further 13% afterthe final treatment.

Bajc et al. (1989) used Tc99m-hexamethylpropyleneamineoxime (HMPAO) SPECT to study11 patients before and during bitemporal ECT, beginningwith stimulus administration. Relative isotope uptake, andtherefore brain perfusion, increased significantly during theseizure over the right, but not over the left, frontal andfrontotemporal areas, increases that were again muchsmaller in magnitude than those reported by Broderson et al.(1973).

In an extension and amplification of the earlier study ofProhovnik et al. (1986), Nobler et al. (1994) examined theeffects of ECT on rCBF in both depressed and manic

patients, using the Xe inhalation technique, and for the firsttime in such a study, specifically assessed the relation ofthe blood flow changes to clinical improvement. Corticalblood flow was measured 30 minutes before and 50 minutesafter a single ECT for both depressed and manic patients,and again during the week following ECT for depressedpatients. At baseline, both patient groups exhibited deficitsin

global and topographic blood flow (including hypofrontality)that were consistent with reports of decreased metabolicfunction in the frontal lobes of untreated depressives (e.g.,Cummings, 1993). ECT accentuated the abnormality byinducing further global and regional blood flow reductions,which reductions were, quite unexpectedly, positivelyassociated with clinical improvement: prefrontal reductions,in particular, were strongly associated with a positiveoutcome. In my view, no amount of post-hoc rationalizationcan satisfactorily account for the paradoxical finding thataggravating the base line hypofrontality of depressivesrelieves their symptoms.

However, subsequent studies have failed to confirm the

P.58

report of Nobler et al. (1994) either that ECT reduces frontalblood flow, or that such a reduction correlates positivelywith clinical treatment response. Bonne et al. (1996) usedHMPAO SPECT to study 20 patients with major depressionseveral days before, and about a week after, a course ofbrief-pulse, bitemporal ECT. Cerebral blood flow increased

significantly in ECT responders and was unchanged innonresponders, using the same definition of treatmentresponse as Nobler et al. (1994). Bonne et al. (1996) alsoobserved an in verse correlation between baseline severityof depression and isotope uptake, and a direct correlationbetween clinical improvement and uptake.

Galynker et al. (1997) used HMPAO SPECT to study adepressed patient with catatonic features before and after acourse of ECT. Cerebral perfusion, which was markedlyreduced at baseline compared with normal control subjects,was markedly increased by a course of 5 ECTs.

Milo et al. (2001) used HMPAO SPECT and obtained similarresults to Bonne et al. (1996): frontal hypoperfusion in 15depressives before brief-pulse ECT (electrode placementunspecified), and increased cerebral perfusion following ECTin those patients who enjoyed the best clinical response.

Most discouraging is the report of Pridmore et al. (2001)who obtained HPAO SPECT scans at baseline andimmediately following high-dose brief pulse unilateral,bitemporal, or bifrontal ECT in 8 patients, and were unableto detect any pre-post differences at all.

To summarize: cerebral blood flow is generally reducedfrontally in depressed patients at pretreatment baselineexamination, increased over baseline during ECT-inducedseizures, and then either increased or reduced belowbaseline following ECT, possibly depending on the time ofreexamination and the method chosen for estimating bloodflow. Blood flow reductions following ECT have been reportedfor Xe-inhalation studies performed during the hours aftertreatment (Silverskiold et al., 1986; Prohovnik et al., 1986;Rosenberg et al., 1988; Nobler et al., 1994), whereas bloodflow increases following ECT have been reported for HMPAOSPECT studies performed at longer intervalsâ!”often daysâ!”following treatment (Bonne et al., 1996; Galynker et al.,1997; Milo et al., 2001). With this division, it is not possibleto tell whether the re-examination interval or the techniqueis responsible for the observed differences.

Glucose MetabolismPositron emission tomography (PET) employing various formsof labeled fluorodeoxyglucose (FDG) provides a measure oflocal glucose uptake, and therefore, cerebral metabolismduring seizures. Unfortunately, the present level oftechnology of this procedure, which only measures utilizationover relatively prolonged periods of time, does not typicallyallow discrimination of ictal from postictal phases. AlthoughEngle, Duhl, and Phelps (1982) were initially able todemonstrate increased glucose utilization over baselineduring ECT-induced seizures and a sharp drop in utilizationbelow baseline during postictal suppression, they failed toreplicate these findings in a later study (Ackermann, Engle,and Baxter, 1986). Neither Volkow et al. (1988) or Guze etal. (1991) were able to detect a significant reduction inbifrontal cortical glucose uptake on PET scans obtained indepressed patients 24 hours after bitemporal ECT.

Yatham, Clark, and Zis (2000) obtained FDG PET scans in 6depressed patients before a course of brief-pulse unilateralor bitemporal ECT and one week later. No significant pre-topost-ECT change in glucose uptake was detected, and therewas no correlation between changes in regional glucoseutilization and treatment response.

The most recent study (Henry et al., 2001) obtained FDGPET scans on 6 depressed patients before and after a courseof brief-pulse, bitemporal ECT. The authors performed 61paired Mests comparing glucose uptake in various brainregions before and after ECT, and found 17 significant pre-post differences; of these, 3 (right and left frontal, andright parietal) correlated with clinical treatment response.Because no correction (e.g., Bonferroni) for multiple f-testswas performed, the authors' assertion of a significantcorrelation between reduced glucose uptake in the frontallobes and treatment response is doubtful.

In summary, FDG PET studies during and after ECT revealneither significant regional changes, nor correlations withtreatment response.

An important caveat that applies to all of the above studiesis that although frontal lobe metabolic changes are generallystressed in the discussion and abstract sections of thearticles, in actuality multiple topographic areas of alteredmetabolism are typically found. Because the results relatingto the frontal lobes chance to fit a particular hypothesis ofdepressive illness and ECT response, these concordantresults are selectively focused on, thereby creating the

P.59

illusion of a highly consistent phenomenon out of resultsthat are, in fact, quite variable.

For example, the article by Nobler et al. (1994) stresses thefrontal and anterior temporal correlations found between

rCBF reductions and clinical treatment response, while

ignoring a significant parietal correlation that was alsodetected. Henry et al. (2001) do likewise, bypassing asignificant correlation between parietal glucose uptake andclinical improvement in favor of joining Nobler et al. (1994)on the frontal bandwagon. Because science provides no apriori reason to believe that parietal lobe functions cannotbe

central to the antidepressant effects of ECT, social, ratherthan scientific, considerations appear to play thedetermining role in this process.

Blood-Brain BarrierCerebral permeability (the blood-brain barrier) might beaffected by ECT. Bolwig and his associates (Bolwig, 1984 ,1988), used a double-isotope technique to study bothtranscapillary escape and capillary diffusion of small tracers(e.g., urea) into the brain during ECT. No increase intranscapillary escapeâ!”and therefore no breakdown of theblood-brain barrierâ!”occurred, although a net increase inpassive diffusion of small molecules across capillaryendothelial cells was observed, which the authorsinterpreted as re sulting from either increased availablecapillary area secondary to hyper-perfusion, or recruitmentof underperfused capillaries.

Mander et al. (1987) used MRI to study cerebral and brain-stem changes after bitemporal or unilateral ECT in 14patients. Scans were obtained before and after the first anda later treatment in the course in 11 patients and beforeand after the last treatment, followed by 6 additional scansat hourly intervals in 3 patients. Immediate post-treatmentscans were obtained as soon as possible after the patientemerged from anesthesia. Tl (spin lattice relaxation) times,reflecting tissue water content, rose immediately after theseizure, peaked at 4 to 6 hours, then returned to baseline;no long-term increases occurred across the treatmentcourse. The rise in Tl time was 83% greater after bitemporalcompared with unilateral ECT, but not significantly so. Theseauthors interpreted their data to support the hypothesis ofBolwig and associates (Bolwig, 1984 , 1988) that ECT

P.60

induces a temporary func tional increase in cerebrovascularpermeability.

Scott et al. (1990) generally confirmed these MRI findings in20 unipolar depressives studied immediately before and 25minutes after right unilateral ECT, with follow-up scansobtained 24 hours later in 13 subjects. The maximal increasein Tl relaxation times occurred between 25 minutes and 2hours post-ECT; the largest fall was observed from 2 to 6hours post-ECT; and all values returned to baseline by 24hours post-ECT.

Zachrisson et al. (2000) examined the ratio of the albuminconcentration in the CSF to that in the serum in a sample of9 depressed patients before and after a course of ECT, as ameasure of blood-brain barrier dysfunction. Although severalpatients had signs of blood-brain barrier dysfunction atbaseline, no change in the CSF/serum albumin ratioâ!”andtherefore no blood-brain barrier dysfunctionâ!”was inducedby ECT.

To summarize, all studies confirm maintenance of thestructural integ rity of the blood-brain barrier during ECT.

Do Persistent Brain Changes Occur?In view of the large and diverse effects of ECT on cerebralphysiology and metabolism described previously, is thereconvincing evidence that ECT is

capable of permanently damaging the brain? Although thereare no data suggesting the possibility of such an occurrence,it remains a question of concern among many patients whoreceive ECT (see Chapter 12) as well as among twooutspoken medical opponents of this form of therapy(Friedberg, 1977; Breggin, 1979). For obvious reasons, thequestion must be considered in light of the present-daypractice of ECT, using barbiturate anesthesia, musclerelaxation, and oxygenation, sharply truncating the availablebody of data that addresses the topic, much of it obtainedbefore modern treatment techniques became standard. MightECT given in the era before these advances were introducedhave caused brain damage in certain patients under certaincircumstances? Conceivably, because patients often becamecyanotic during treatment, generally received substantiallylonger courses of treatment than are administered today,and may possibly, although rarely, have developed tardiveseizures long after the treatment course was terminated(Fink, 1979). Even in the absence of oxygenation and

P.61

muscular relaxation, however, studies of electrically inducedseizures in animals demonstrate that cerebral lesions do notoccur unless seizures are prolonged for many mul tiples ofthe duration of those encountered during the administrationof ECT (Weiner, 1981).

Electroconvulsive therapy could only cause brain damagethrough the electrical stimulation or the induced seizure.Because an electrical stimulus can only damage brain tissueby burning, the calculations of Swartz (1989a), which showthat even under a worst-case scenario, the maximum outputof modern brief-pulse ECT devices is incapable of elevatingbrain tissue tem perature by even one tenth of a degreecentigrade, effectively eliminate the possibility of electricaldamage.

In reviewing his own and other studies of theneuropathologic consequences of induced seizures (primarilyin baboons), Meldrum (1986) pointed out that selective braindamage involving neuronal loss and gliosis in thehippocampus (the brain region most susceptible to anoxia)requires sustained generalized seizures lasting more than 90minutes, or more than 26 recurrent seizures in an 8-hourinterval, or continuous limbic seizures lasting longer than 3to 5 hours. These figures are for unmodified seizures; whencurarization and oxygenation are employed during theprocedure, continuous seizures for 3 to 7 hours are requiredto produce permanent damage. Although he raises thepossibility that a prolonged period of limbic statusepilepticus might conceivably be triggered in a susceptiblepatient, Meldrum also ac knowledges that the potentanticonvulsant effects of ECT render such an occurrenceextremely unlikely.

O'Connell et al. (1988) used mercaptopropionic acid toproduce status epilepticus in paralyzed, ventilated rats underEEG monitoring and found that, compared with controls,lesions in the substantia nigra occurred after as little as 10minutes of continuous seizure activity. Although this paperis the first to demonstrate neuronal damage in animalssubjected to continuous seizures of less than 30 to 60minutes' duration, its use of a neurotoxic chemical to induceseizures makes its relevance to ECT questionable.

Virtually all of the animal studies literature reviewed byWeiner (1984) had to be rejected because of the excessiveelectrical doses used, the fact that the seizures studied wereunmodified by muscle-paralysis and oxygenation, the lack of

unshocked control animals for comparison, and mostimportantly the inappropriate tissue fixation methods used.Two methodological features particularly characterize theanimal studies of the 1940s that purported to find thatelectrically induced seizures could cause neuropathologicchanges (Alpers and Hughes, 1942; Heilbrunn and Weil,1942; Neuberger et al., 1942; Lidbeck, 1944; Winkelmannand Moore, 1944; Ferraro, Golden, and Hare, 1946): Allused the technique of immersion fixation in formalin topreserve the brains for study, and none used unshockedcontrol animals.

Because postmortem cellular degeneration begins within 30minutes after death, studies in which the brain is removedand fixed by immersion in formalin require a control groupof untreated animals examined after the same postmorteminterval to ensure that any neuronal changes observed inthe experimental animals are the result of ECS rather thannonspecific postmortem degeneration. Moreover, crucial dataon the lapse of time between death and fixation are lackingin each of the uncontrolled immersion-fixation investigationsof ECS-induced neuropathology. The fact is, that inimmersion fixation the formalin penetrates the tissues soslowly that autolytic changes take place before the tissueelements are adequately fixed. The definitive method foravoiding this problem is in vivo formalin perfusion fixation, atechnique that rapidly and simultaneously fixes the entirebrain in the living, anesthetized, animal: The formalin isinjected arterially, under pressure, in combination with otheragents to prevent tissue shrinkage (Windle, Kreig, andArieff, 1945; Cammermeyer, 1972). None of three studiesusing the in vivo perfusion fixation method found anyevidence for ECS-induced brain damage in animals(Fetterman, 1942; Windle, Krieg, and Arieff, 1945; Siekert,Williams, and Windle, 1950). It is further noteworthy thatthe same perfusion fixation technique was sensitive enoughto detect the subtle neuronal damage in animals subjectedto cerebral concussion (Windle, Groat, and Fox, 1944),asphyxiation (Jensen, Becker, and Windle, 1948), andinanition (Liu, 1949).

Unshocked animals, prepared and handled identically toexperimental animals after the same postmortem interval,provide a critical control for the problems encountered withimmersion fixation. Thus, Globus et al. (1943) were able tocorrectly characterize as â!œpseudo-defectsâ! ! the minimalchanges they found after electrically induced seizures indogs because such changes were as common in the control

P.62

as in the experimental animal. Likewise, Dam and Dam(1986) performed a meticulous cell-counting study in agroup of rats given 3 electrically induced seizures a day upto 140 seizures and found no difference in hippocampal andPurkinje cell densities compared with unshocked control rats.

Friedberg (1977) and Breggin (1979) have made much of thepetechial hemorrhages noted in the brains of animals givenECS in some uncontrolled studies (Alpers and Hughes, 1942;Heilbrunn and Weil, 1942; Neurberger et al., 1942; Lidbeck,1944).

Because these authors found no neuronal changes outside ofthe immediate petechial areas, which would have indicatedelectrically or seizure-induced damage, it is likely that thesehemorrhages resulted from craniocerebral trauma caused bythe unrestrained muscular movements in the nonparalyzedanimals studied (Windle, Krieg, and Arieff, 1945; Siekert,Williams, and Windle, 1950). When Siekert, Williams, andWindle (1950), during administration of ECS restrained theirmonkeys in a special chair in which the monkeys' headsprojected through an adjustable aperture, they found nosuch hemorrhages. Moreover, when studies includeunshocked control animals for comparison (e.g., Hartelius,1952), hemorrhages appear in the control animals as well.

Hartelius (1952) was the sole investigator to use carefulmethods, albeit immersion fixation, and a control group toreport any ECS-induced neuronal changes. He gave catseither 4 ECSs at 2-hour intervals or 11 to 16 ECSs,administered 4 per day for 3 to 4 consecutive days. Hefound what he characterized as â!œfairly subtleâ! ! neuronalchanges in the brains of the shocked cats compared withthose of unshocked controls, especially a greater variabilityin the affinity of cells for the stain and a tendency for thecells to show more darkly stained nuclei. The significance ofthese changes is doubtful, because, as Hartelius was wellaware, darkly staining cells are widely considered to bepostmortem artifacts (Cammermeyer, 1961, 1972). In fact,Hartelius warns that â!œthe risk of faulty and biasedevaluation is still greater in the case of the nerve cellchanges in question, since they are less distinct.â! ! Dam andDam (1986) agree, characterizing Hartelius' (1952) find ingsas slight and mostly reversible.

KindlingAlthough it has been suggested hypothetically that ECT

P.63

might kindle (create by repeated low-dose intracerebralelectrical stimulation) epileptic foci (Pinel and Van Oot,1975, 1977), several studies show that electroconvulsiveshock (ECS) in animals actually does the reverse of kindling:it exerts a pronounced anticonvulsant effect on amygdala-kindled seizures. Babington and Wedeking (1975) showedthat one electrically induced seizure administered from 15minutes to 2 hours before an amygdala-kindled seizure inanimals markedly reduced seizure duration, an effectconfirmed by Handforth (1982). Post et al. (1984)demonstrated a marked protective effect against kindlinglasting up to 5 days after a 7-day course of once-daily ECS.Moreover, when ECS was administered before the stimulationgiven to induce kindling, this phenomenon was completelyblocked.

Several recent studies show that ECS induces neurogenesisin rats (Vaidya et al, 1999; Vaidya, Terwilliger, and Duman,2000; Gombos et al., 1999; Scott, Wojtowicz, and Burnham,2000), a phenomemon also seen with kindling. However,kindling is associated with neuronal cellular loss in the

hippocampal hilar region, a phenomenon that does not occurafter ECS in animals (Dam and Dam, 1986; Dalby et al.,1996; Duman and Vaidya, 1998) or ECT in man (Ende et al.,2000). Repeated ECS is associated with significant mossyfiber sproutingâ!”but not significant cell lossâ!”inhippocampal granule neurons of the dentate gyrus of treatedrats compared with untreated controls (Vaidya, Terwilliger,and Duman, 2000; Gombos et al., 1999). The same ECSschedule is also associated with an increase in the numberof bromodeoxyuridine-containing cells in the dentate gyrusof the hippocampus, reflecting increased new cell formation(Scott, Wojtowicz, and Burnham, 2000). Since ECS does notcause kindling, and in fact exhibits potent antikindlingproperties, it is possible that these changes may relate tothe therapeutic action of ECT, and perhaps even to aneuroprotective effect as well.

Brain Studies in ManAutopsy material from patients dying during or just afterECT is rare to begin with, considering the low mortalityrates (see Chapter 5), and almost impossible to interpretbecause of the lack of any information on the structure ofthese patients' brains before ECT. There is, in addition, agreat likelihood of extensive coexisting agonal changesbecause most ECT deaths are cardiovascular in origin and

P.64

therefore likely to produce severe anoxic changes. Moreover,depressive illness itself is associated with MRI evidence ofhippocampal atrophy (Sheline et al., 1999): patients with ahistory of depression exhibit smaller hippocampal volumesbilaterally than controls, a result that is unrelated to age.

The recent and widespread availability of brain imagingtechniques makes it possible to study directly the effects ofECT on ventricular size and cortical structure. Earlyretrospective studies (Weinberger et al., 1979; Galloway etal., 1981; Owens et al., 1985; Kolbeinsson et al., 1986;Bergsholm et al., 1989) that attempted to relate lateralventricular size on CT scan to a history of ECT in differentpatient samples are uninformative because this approachsuffers from serious methodological problems, in cludingespecially selection bias (Schwartz, 1985).

For this reason, this section will consider only prospectivebrain-imaging studies of patients undergoing ECT. Bothmagnetic resonance imaging (MRI) studies of Mander et al.(1987) and Scott et al. (1990), already cited above in thesection on the blood-brain barrier, showed a return offunctional brain changes to baseline by 24 hours post-ECT,providing no support for the existence of long-term orpersistent damage.

Pande et al. (1990) conducted a prospective, blind MRI studyof seven major depressives who received right unilateralECT. MRIs obtained within 1 week after the course of ECTrevealed no differences from baseline; as in the study byCoffey et al. (1991) cited below, about two thirds of patients

showed multiple areas of signal hyperintensity in theperiventricular white matter at baseline, which remainedunchanged after ECT. Such changes, however, areassociated with ECT-resistance and an increased risk of cognitive side-effects and delirium (Hickie et al., 1995;Vidbech, 1997; Simpson et al., 1998).

In a prospective single case study, Scott and Turnbull(1990) followed a 77-year-old hypothyroid woman with MRIsobtained at baseline and seriatim across several courses ofECT. Despite the emergence of significant cognitiveimpairment following the last course, no evidence for braindamage was detected with MRI. Adding 4 more consecutiveprospectively studied cases using quantitative MRI, Scott etal. (1991) found that 2 or 3 courses of ECT showed noconsistent effect on brain structure, a result that is quiteconsistent with the report of Coffey et al. (1991) in patients

P.65

receiving single courses of ECT. As in the latter study, Scottet al. (1991) recorded the development of a 2-mmsubcortical focus of increased signal intensity be tween thefirst and second MRI in one patient, an occurrence that isquite frequent in healthy elderly subjects (Scott, 1995).

Coffey et al. (1991), provide the definitive prospective studyof the brain anatomic effects of ECT. Blindly analyzed serialMRIs obtained in 35 patients at baseline, 2 to 3 days post-ECT, and 6 months later, revealed no acute or delayedchanges in total volumes of the third or lateral ventricles,frontal or temporal lobes, or the amygdala-hippocampalcomplex. Pairwise global comparisons in 5 subjects revealedan apparent increase in preexisting subcortical white matterhyperintensity, which the authors attributed to pro gressionof ongoing cerebrovascular disease, unrelated to ECT.

Ende et al. (2000) obtained proton magnetic resonancespectroscopic images of the hippocampus in 14 depressedpatients before and after courses of right unilateral ECTadministered with age-based dosing. ECT caused no changesin hippocampal N-acetylaspartate signals, thus providing noevi dence that ECT can cause hippocampal atrophy or celldeath.

Although little definitive scientific information can begleaned from individual cases, it is nevertheless reassuringto note that a 63-year-old woman described by Kendell andPratt (1983) received 325 ECTs over 4 years without anyevidence of brain atrophy on CT scans obtained after a fewtreatments and at the end of the 4 years. In a similar vein,Menken et al. (1979) reported that a 30-year-old womanwho received 10 ECTs in a single 45-minute session showedno brain changes as measured by CT scans obtained beforethe session and 3 hours afterward. Most striking, however,is the report by Lippmann et al. (1985) of a patient whoreceived 1250 documented bitemporal ECTs over 26 years,with an additional 800 ECTs claimed by her withoutsupporting records. When she came to autopsy after herdeath at 89 years of age, the results of the neuropathologicexamination were normal. Also impressive are the long-termfollow-up neuropsycho logical test data from patients whohave received large numbers of ECTs (Chapter 10).

Neurochemical StudiesMyelin basic protein is an antigen that constitutes 30% ofthe myelin sheath, and its immunoreactivity in serum and

cerebrospinal fluid correlates with the degree of centralnervous system damage that occurs with stroke and cerebraltrauma. Hoyle, Pratt, and Thomas (1984) found nodifference in serially sampled serum myelin basic proteinimmunoreactivity between a sample of 13 patientsundergoing ECT and a sample of 14 normal controls, nor wasany pre-to post-ECT increase in mean reactivity observed inthe patient sample.

Serum neuron-specific enolase is elevated after stroke,cardiac arrest, and epileptic seizures. Greffe et al. (1996)presented preliminary evidence for increased enolase levelsinterictally in epileptics and patients with abnormal interictalEEGs, and in one sixth of the patients post-ECT. In a moresystematic study, Berrouschot et al. (1997) obtained bloodsamples for this enzyme at baseline, 1 minute post-seizure,and every 5 or 10 minutes thereafter for 2 hours, in 7depressed patients receiving ECT; no effect on serumenolase was found, and the authors concluded that ECT doesnot cause neuronal damage.

The study of Zachrisson et al. (2000) cited above in theblood-brain barrier section also examined cerebrospinal fluidconcentrations of three established markers of neuronal/glialdamage: tau protein, neurofilament, and S-100 beta protein.In 9 depressed patients studied before and after a course of6 ECTs, no ECT-induced changes in any of these markerswas detected; the authors concluded that ECT did not causeneuronal/glial damage.

SummaryFollowing publication of the 3rd edition of this volume(Abrams, 1992), Devanand et al. (1994) reviewed all of thesame material and reached the same conclusions, oftenexpressed with the same phrasing. There is simply noevidence, and virtually no chance, that ECT as presentlyadministered is capable of causing brain damage. Althoughabsence of proof does not constitute proof of absence, proofof absence is precisely that which science is incapable ofproviding: No experiment can prove that something does notexist. Reassuring in this regard are the overwhelminglypositive subjective assessments of post-ECT memoryfunction provided by those who have received ECT (seeChapter 10), a majority of whom experience their memoryafter ECT to be better than at any previous point in theirlife.

Considering that many depressed patients coming to ECT

P.66

already exhibit MRI and neuropsychological evidence forhippocampal atrophy (Sheline et al., 1999), the experienceof substantial improvement in memory function with ECT isconsistent with the phenomenon of ECS-inducedneurogenesis described above in rodents, which is associatedwith reversal of atrophy of stress-vulnerable neurons,protection from further damage, and increased neuronalsynaptic strength, survival, and growth (Duman and Vaidya,1998).

These authors make the further point that a course of ECS inrats increases the expression of brain-derived neurotrophicfactor and its receptor, leading to increased synapticstrength, survival, and growth of adult neurons (Lindefors,Brodin, and Metsis, 1995; Nibuya, Morinobu, and Duman,1995; Smith et al., 1997).

Based on important new evidence for the continuedexistence of the phenomenon of neurogenesis ofhippocampal neurons in adult humans (Eriksson et al., 1998;Gould et al., 2000), Jacobs, Praag, and Gage (2000) recentlyproposed that â!œtherapeutic interventions for depressionthat increase serotonergic neurotransmission [of which ECTis a putative example] act at least in part by augmentingdentate gyrus neurogenesis and thereby pro moting recoveryfrom depression.â! !

If this hypothesis is confirmed, it would not only contravenethe possibility of ECT-induced brain damage, but wouldsuggest a potential application of ECT for the purpose ofameliorating or partially reversing the structurally basedimpairment of memory function that characterizesneurodegenerative conditions such as Alzheimer's andParkinson's diseases.

Cardiovascular Physiology andMetabolism

Heart RateAnesthesia induction by itself increases baseline heart rateby about 25 (Usubiaga et al., 1967; Rollason, Sutherland,and Hall, 1971; Kitamura and Page, 1984; Wells and Davies,1987). During and immediately after administration of theelectrical stimulus there is a sharp vagal parasympatheticoutflow that is both neurally mediated and consequent to aValsalva effect induced by forced expiration against a closedglottis. Without anticholinergic premedication (e.g.,

P.67

atropinization), an intense but transient sinus bradycardiaoccurs (Bellett, Kershbaum, and Furst, 1941; Perrin, 1961;Clement, 1962), with periods of sinus arrest (cardiacasystole) averaging about 2 seconds but occasionallyrecorded in excess of 7 seconds (Clement, 1962;Gravenstein et al., 1965; Welch and Drop, 1989). The factthat even longer periods of bradycardia/asystole can occurin association with subconvulsive stimuli (Decina et al.,1984; Wells, Zelcer, and Treadrae, 1988) demonstrates thatthe electrical stimulusâ!”rather than the induced seizureâ!”is responsible for this vagal effect. Wyant and MacDonald(1980), who were unable to demonstrate significantbradycardia during ECT, reported the phenomenon in only 1of 39 patients observed during 297 seizures; however, theydid not specifically focus on heart rate during orimmediately after stimulus administrationâ!”which is thepoint of maximum bradycardiaâ!”but inexpli cably took meanhighest heart rate during the induced seizure as their primary measure of interest.

A sympathoadrenal tachycardia then supervenes (Figure 4-2), which is initially driven predominantly by directsympathetic neural outflow of discharging

cardioaccelerator areas in the hypothalamus, descendingipsilaterally by way of the medulla, upper thoracic cord,paravertebral stellate ganglia, and postganglionic cardiacnerves to the heart (Berne and Levy, 1981; Welch andDrop, 1989). Adrenal medullary catecholamine release laterin the seizure is supposed to contribute to maintaining heartrate above baseline during the late ictal and postictalphases, although the mean duration of the maximal phase ofthe ECT-induced tachycardia is significantly shorter thanthat of total paroxysmal EEG seizure activity (Larson,Swartz, and Abrams, 1984). Most authors (Gravenstein etal., 1965; Jones and Knight, 1981; Mulgaokar et al., 1985;Griffiths et al., 1989; Cuche et al., 1990) find plasmacatecholamine levels in patients receiving ECT to be higherduring the seizure than at baseline or 5 to 10 minutes later.

P.68

Figure 4-2 Heart rate and blood pressure changes with ECT.

However, the relevance of plasma catecholamine levels tothe ECT-induced tachycardia, however, is nullified by thereport of Listen and Salk (1990) of a 73-year-old womanwith bitemporal adrenalectomy who nonetheless exhibited aclassic hemodynamic response to ECT despite concurrentnifedipine administration: The mean rate-pressure productincreased by more than 50%, from a baseline value of13,897 to a postseizure peak of 21,069, and returned tobaseline over several minutes postseizure. The authors concluded that â!œa functionally intact adrenal medulla is notnecessary for the pressor response to ECT.â! !

A neural -neurohumoral model of the effects of ECT on heartrate is consistent with the results of 3 separate studies fromour group. In the first, Larson, Swartz, and Abrams (1984)found a close correspondence between the durations of theECT-induced tachycardia and the concurrent paroxysmal

EEG activity; the correlation between the point of maximalheart rate deceleration and cessation of paroxysmal EEGactivity was 0.75 (p < 0.001), a result consistent withprimarily neural chronotropic effects during the ictal phaseof ECT. In a second study (Figure 4-3), Lane et al. (1989)found greater postictal heart rates with bitemporal comparedwith right unilateral ECT and attributed this result to greater

P.69

brain-stem stimulation, and resultant greater inducedadrenal catecholamine release, with the former method.

In the most recent of the 3 studies (Swartz et al., 1994b),ECT-induced heart rate elevations after occurrence of thepeak heart rate were greater following right compared withleft unilateral ECT (Figure 4-4), a result consistent withreports of asymmetric autonomic innervation of the humanheart, with lateralization of sympathetic control of heartrateâ!”specifically, cardioaccelerationâ!”to the right cerebralcortex, diencephalon, medulla, spinal cord, and stellateganlgion (Rogers et al., 1978; Berne and Levy, 1981; Rosenet al., 1982; Cinca et al., 1985; Lane et al., 1988; Zamriniet al., 1990). This is the first study in neurologically intacthumans to demonstrate right-hemisphere superiority in thecontrol of heart rate within a paradigm of neuronalactivation, and it demonstrates that lateralization ofelectrode placement during ECT is reflected in lateralizationof brain-stem as well as cortical stimulation.

Heart rate drops rapidly at the termination of the seizure(Larson, Swartz, and Abrams, 1984)â!”falling to 50% of theictal rate within 5 seconds (Welch and Drop, 1989)â!”andgenerally returns to baseline or below it within minutes afterthe seizure ends (Tewfik and Wells, 1957; Perrin, 1961;

Deliyiannis, Eliakim, and Bellet, 1962; Kitamura and Page,1984; Wells and Davies, 1987; Huang et al., 1989),concomitant with declining plasma catecholamine levels(Gravenstein et al., 1965; Jones and Knight, 1981;Mulgaokar et al., 1985; Griffiths et al., 1989).

Figure 4-3 Heart rate effects of bitemporal compared with right unilateral ECT. (Lane et al., 1989.)

Figure 4-4 Heart rate after left compared with right unilateral ECT.

P.70

Blood PressureBlood pressure generally parallels heart rate throughout thetreatment (Figure 4-2), dropping sharply during the initialvagal hypertonic phase and then rapidly increasing 30% to40% over baseline to reach peak systolic values that oftenexceed 200 mm Hg in the presence of atropinepremedication (Perrin, 1961; Bodley and Fenwick, 1966;Rollason, Sutherland, and Hall, 1971; Mulgaokar et al.,1985; Prudic et al., 1987; Wells, Davies, and Rosewarne,1989). The fact that intraarterial systolic pressures peakwithin 7 seconds of the first poststimulus heartbeat (Welchand Drop, 1989) suggests that, as for heart rate, the firstphase of the ECT-induced hypertensive response isneuronally mediated. The initial rise in systolic pressure isproportionately greater than the initial rise in diastolicpressure, is more pronounced in hypertensive thannormotensive patients, and in males than females (Prudic etal., 1987).

Early reports that the hypertensive response to ECT wasrelated to its cognitive side effects (Hamilton, Stacker, andSpencer, 1979; Stoker, Spencer, and Hamilton, 1981) werenot confirmed in subsequent studies of the hemodynamicresponse to ECT (Taylor, Kuhlengel, and Dean, 1985;O'Donnell and Webb, 1986; Webb et al., 1990).

Rate Pressure ProductThe rate pressure product (RPP), systolic arterial pressure Ã— heart rate, is a rough indicator of myocardial oxygenconsumption. It increases 30% to 140% during the seizure(Jones and Knight, 1981; Mulgaokar et al., 1985; Huang etal., 1989; Webb et al., 1990), reaching its maximum about30 seconds after the ECT stimulus; this response issubstantially attenuated by beta adrenergic receptorblockade (Jones and Knight, 1981). The RPP and, of course,heart rate and blood pressure, vary inversely with age andbaseline RPP (Huang et al., 1989; Webb et al., 1990).

Cardiac OutputWells and Davies (1987) used noninvasive electricalbioimpedance monitoring in 10 patients to determine thatcardiac output (ventricular stroke volume Ã— heart rate)immediately following the ECT-induced seizure increasedover preanesthesia baseline by an average of 81% and

P.71

returned to baseline within 2 minutes. Using the same makeand model of equipment, however, Huang et al. (1989)found that cardiac output divided by body surface area (theCardiac Index) remained unchanged throughout the ictal andpostictal period in 13 patients receiving ECT. These authorspostulated an increase in peripheral vascular resistance asan explanation for their findings.

Electrocardiographic EffectsAbnormalities of cardiac rhythm and conduction are recordedmuch more frequently just after the induced seizure thanduring it, and are classified as either vagal or sympathetic.The vagal arrhythmias are of atrial, junctional, or nodalorigin and include sinus bradycardia, sinus arrest, atrialpremature contractions, paroxysmal atrial tachycardia(atrioventricular junctional tachycardia), atrial flutter, atrialfibrillation, atrioventricular block (first-, second-, and third-degree), and premature ventricular contractions duringperiods of sinus bradycardia. The sympathetic arrhythmiasoriginate in the ventricles as premature contractionsoccurring during sinus tachycardia, bigeminy, trigeminy,ventricular tachycardia, and ventricular fibrillation (Perrin,1961; Elliot, Linz, and Kane, 1982; Pitts, 1982; Dennisonand French, 1989).

Electrocardiographic (ECG) repolarization abnormalitiesreported during and immediately after ECT include increasedT-wave amplitude, T -wave inversion, and ST-segmentdepression of nonischemic and ischemic types (Lewis,Richardson, and Gahagan, 1955; Green and Woods, 1955;Deliyiannis, Eliakim, and Bellet, 1962; Graybar et al., 1983;Dec, Stern, and Welch, 1985; Khoury and Benedetti, 1989),all of which are reported to be entirely benign. Theprevalence of ECT-induced ECG abnormalities is increased in

patients with preexisting cardiac pathology and was reportedby Pitts et al. (1965) to be significantly greater afterthiopental than after methohexital barbiturate narcosis.Although this latter point has been repeatedly stressed inthe ECT literature (Fink, 1979; Abrams, 1988b; AmericanPsychiatric Association, 1990), it is far from universallyaccepted (Selvin, 1987), and was not confirmed in therecent study of Pearlman and Richmond (1990).

The occurrence of ECG abnormalities is essentially limited tothe ictal and immediate postictal periods. Extensiveexaminations, including Holler monitoring, conducted up to

24 hours post-ECT do not reveal persistent ECG changes,even in patients with preexisting cardiovascular disease(Troup et al., 1978; Kitamura and Page, 1984; Dec, Stern,and Welch, 1985). These latter authors have also stressedthe neural origin of these phenomena, attributing them todirect electrical stimulation of brainstem subcorticalstructures and thalamic and hypothalamic nuclei. Many ofthese phenomena (e.g., ectopy, repolarization abnormalities)are so regularly observed in young patients with no historyor symptoms of cardiovascular disease that they must beconsidered normal physiologic concomitants of ECT and notin any way contraindications to continuing it (Deliyiannis,Eliakim, and Bellet, 1962).

Cardiac EnzymesDuring the hours after ECT, significant elevations occur inserum levels of creatine phosphokinase and lactatedehydrogenase, but not in glutamic oxalaminasetransaminase (Rich et al., 1975a,b; Braasch and Demaso,1980; Dec, Stern, and Welch, 1985). Creatinephosphokinase is found in the skeletal muscle, myocardium,brain, and gastrointestinal tract; only the creatinephosphokinase muscle-brain isoenzyme is specificallyelevated after myocardial damage (and the brain creatinephosphokinase isoenzyme does not cross the blood-brainbarrier in the absence of brain damage). Lactatedehydrogenase is found in most human tissues, includingskeletal muscle and myocardium; its lactate dehydrogenase-1 and lactate dehydrogenase-2 isoenzymes are consideredfairly specific for myocardial damage. None of themyocardiospecific isoenzymes are significantly elevated whentested at mul tiple intervals up to 96 hours after ECT(Braasch and Demaso, 1980; Taylor, von Witt, and Fry,1981; Dec, Stern, and Welch, 1985).




> Table of Contents > Chapter 5 - Electroconvulsive Therapy in the

High -Risk Patient

Chapter 5

Electroconvulsive Therapy inthe High-Risk Patient

Electroconvulsive therapy is a low-risk procedure. Itsinherent safety, combined with an increasingly preciseunderstanding of its medical physiology and the widespreadavailability and application of advanced monitoringtechniques, now enables its routine and successfulapplication in a group of patients previously believed to betoo old or too physically ill to undergo the stress of inducedconvulsions: the â!œhigh-riskâ! ! patient to whom thischapter is devoted. The most recently published mortalitystatistic of about 2 deaths per 100,000 treatments (Kramer,1985; 1999) has remained constant over almost 20 years,and places ECT at the low end of the risk range reported foranesthesia induction alone (Fink, 1979)â!”lower, in fact,than the risk for childbirthâ!”a development that is all themore gratifying in view of the steadily increasing averageage of patients for whom ECT is pre scribed.

It is instructive to compare these figures with recentlypublished spontaneous death rates in an age rangecompatible with that of typical ECT samples. In a 10.5-yearfollow-up of 3657 community residents aged 65 and older,Glynn et al. (1995) found that 1709 had died, equivalent toa rate of 4.45% per year, or 1 in 195 over each 6-weekinterval, a death rate that was linearly related to restingblood pressure. It is thus apparent that the death ratereported for patients undergoing ECT is orders of magnitudesmaller than the spontaneous death rate of a comparablyaged sample.

P.73

Many published studies abundantly confirm the fact that ECTreduces age-specific mortality rates independent of anyeffect in reducing suicide rates (Prudic and Sackeim, 1999).It would be tempting to speculate that the routine medicalscreening procedures for ECT are effective in eliminating thehigh-risk patients, were it not for the fact that the high-risk, geriatric cardiac patient is fast becoming the modalcandidate for ECT. Because the presence of depression is asignificant risk factor for death from cardiac disease inpatients both with and without preexisting cardiac disease,and because sudden cardiac death accounts for most of theexcess mortality in patients with estabished heart disease(Carney, Freedland, and Jaffe, 2001), it is reasonable toguess that ECT exerts its beneficial effects on over-allmortality rates by reducing depression, although ECT-induced reduction of resting blood pressure (Swartz andInglis, 1990) may also play a role.

It is further instructive to view the ECT-specific mortalityrates cited above in light of the report of a 17-fold overallincrease in the risk of fatal myocardial infarction, and a 3-fold increase in the risk of subarachnoid hemorrhage, inyoung women (aged 16 to 39 years) currently takingpsychotro pic drugs, particularly tricyclic antidepressantsand benzodiazepines (Thor ogood et al., 1992).

As noted in the previous chapter, the brain, heart, andblood vessels bear the brunt of the physiologic impact ofECT, hence it is patients with compromised cerebral andcardiovascular function who constitute the pop ulation atgreatest risk during the procedure.

Management of Cardiovascular RisksThe gravest potential cardiovascular complications of ECTvirtually never occur during the procedure; these includeacute myocardial infarction, acute coronary insufficiency,ventricular fibrillation, myocardial rupture, cardiac arrest,cardiovascular collapse, stroke, and ruptured cerebral oraortic aneurysm. They are so rare, in fact, that none werereported in a multihospital study in Denmark of 22,210consecutive ECT treatments (Heshe and Roeder, 1976).

The detection and management of significant cardiovascular

P.74

disease before administering ECT is overwhelmingly the mostimportant factor in reducing consequent cardiovascularmorbidity and mortality: One need only contemplate thedifference in the risk of ECT to a patient in acute congestiveheart failure before and after he has been stabilized ondigitalis and diuretics. The increasing number of olderindividuals with significant cardiovascular disease is amplyrepresented among patients referred for ECT, and a greatdeal of experience has been accumulated in recent years inthe pharmacologic management of such â!œhigh-riskâ! !patients (Elliot et al., 1982; Weiner, 1983b; Alexopolous etal., 1984a; Dec, Stern, and Welch, 1985; Regestein andReich, 1985; Welch and Drop, 1989). The management ofECT-induced alterations in cardiac rate, rhythm, and bloodpressure has received the most attention because thesephenomena have multiform potential adverse effects in thepresence of preexisting cardiovascular disease.

Vagal Cardiovascular EffectsAlthough it has been routine practice for many years toattempt to attenuate or abolish the vagal effects of ECT byadministering an anticholinergic agent (e.g., atropine) beforetreatment, some authors, including those of the RoyalCollege of Psychiatrists' ECT Handbook (1995), havesuggested that such drugs might best be reserved forpatients with hypodynamic cardiac states or a tendency toprolonged asystole during ECT and should be avoided inpatients with hyperdynamic states, such as hypertension ortachycardia (Bouckoms et al., 1989; Shettar et al., 1989).

A study of the effect of atropine premedication on the rate-pressure product (RPP) recorded an unsurprisingly largerincrease in this measure with atropine than with placebo,but the authors' recommendation to avoid atropine for ECTpremedication except prior to seizure thresholddetermination was not supported by evidence for anyharmful effects of the increased RPP. Indeed, the potentiallyfatal cardiac arrest that has been reported in patientsreceiving ECT without anticholinergic premedication wouldscarcely have been expected to occur in the small sample of30 patients studied, so the authors' failure to detectclinically significant bradyarrhythmia in the placebo-treated

group is not at all reassuring.

Atropine remains the drug of choice for attenuating orblocking the direct vagal effects on the heart during andimmediately after the passage of the electrical stimulus andin the immediate postictal period: sinus bradycardia andarrest (and the consequent sharp drop in blood pressure),the atrial and junctional arrhythmias, and ventricularpremature contractions dur ing sinus bradycardia.

Glycopyrrolate is an anticholinergic alternative to atropinethat does not cross the blood-brain barrier and is thereforepotentially less likely to exacerbate post-ECT confusion thanatropine, a known deliriant when given at higher doses thangenerally used during ECT. However, although oneprospective comparison of these two agents for ECTpremedication (Kramer, Allen, and Friedman, 1986) foundglycopyrrolate to be associated with less post-ECT confusionthan atropine, a cognitive advantage of glycopyrrolate wasnot confirmed by other investigators (Kelway et al., 1986;Sommer, Satlin, and Friedman, 1989; Swartz and Saheba,1989). Moreover, Calev et al. (199la), in an open clinicaltrial, were unable to detect any adverse cognitive effects ofatropine, 0.5 mg intravenously, on a comprehensive testbattery administered at baseline and sequentially over 110minutes postic tally. Thus, there is no rationale forsubstituting glycopyrrolate for atropine for ECTpremedication.

Although Cropper and Hughes (1964) recommend no lessthan a vagolytic dose of atropine (2.0 mg), a systematicstudy of the cardiac effects of four dosage levels of atropine(1.0 mg, 1.5 mg, 2.0 mg, and 2.5 mg) for ECTpremedication revealed no advantage of exceeding a 1.0 mgdose, administered intramuscularly 45 to 60 minutes beforetreatment (Rich et al., 1969).

Prolonged Asystole/Cardiac ArrestThe unopposed vagotonicity of subconvulsive stimuli haslong been known (Perrin, 1961) and is responsible for theadmonition (Wells, Zelcer, and Treadrae, 1988; Welch andDrop, 1989) that such stimuli may represent a threat to thecardiac patient by inducing a strong parasympatheticstimulus that remains unopposed by a seizure-induced

P.75

sympathetic response. In the case reported by Wells, Zelcer,and Treadrae (1988), a 15-second asystole

occurred despite premedication with 0.4 mg atropineintravenously during the first ECT given to a 75-year-oldhypertensive man who had sustained an inferior myocardialinfarction a year earlier. McCall (1996) described a 55-year-old woman who sustained a 30-second asystole immediatelyfollow ing a subconvulsive stimulus, despite pretreatmentwith 0.2 mg glycopyr rolate intramuscular 1 hour beforeECT.

These instances of prolonged asystole/cardiac arrestfollowing subconvulsive stimulation suggest that the â!œmethod of limitsâ! ! titration procedure for determining theECT seizure threshold (Sackeim et al., 1987a) which requiresthe administration of up to several subconvulsive stimuli in asingle treatment sessionâ!”provides a degree of cardiac risk.As Wells, Zelcer, and Trendrae (1988) found duringsubsequent treatments of their patient, even maximal dosesof atropine (e.g., 2.0 mg intravenously) may not prevent thesevere bradycardia that can result from subconvulsivestimuli.

Although the chart-review study of Zielinski et al. (1993)â!”who compared 40 patients with and 40 without cardiacdisease who underwent ECT â!”found no excess in theoverall rate of cardiac complications for sessions in whichsubconvulsive stimuli were used, all 5 episodes ofbradycardia occurred in association with subconvulsivestimuli (p = 0.016). Two of the episodes required medicalintervention, and 1 (see Decina et al., 1984) re sulted in acode 99 being called for cardiac arrest. (It is further notablethat an anticholinergic had not been given in 4 of theinstances.) In a prospective study of ECG and cardiovasculareffects of subconvulsive stimulation during titrated ECT in 40patients (22 with a history of active cardiovascular disease),McCall, Reid, and Ford (1994) found that despiteintramuscular premedication with low-dose glycopyrrolate 90minutes prior to ECT, subconvulsive stimuli prolonged the R-R interval and slowed the heart rate compared withbaseline, and were associated with occasional junctionalblock and â!œbrief sinus pauses,â! ! otherwise unspecified.(However, convulsive stimuli had an even greater effect on

P.76

the R-R interval.) Bradycardia/cardiac arrest can also occurfollowing the administration of beta-adrenergic blockingagents, which prevent the tachycardia response of theinduced seizure from reversing the initial vagal bradycardia.The cases reported by Liebowitz and El-Mallakh (1993) andKaufman (1994), and 2 of those reported by McCall (1996),clearly illustrate this phenomenon be cause none wasassociated with a subconvulsive stimulus.

Liebowitz and El-Mallakh (1993) reported a 15-secondasystole (characterized as cardiac arrest) during the fifthECT of an 80-year-old man with hypertension, coronaryartery disease, and an old anterior myocardial infarction,who had received premedication with a 10-mg dose oflabetalol intravenously without concomitant anticholinergicpremedication. Kaufman's (1994) patient tolerated the first 4ECTs without difficulty, but then unaccountably received 15mg of labetalol intravenously just prior to the fifth ECTstimulus (given without anticholinergic premedication), whichwas immediately followed by a 30-second asystole. McCall(1996) described 2 cases of prolonged asystole occurring inthe absence of subconvulsive

stimulation. These occurred in association withadministration of the beta-blocker esmolol, despiteintramuscular premedication with 0.1 mg glyco pyrrolate 1hour before pretreatment in one instance, and atropine, 0.2mg intravenously, immediately before treatment in theother.

The 15-second asystole reported by Decina et al. (1984),also cited by them as an instance of cardiac arrest, couldhave resulted from either mech anism, as the patientreceived both a beta-blocker and a subconvulsive stim ulus(again, in the absence of anticholinergic premedication).

The fact that several instances of prolonged asystoleoccurred despite premedication with 0.2 mg of atropineintravenously, or intramuscular administration of 0.2 mgglycopyrrolate 1 hour before ECT, suggests that theseanticholinergic regimens are insufficient for the purpose. Asnoted above, I consider intravenous atropine, 0.4 to 1 mg,administered immediately prior to anesthesia induction, tobe the anticholinergic method of choice for block ing thevagal effects on heart rate.

P.77

Although all the patients described in the studies citedabove survived their episodes of prolonged asystole orcardiac arrest, I nevertheless have some difficultyconsidering the phenomena in question benign. I certainlyagree with the recommendation of Zielinski et al. (1993) toadminister atropine before treatment whenever stimulustitration is to be performed, but it is clear from some of thecases cited that this precaution may be insuffi cient.

In summary, the occurrence of subconvulsive responses, theadministration of beta-adrenergic blocking agents, and theomission of adequate anticholinergic premedication allincrease the risk of bradycardia, prolonged asystole, andcardiac arrest during ECT. The following recommendationsshould virtually eliminate the risk of such cardiac events,without in any way reducing efficacy.

1. Always administer a vagal-blocking dose of atropine(e.g., at least 0.4 mg) intravenously immediately priorto anesthesia induction in patients who are receivingbeta-adrenergic blocking agents, or for whom stimulustitration is planned. Glycopyrrolate should not beconsidered equivalent to atropine for this purpose.

2. Avoid stimulus titration in patients with clinicallysignificant cardiac dis ease.

Because no one claims or has demonstrated that therelatively frequent, entirely unpredictable, and occasionallyprolonged periods of ECT-induced cardiac asystole arewithout risk, and until and unless adverse consequences ofgiving anticholinergic premedication for ECT aredemonstrated, a risk-benefit analysis stipulates that ananticholinergic agent should be given before ECT. Moreover,because excessive salivation is not known to complicateECT, and parenteral administration of anticholinergics 30 to60 minutes before ECT serves only to cause anuncomfortable dry mouth that may stimulate the patient todrink water, I further recommend that the anticholinergic

be given intravenously immediately preceding anesthesiainduction. Finally, since glycopyrrolate has no demonstrableadvantage over atropine, a drug that has been intensively

studied for ECT premedication over several decades, thelatter remains, in my view, the anticholinergic agent ofchoice for ECT.

Sympathetic Cardiovascular EffectsAttenuation or blockade of the acute hemodynamic andmyocardial consequences of ECT may be desirable in patientswith brain tumors, cardiac conduction defects or ectopy,hypertension, recent myocardial infarction or hemorrhagicstroke, and aortic or cerebral aneurysms.

Agents that Lower Blood PressureIn the previous edition, I devoted considerable text to theconsideration of several potent intravenous agents forlowering blood pressure, such as trimethaphan (no longeravailable), sodium nitroprusside, and hydralazine. Theintervening years have shown less heroic measures to besafer and equally effective for this purpose: nitrates,calcium channel blockers, and beta-adrenergic blockingagents, in order of increasing efficacy. (An exception may bethe use of intravenous nitroprusside to control bloodpressure in the presence of an intracranial aneurysm, asdescribed below.)

1. Of the nitrates, both nitroglycerine ointment andsublingual spray provide modest to moderateattenuation of the hypertensive response to ECT (Leeet al., 1985; Villalonga et al., 1989; Parab, Chaudhari,and Apte, 1992). A typical regimen of 2 inches of 2%nitroglycerine ointment applied to the anterior chest 45minutes before ECT achieves a 15% to 20% reductionin the systolic pressure recorded 2 minutes post-ECT(Parab, Chaudhari, and Apte, 1992). However,nitroglycerine can induce substantial heart rateelevations (O'Flaherty et al., 1992), and thereforeincrease myocardial oxygen consumption, an effect thatis likely to be undesirable in many cardiac patients.

2. Nifedipine is a calcium channel blocking agent thatrelaxes the smooth muscle of coronary arteries andperipheral arterioles. Its vasodilating action has beensafely and effectively used to manage acute

P.78

hypertensive crises (Schillinger, 1987), and to preventor attenuate the ECT-induced hypertensive response.Sublingual administration of the contents of one 10-mgnifedipine capsule 20 minutes before anesthesiainduction reduced the maximum ECT-induced systolicblood pressure elevation by almost two thirds (24 mmHg vs. 62 mm Hg) in a placebo-controlled study of 5previously hypertensive patients who served as theirown controls (Wells, Davies, and Rosewarne, 1989). Ina single case report (Kalayam and Alexopoulos, 1989),

the same method reduced the maximum systolicpressure increase by almost 90%, from 300 mm Hg to160 mm Hg. Heart rate was unaffected in both studies,and no complications were reported, but tachycardiacan occur.

3. The beta-blockers are generally the most useful andwidely used of the 3 classes of medications describedin this section. Although propranolol was the first to beused in association with ECT, subsequent experiencewith this compound for this purpose has been neitherextensive nor particularly favorable (Gaitz and Essa,1991; Maneksha, 1991), and it has been abandoned.

Labetalol is the most systematically studied beta-blocker foruse during ECT. In a single case report of a 74-year-oldwoman whose post-ECT hypertension (e.g., 240/140 mm Hg)was resistant to clonidine and hydralazine, Foster and Ries(1988) achieved success with labetalol, 15 mg intravenouslyimmediately after her seizure, followed by maintainance oraltherapy. In a randomized, double-blind, placebo-controlledstudy in elderly depressive patients of the effects of 5 to 10nig labetalol given intravenously 90 seconds before seizureinduction, Stoudemire et al. (1990) reported significantreductions in ECT-induced mean arterial blood pressure,tachycardia, and atrial and ventricular ectopy over a 30-minute observation period, without untoward side effects. Ina subsequent dose-response study of similar design, McCalland coworkers (1991) found that 5-mg and 10-mg doses oflabetalol safely and effectively lowered blood pressure andrate-pressure product (but not heart rate), withoutshortening seizure duration.

P.79

It is notable that Holtzman et al. (1986) found thathypertensives who receive single doses of atenolol,propranolol, and labetalol significantly reduced their cardiacoutput and heart rateâ!”but not their blood pressureâ!”with the two former drugs, but exhibited precisely thereverse pattern with labetalol: It lowered blood pressurewithout affecting heart rate or cardiac outputâ!”primarily byreducing peripheral vascular resistance. This ability oflabetalol to lower blood pressure without reducing cardiacoutput is of obvious import in the treatment of high-risk orgeriatric patients who need to maintain their cerebral andmyocardial perfusion to meet the increased met abolicdemands of the induced seizure.

Esmolol, a selective beta-1 receptor blocker and theshortest -acting of the beta-blockers (distribution half -life, 2minutes, elimination half -life, 9 minutes), was used byKovac et al. (1990) to attenuate the hemodynamic responseto ECT in a randomized, within-patient comparison againstno esmolol in 17 subjects. An intravenous bolus of 80 mgesmolol, followed by a 24-mg/minute infusion, bluntedmaximal heart rate response by 26%, blood pressureresponse by 14%, and RPP by 37%, during the 4 minutesimmediately following the stimulus. Equivalent effects wereobtained with a 100-or 200-mg bolus injection of the druggiven 2 minutes before ECT. Compared to the no-esmololcondition, however, esmolol substantially, but

not significantly, shortened seizure duration by 15% to 25%,depending on the dose.

Similar results were obtained by Howie et al. (1990), whoused an analagous design to study 20 patients in a double-blind, randomized, within-subjects study of placebocompared with esmolol. The authors administered a 10-to15-minute infusion of either placebo or esmolol at an intialrate of 500 (jLg/kg/min, decreasing to 300 jjug/kg/min for 3minutes postictally, before ceasing altogether. Heart rateand blood pressure were monitored every minute during theinfusion and every 5 minutes thereafter. Esmololsignificantly reduced baseline (prestimulus) heart rate belowthat achieved with placebo and sustained this reduction over2 to 15 minutes poststimulus, reducing maximum heart rateby 24% (from 152 to 115 beats/minute). Arterial blood

pressure was also significantly lower with esmolol, systolicmore than diastolic. Like Kovac et al. (1990), these authorsfound that esmolol reduced EEG seizure length by about22% (from a mean of 86 to 67 seconds). In a subsequentdose-response study of 3 dosage levels of esmolol, Howie etal. (1992) found that an initial intravenous bolus of 500|Jig/kg followed by an intravenous infusion of 100(xg/kg/min was as effective as higher dosages in controllingthe hemodynamic response to ECT, and with out significantlyreducing seizure duration.

Weinger et al. (1991) compared the relative efficacies oflabetalol and esmolol in blocking the hemodynamic responseto ECT. In a double-blind, intrapatient, balanced randomizedcomparison among fixed, single-bolus doses of intravenousesmolol, labetalol, fentanyl, lidocaine and saline placeboconducted in 10 patients, these authors found labetalol (0.3mg/kg) and esmolol (1 mg/kg) to be similarly effectivecompared with saline in attenuating the ECT-inducedhypertension and tachycardia (e.g., RPP increase wasattenuated 46% after labetalol, and 64% after esmolol).Both compounds shortened seizure duration (esmolol by 19,labetalol by 35), an effect that reached significance only forthe latter drug. Although these authors conclude byexpressing a preference for esmolol, it is clear from theirstudy and the others discussed earlier that with properadjustment of dosage, either drug is capable of safely andeffectively attenuating the hemodynamic re sponse to ECTwithout unduly shortening seizure duration.

Castelli et al. (1995) studied 18 patients with at least onecardiovascular risk factor during 5 ECT sessions in which 5different pretreatment regimens were administered in arandomized block design: no drug; esmolol, 1.3 or 4.4mg/kg; labetalol, 0.13 or 0.44 mg/kg, all administered bybolus push within 10 seconds of anesthesia induction.Systolic and diastolic blood pressure, heart rate, and ST-segment deviation were measured at baseline, and 1, 3, 5,and 10 minutes postictally. Compared with no drug, the low-dose beta-blocker conditions significantly reduced, and thehigh-dose conditions virtually eliminated, the peak heartrate and systolic blood pressure responses. ST-segmentdeviations were not affected by either compound at eitherdosage level. The effects of labetalol, but not esmolol, on

P.80

systolic blood pressure could still be observed 10 minutespostictally.

Labetalol and nifedipine have also been combined to takeadvantage of the heart rate control achieved with theformer, and the blood pressure control with the latter (Figielet al., 1993). In a sample of 38 patients over the age of 50years, these authors gave 10 mg of labetalol intravenouslyat the first ECT treatment and observed 10 patients whonevertheless exhibited sustained systolic blood pressureelevations (2 consecutive readings > 210 mm Hg) despiteadequate control of heart rate. These 10 were then treatedwith nifedipine, 10 mg sublingually 15 minutes pre-ECT, incombination with the dose of labetalol already described.The addition of nifedipine appeared to achieve substantialcontrol of blood pressure at all measurement intervalswithout further affecting heart rate; no episodes ofhypotension or bradycardia occurred. Unfortunately, thestatistical analysis is flawed (multiple paired t -tests ratherthan repeated-measures ANOVA were used), and the authorsfailed to describe the response vis -a-vis the criterion forentry in the studyâ!”that is, how many of the 10 patientsexhibited sustained blood pressure elevations despite thedrug combination. The relatively low dose of labetalol usedâ!”equivalent to the low-dose condition of the Castelli et al.(1995) study reviewed in the previous paragraphâ!”alsoraises the ques tion whether simply increasing the labetaloldose would have achieved the same results.

The beneficial effects of beta-blockers on cardiac ectopy,blood pressure, and myocardial oxygen consumption duringECT are, however, entirely safely enjoyed only aftertreatment with atropine. Based on studies in dogs that hadelectrically induced seizures in the presence of high spinalanesthesia, Anton, Uy, and Redderson (1977) predicted thatECT-induced â!œactivation of the autonomic nervous systemin the presence of a sympathetic block would lead to avagally induced protracted asystole.â! ! The aptness of thiswarning is illustrated by several of the case reportsdescribed above (Decina et al., 1984; London and Glass,1985; Kaufman, 1994; McCall, 1996).

Agents that Reduce Heart Rate and

P.81

Agents that Reduce Heart Rate andEctopyPrevention and management of the sympathoadrenal cardiacarrhythmias of ECT involve the use of lidocaine or beta-blockers. The most frequent of these extravagaltachyarrhythmias are ventricular premature contractions(VPCs) occurring late in the seizure or in the immediatepostictal phase. Occasional VPCs are of no concern; it is thefrequent or multifocal ones that present a risk because ofthe increased chance that one might coincide with the apexof the T wave of the ECG and precipitate ventricular tachycardia or fibrillation (Lown, 1979).

1. Lidocaine (xylocaine) is a local anesthetic agent thatprolongs the refractory period of the cardiac conductionsystem and increases the myocardial threshold toabnormal stimulation; it has been used successfully formany years to control ECT-induced tachyarrhythmias(Usubiaga et al., 1967;

McKenna et al., 1970; Hood and Mecca, 1983; Londonand Glass, 1985). It is available for intravenousadministration in 100-mg ampules. In patients withpreexisting multiple ventricular premature contractions,a constant intravenous infusion is administered at arate of 1 to 5 mg/min, permitting moment-to -momenttitration of ventricular ectopic activity. For quickcontrol of multiple ventricular premature contractionsthat develop for the first time during ECT, a rapidbolus push of 50 to 100 mg of lidocaine (1 to 2 mg/kgbody weight) is safe and effective. A significantdrawback of intravenous lidocaine, however, is that itshortens (Ottosson, 1960; Usubiaga et al., 1967;Weinger et al., 1991) or even abolishes (Hood andMecca, 1983; London and Glass, 1985) ECT-inducedseizure activity when given at the preceding doses,thus partially or entirely blocking the therapeutic effectas well.

2 . Of the beta-blockers, labetalol and esmolol aretherefore of particular interest because of their moremodest effects on seizure duration (McCall et al., 1991;Weinger et al., 1991).

P.82

Treatment of the High-Risk GeriatricCardiac PatientBecause virtually all cardiac patients referred for ECT areover 50 years of age (Zielinski et al., 1993; Rice et al.,1994), it is reasonable to combine geriatric and cardiacpatients for purposes of this discussion.

Burke et al. (1985, 1987) employed a retrospective chart-review method to assess the safety and efficacy of ECT in atotal sample of 70 patients over age 60 years, many withcardiovascular disease. In the first study, 5 of 30 patientswere described as having â!œcardiovascularâ! ! complicationsfrom ECT (including 1 death); in the second study, 15 of 40were described as having â!œcardiorespiratoryâ! !complications, for an apparent overall risk of about 30%,which led the authors to warn in the first paper that ECT â!œshould be used with caution, particularly in those overthe age of 75 years with cardiovascular disease,â! ! and inthe second paper that â!œat particular risk [with ECT] arethe very old, those in poor general health, and those takingmultiple medications, particularly cardiovascular agents.â! !Because they bear the imprimatur of Washington University,these articles are often citedâ!”together with the article byGerring and Shields (1982), discussed belowâ!”in support ofwithholding ECT in the aged patient with cardiovasculardisease.

However, the lack of a control group of elderly, medically illhospitalized psychiatric patients who did not receive ECTmakes it impossible to assess whether allâ!”or even anyâ!”of the adverse affects claimed by Burke et al. (1985,1987) can be attributed to ECT. Moreover, insufficientdetails of the claimed adverse cardiac effects are presentedto allow the reader to judge their severity. For example, inthe first article (Burke et al., 1985), one of the 5 adverseeffects is described only as â!œan episode of hypotension,â! ! without providing even a blood pressure value in supportof this diagnosis. In another instance, a sustained elevatedblood pressure of

230/110 is claimed as an adverse effect, without anymention of what the patient's baseline values were or howlong the elevation lasted. Worse yet, the death these

authors attributed to ECT actually occurred several weeksafter treatment, and in a patient who was takingtheophylline at the time of ECT, a drug known to havecaused severe complications and death when coadministeredwith ECT. The second article (Burke et al., 1987) commitsthe same errors as the first, claiming 2 episodes of â!œmarked, sustained elevation of blood pressureâ! ! withoutproviding the actual values or comparison with baseline; anepisode of â!œhypotension following treatment,â! ! otherwiseunspecified; another episode of â!œsustained tachycardiafollowing treatment,â! ! otherwise unspecified; and so forth.In sum, both articles have to be excluded from considerationas lacking even a modicum of objectivity or scientificvalidity.

In addition to the question of the cardiac risk ofsubconvulsive stimulation, Zielinski et al. (1993) comparedthe overall rates of cardiac complications in their samples ofpatients with and without cardiac disease who received ECT.Cardiac medications were adjusted as necessary, and whenthe regimens were stable for 7 to 10 days, ECT wasadministered without specific additional cardiotropicmedications. Although the patients with cardiac disease hadsignificantly higher rates of cardiac complication during ECT,no deaths occurred in this group, and only 5% failed tocomplete the treatment course. This is remarkable in view ofthe criteria for inclusion in the cardiac group, which requiredat least one of the following: left ventricular ejectionfraction 50% by radionucleide angiography; QRS complex ofat least 0.1 second by ECG; at least 10 VPCs per hour byHolter monitoring, and the fact that 15 patients had ahistory of myocardial infarction.

The cardiac complications observed in the cardiac groupâ!”many of whom were referred for ECT because of theintolerable cardiovascular side effects of tricyclicantidepressantsâ!”included 15 instances of ventriculararrhythmias (2 of ventricular tachycardia), 9 ischemic events(including one episode of myocardial infarction); 6 instancesof atrial arrhythmias; and 5 episodes of bradycardia. Eightpatients had major complications consisting of persistentECG changes accompanied by chest pain, asystole, orpersistent arrhythmia.

Thus, although the 40 patients with preexisting cardiac

P.83

disease suffered more, and more severe, cardiaccomplications during ECT than a matched control group ofnoncardiac patients, and had received no specific therapywith cardiotropic medications at the time of ECT, intensivemedical man agement of complications once they occurredenabled the vast majority to safely and effectively completetheir course of treatment.

Rice et al. (1994) conducted a case-control chart review ofconsecutively treated patients over age 50 years that yieldeda group of 26 at increased risk for cardiac complications,and a group of 21 at standard risk. No specific cardiotropicmedications were administered at the time of ECT. Unlikepatients in the study of Zielinski et al. (1993), patients inthe cardiac risk group were no more likely to experience amajor complication during

ECT than patients in the no-risk group; minor cardiaccomplications, however, were twice as frequent in the riskgroup. Three of 26 risk group patients suffered complicationssevere enough (e.g., ventricular tachycardia) to necessitatetransfer to the medical service; 2 of these (8%)discontinued ECT, and none sustained lasting sequelae. Theresults of this study stand in striking contrast to the widelycited report of Gerring and Shields (1982), two residentswho unaccountably found an extreme and unacceptable riskof cardiac morbidity and mortality in cardiac patientsreceiving ECT at the same institution as Rice et al. (1994),but 20 years earlier. Gerring and Shields' (1982) findingshave never been replicated and must therefore beconsidered aberrant.

As might be expected, the stress of ECT on the heart issimilar to that which might result from vigorous exercise.Messina et al. (1992) assessed left ventricular regional wallmotion via echocardiogram in 11 patients (mean age, 54years) before and immediately after ECT, and found that in3 patients (2 of whom had coronary risk factors and a familyhistory of heart disease), the stress of ECT on the heart wassimilar to that typically observed during treadmill exercisetesting, producing signs suggestive of reversible myocardialischemia. No patient experienced myocardial infarction, andall ECGs and echocardiograms returned to normal by the endof the study.

P.84

Thus, one may conclude that ECT is generally safe andeffective in geriatric patients with or without preexistingcardiac disease; that patients with cardiac disease have amoderately increased risk of untoward cardiac events (and asmall risk of death) during ECT; and that these risks areeffectively managed with appropriate cardioactivemedications administered before, during, or after treatment.Indeed, it is precisely in this group of older cardiac patientsreferred for ECT that Figiel and Stoudemire (1994) haverecommended routine pretreatment with beta-blockersâ!”particularly la-betalolâ!”alone or in conjunction withnifedipine. Considering the relatively benign nature of thesemedications, and the potential severity of some of theadverse cardiac events reported (e.g., ventriculartachycardia, myocardial infarction), this recommendationappears both reasonable and prudent. The initialprophylactic dose of labetalol recommended by these authorsfor this purpose is 10 mg given intravenously 1 to 2 minutesprior to anesthesia induction, and adjusted at subsequenttreatments to maintain peak systolic blood pressure below210 mm Hg. If increases up to 20 mg of labetalol proveinsufficient for this purpose, the contents of one 10 mgcapsule of nifedipine can be placed under the patient'stongue 25 minutes before administering the labetalol. Fewerthan 5% of patients develop post-ECT or thostatichypotension on this regimen, which responds to continuedbed rest and intravenous fluids.

Cardiac PacemakersThere are numerous reports of the successful use of ECT inpatients with implanted cardiac pacemakers (Youmans et al.,1969; Gibson et al., 1973;

Abiuso, Dunkelman, and Proper, 1978; Jauhar, Weller, andHirsch, 1979; Alexopolous and Frances, 1980; Alexopolouset al., 1984a; Tchou et al., 1989). Although the ECTstimulus itself is normally prevented from reaching the heartby the high resistance of the intervening body tissues, andpacemakers are constructed with protective electricalcircuitry and shielding to withstand electrical stimuli withinthe range of those used for ECT, the low-resistance pathwayto the myocardium created by an endocardial electrode mayallow a large current to pass through the heart during ECT if

the patient is in contact with ground. Thus, all monitoringequipment must be properly grounded; the stretcher onwhich the patient is lying must be completely insulated fromground (e.g., by rubber wheels); and the patient must notbe held or touched during treatment by anyone who is incontact with ground. (Contact with an improperly groundedmonitor is dangerous even if it is turned off at the time.)Pacemaker wires should be checked for breaks or faultyinsulation because these also provide ready entry of currentsinto the heart. Previous editions of this book have suggestedusing a magnet to convert fixed-mode pacemaker operationto demand mode during ECT to avoid the theoreticaloccurrence of either bradycardia or tachycardia, dependingon the circumstances. Because I am aware of no reportssince 1980 in which the feared episodes actually occurred,or in which the magnet technique was used, I have omittedthis recommendation in the present edition. External(temporary) cardiac pacemakers must, of course, begrounded; they are, therefore, substantially riskier thanimplanted ones because they provide a ready conduitthrough the heart for current flow from improperly groundedmonitoring equipment, on or off, even if unattached to thepatient and only touched by an assistant who is touching thepatient at the same time.

Aortic AneurysmOf the many published reports of patients with aorticaneurysms who received ECT (Monke, 1952; Wolford, 1957;Weatherly and Villein, 1958; Moore, 1960; Chapman, 1961;Pomeranze et al., 1968; Abramczuk and Rose, 1979;Alexopolous and Frances, 1980; Rosenfeld, Glassberg, andSherrid, 1988; Goumeniouk, Fry, and Zis, 1990), either foruntreated aneurysms or after dissection or surgical grafting,no untoward effects occurred in any patient, despite the factthat no particular efforts were made to augment the usualdegree of muscle relaxation or to reduce the blood pressureresponse to treatment. In fact, one patient with multipleaortic homografts (Greenbank, 1958) safely received 10ECTs without any muscle relaxant at all. Although thesecases suggest that aortic aneurysm is not an important riskwith ECT, it seems prudent to recommend thatsuccinylcholine doses in such patients be adequate toprovide full relaxation of the abdominal musculature (e.g.,

P.85

60 to 80 mg), especially in the presence of an untreatedabdominal aneurysm.

SurgeryThere are 6 reported instances of patients who received ECTat varying intervals after cardiac valvular surgery, 2 of themafter replacement of both aortic and mitral valves (Blachlyand Semler, 1967; Weinstein and Fischer, 1967; Hardmanand Morse, 1972; Viparelli et al., 1976). Again, no specialprecautions were taken and no untoward cardiovasculareffects of ECT were observed, even in one patient who wastreated only 27 days after surgery.

Myocardial InfarctionVentricular arrhythmias and cardiac rupture constitute theprimary fatal risks of ECT in the presence of recentmyocardial infarction. Although the risk is greatest duringthe first 10 days postinfarction (Willerson, 1982), andprobably least after 3 months have elapsed (Perrin, 1961),there are no hard data to support the safety (or lack of it)of administering ECT at any given postinfarction interval.Ungerleider (1960) reported the case of a 68-year-oldwoman who inadvertently received an ECT without any illeffect during an acute myocardial infarction that wasdocumented electrocardiographically, and described anotherinstance personally communicated to him of a patientsuccessfully treated with ECT 3 days postinfarction. Despitethese lucky outcomes, the risks of such treatment aresubstantial and may be reduced by waiting as long aspossible after infarction before giving ECT and by thejudicious use of antiarrhythmic and antihypertensive agents,and the administration of 100% oxygen by positive pressurebefore, during, and after the seizure.

Intracardiac ThrombiMensah, Schoen, and Devereux (1990) presented 2 patientswith echocardiographically detected left ventricular apicalmural thrombi who were safely and successfully treated withcourses of 10 and 12 ECT, respectively. Short-acting beta-blockers (otherwise unspecified) were given before each

P.86

treatment to blunt the sympathetic response, butanticoagulation was not given because neither patient had ahistory of embolization.

Myocardial StunningMyocardial stunning is an ECG diagnosis of doubtful validitythat is purported to reflect reversible myocardialdysfunction. I am mentioning it here because of 2 casereports (Zhu et al., 1992; Eitzman, Bach, and Rubenfire,1994) of alleged myocardial stunning during ECT; neitherpatient experienced chest pain, and both went on tosuccessfully complete their course of

ECT without further episodes, one receiving nifedipine duringsubsequent ECTs, the other labetalol.

Management of Central NervousSystem Risks

Brain TumorConsidering the fixed cranium, the noncompressibility ofcerebrospinal fluid and blood, and the substantial increasesin cerebral blood flow and blood-brain barrier permeabilitythat occur during ECT, it is not surprising that majorneurologic deteriorationâ!”and deathâ!”have been reportedwhen this treatment is administered in the presence of aspace-occupying lesion of the brain. The difficulty ininterpreting the older literature on the subject, however,derives from the fact that most of the case reportsdescribing the consequences of such a procedure are subjectto ascertainment bias, having come to the attention ofphysicians precisely because of neurologic deterioration thatoccurred in association with a course of ECT. Thus, Savitskyand Karliner's (1953) patient with occult glioblastoma whodeveloped stupor, papilledema, and hemiparesis after ECT;Shapiro and Goldberg's (1957) 6 patients with previouslyundiagnosed brain tumors who deteriorated rapidly afterECT; Gassel's (1960) three patients with occult meningiomaswho fared likewise; and Paulson's (1967) three cases ofrapid neurologic progression after ECT in the presence ofundiagnosed brain metastases or cerebellar sarcoma, arelargely responsible for the widely quoted but longoutdated

P.87

clinical dictum that â!œbrain tumor is an absolutecontraindication to ECT.â! !

Unique is the extremely negative article of Maltbie et al.(1980), who conducted a retrospective review of 28 casesreported in the literature, plus 7 from their own files. Only34 of the 35 cases exhibited improvement in theirpsychiatric symptoms with ECT, whereas 74 exhibitedneurologic deterioration, providing a risk about twice aslarge as the possible benefit. Ascertainment bias renders thevalidity of their report doubtful. Since the time of thiswriting, numerous instances of successful prospectiveadministration of ECT to patients with known brain tumors(mostly meningiomas) amply demonstrate the safety of thisprocedure when performed cautiously and withforeknowledge (Dressier and Folk, 1975; Hsiao and Evans,1984; Alexopolous et al., 1984; Greenberg, Mofson, andFink, 1988; Fried and Mann, 1988; Goldstein andRichardson, 1988; Malek-Ahmadi and Sedler, 1989; Zwil etal., 1990; McKinney et al., 1998).

The risk in administering ECT in the presence of a braintumor has been attributed to aggravation of increasedintracranial pressure by the cerebral hemodynamic effects ofECT. It is notable that Dressier and Folk's (1975) patienthad normal spinal fluid pressure and dynamics and that allof the other prospectively treated patients describedpreviously had meningiomas that were unlikely to havecaused increased intracranial pressure. An

ECT-induced increase in peri-brain tumor edema mightcontribute significantly to increased intracranial pressure.Although steroids (usually dexamethasone) can effectivelyreduce such edema and reduce the risks of treatment(Carter, 1977; Fried and Mann, 1988; Zwil et al., 1990),these drugs were used in only 2 of the 10 cases described.

In considering whether to give ECT to a patient with a braintumor, the risks are likely to be least in the presence ofsmall, slow-growing (or calcified) lesions and in the absenceof increased intracranial pressure. The administration ofdexamethasone in doses sufficient to reduce peri-braintumor edema seems prudent, although no prospectivecontrolled trials exist to demonstrate the effectiveness ofthis procedure. Oral dexamethasone, 40 mg/ day, for

example, was introduced several days before ECT in apatient who had no evidence of increased intracranialpressure (Zwil et al., 1990).

Prospective administration of ECT in the presence of a braintumor accompanied by increased intracranial pressure hasonly recently been described in the literature, in a depressedpatient with a left temporal anaplastic astrocytoma andbrain edema, both demonstrated on MRI (Patkar et al.,2000). Although papilledema was absent, the attendingneurologist diagnosed increased intracranial pressure on thebasis of the clinical and MRI findings, and the course of ECTwas preceded by one week's treatment with parenteraldexamethasone, 12 mg/day, tapered shortly prior to ECT to2 mg/ day, and continued throughout the treatment course.In addition, oral furosemide and intravenous esmolol wereadministered each day prior to ECT. Substantial clinicalimprovement in the patient's depressive symptoms wasobserved throughout the course of 8 ECTs, withoutaggravation of preexist ing neurological deficits or thedevelopment of new ones.

Other Space-Occupying LesionsCerebral compression by subdural hematoma should increasethe risk with ECT in the same fashion as a brain tumor.Paulson (1967) reported a 47-year-old agitated depressedwoman who became unresponsive immediately after her firstECT and remained comatose for 2 days. A large subduralhematoma was demonstrated on angiography and evacuated,at which point the patient immediately became responsiveand alert. More recently, however, Malek-Ahmadi et al.(1990) successfully used bilateral ECT to relieve a majordepressive illness in an elderly woman with a chronicsubdural hematoma, without any resultant neurologicdeterioration.

Patients with intracranial arachnoid cysts can also presentwith increased intracranial pressure. Escalona, Coffey, andMaus-Feldman (1991) described their treatment of adepressed man with an asymptomatic intracranial arachnoidcyst in whom 11 right unilateral ECTs relieved thedepression without undue cognitive impairment, emergenceof neurologic signs or symptoms, or change in any aspect ofmagnetic resonance images ob tained before and after

P.88

treatment.

Craniotomy, Cerebral Trauma, andSkull DefectYears ago, Savitsky and Karliner (1953) reported thesuccessful use of ECT in a patient whose skull fracture wasfollowed by 5 days of coma. Additional successful reportshave accumulated during the 1990s. Ruedrich, Chu, andMoore (1983) successfully gave ECT to a 21-year-old womanwith a depressive psychosis 3 weeks after she had shotherself in the right parietal lobe in a suicide attempt; thewound required debridement of her right motor and sensorycortex and resulted in left hemiparesis. A total of 17treatments were given with bitemporal treatment electrodeplacement to avoid the parietal skull defect and withconcomitant anticonvulsant therapy to prevent prolongedseizures or status epilepticus. Not only was there noevidence of ECT-induced aggravation of her hemiparesis, butthere was steady improve ment following the course of ECTuntil she had regained nearly full use of her left arm andleg.

The patient of Hsiao and Evans (1984) described earliersubsequently underwent elective craniotomy for removal ofher meningioma, which left a left parietal calvarial defect.Four months postcraniotomy, a second course of ECT wasgiven for a recurrence of her psychotic depression, usingbifrontotemporal treatment electrodes placed well away fromthe cranial defect; no complications resulted from a courseof 9 treatments. Levy and Levy (1987) reported the case ofa 72-year-old man with a plastic plate covering a skulldefect over the entire right cerebral hemisphere resultingfrom the earlier removal of a meningioma and more recentsurgery for a cerebral abscess. A course of nine leftunilateral ECTs induced full remission of his depressionwithout any neurologic sequelae. Ries and Bokan (1979)described a 39-year-old woman who received a course of 12right unilateral ECTs for a depressive psychosis 30 daysafter undergoing the transsphenoidal-transnasal removal of abasophil adenoma of the pituitary. She enjoyed a fullpsychiatric recovery without neurologic complications orsequelae. Roccaforte and Burke (1989) reported a dramatic

P.89

response to a course of 11 ECT (bilateral at first, then rightunilateral) in a 26-year-old, right-handed man whodeveloped psychotic depression 2 months after surgicalremoval of a third ventricle colloid cyst. Notably, the authorscontinued diphenyl -hydantoin at therapeutic serum levelsthroughout the treatment course.

Hartmann and Saldivia (1990) successfully used ECT to treata patient who had sustained large neurosurgical defects ofthe cranium many years earlier secondary to a hand-grenadeexplosion. Following Gordon's (1982) advice, the authorsplaced the ECT treatment electrodes well away (andequidistant) from the 6 Ã— 4.5 cm right temporoparietaldefect to avoid local intracerebral concentration of thestimulus current through the defect. (The bony flapstemporarily raised during most craniotomies do not requireany special precautions in this regard.) Everman et al (1999)also effectively modified their stimulus electrode placementsin 2 patients with neurosurgical skull defects to avoiddelivering the stimulus over the defect.

Cerebral Vascular Malformations

Intracranial AneurysmHusum et al. (1983) successfully administered a course of 10ECTs to a severely melancholic 42-year-old woman who 6months earlier had undergone a craniotomy for surgicalclipping of a bleeding sacculate aneurysm of the upperinternal carotid artery and muscular wrapping of anotherlocated on the right medial cerebral artery. Although thepatient was normotensive, these authors elected to block thesympathoadrenal response to ECT with a combination ofhydralazine (to relax arteriolar smooth muscle) andpropranolol (to prevent hydralazine-induced reflexsympathetic activation and ameliorate the effects of ECT-induced catecholamine release). Under this regimen, nosignificant blood pressure response to ECT was observed,with the maximum increase being 10 mm Hg. As notedpreviously, Drop, Bouckoms, and Welch (1988) successfullytreated a depressed patient with multiple intracranialaneurysms whose dramatic hypertensive response to ECTrequired combined timolol -nitroprusside therapy for its

P.90

control. Notably, this patient experienced no untowardeffects of ECT despite systolic blood pres sure levelsreaching 340 mm Hg.

Bader et al. (1995) safely and effectively gave ECT to twopatients with cerebral aneurysm: the first, with a 1-2 mmaneurysm of the left anterior communicating artery that wasfirst discovered 3 years after a course of ECT (which hadbeen given without any special attempt to control bloodpressure); the second, with a 5 Ã— 4 Ã— 4 mm residualaneurysm of the internal carotid artery (the site of apreviously unsuccessful attempt at balloon oc clusion), whichled her physicians to control blood pressure at the time ofECT with an intravenous esmolol drip, 10 mg/ml.

Viguera et al. (1998) successfully administered ECT to a 54-year-old depressed woman whose 5 mm saccular aneurysmof the basilar artery had been demonstrated 4 years earlierby cerebral angiography. She was given an oral dose of 50mg atenolol daily, and each of the 7 ECTs administered wasimmediately preceded by an intravenous nitroprussideinfusion titrated to a systolic arterial pressure of 90â!”95mm Hg, and monitored with trans-cranial Dopplerultrasonography. During ECT, systolic pressures reachedmaximal values of 110-140 mm Hg, and systolic flowvelocities remained under 76 cm/s. Excessive tachycardiaand hypertension did not occur under the combined betablocker-nitroprusside regimen, and the patient enjoyed a fullrecovery without untoward neurological event.

Hunt and Kaplan (1998) gave ECT to an 83-year-oldpsychotically depressed woman who had a 2 cm aneurysm inthe left middle cranial fossa (adjacent to the left carotidartery), demonstrated on CT scan. Pretreatment with oralbeta-blockers alone allowed her to successfully complete hertreatment course without difficulty.

Najjar and Guttmacher (1998) likewise reported thesuccessful ECT treatment of 2 patients with intracranialaneurysms. Although in both instances pretreatment withintravenous esmolol at doses of 50 to 150 mg

failed to prevent substantial hypertensive responses to ECT,neither patient suffered sequelae therefrom. The patient ofGardner and Kellner (1998) safely received ECT in the

presence of a cerebral aneurysm with nothing more drasticthan a sublingual nitroglycerine spray administeredimmediately before each treatment; a patient of Salaris,Szuba, and Traber (2000) with an internal carotid aneurysmdid equally well with intravenous esmolol.

Although the 10 cases described above abundantlydemonstrate that ECT can safely be given in the presence ofunruptured intracranial aneurysm â!”usually withconcomitant control of the blood pressure surgeâ!”Drop,Viguera, and Welch (2000) nevertheless recommend theroutine administration of the drug regimen described abovein the case of Vigueroa et al. (1988): the combination ofbeta blockade and intravenous nitroprusside.

Many thousands of patients with occult unruptured berryaneurysms must have received ECT without any ill effectbecause such aneurysms are found at autopsy in up to 8%of the population (McCormick and Acosta-Rua, 1965) buthave never been implicated in an ECT-related death.Untreated aneurysms that have previously bled or areactively bleeding undoubtedly constitute a more substantialrisk with ECT, although a direct correlation betweenaneurysm rupture and hypertension has not been established (Drop, Viguera, and Welch, 2000), and there are noreported instances of ECT having been administered undersuch circumstances.

Cerebral AngiomaGreenberg et al. (1986) reported a 24-year-old man with aleft parietal cerebral venous angioma who received a courseof 12 ECTs without special precaution and without incident.Venous angiomas are low-pressure malformations, and themaximum ECT-induced systolic blood pressure levelsrecorded in this patient rarely exceeded 200 mm Hg. An MRIperformed 18 days after the last ECT revealed no changecompared with the pre-ECT record. Salaris, Szuba, andTraber (2000) gave 10 right unilateral ECTs to a 74-year-oldwoman with a left cerebellar venous angioma, with similarlygood clinical results and no neurological sequelae. Eachtreatment was given under intravenous beta-blockade, andfollowed by oral steroids.

P.91

StrokeShapiro and Goldberg (1957) described 6 patients who weregiven ECT for severe depressive states occurring 4 weeks to2 years after a cerebrovascular accident (CVA). Four ofthese patients, including 2 who were treated 4 and 9 weekspoststroke, had remission of their depression withoutneurologic complications. Two patients died duringtreatment: A 55-year-old woman with severe diabetes whoseright-sided CVA preceded ECT by 6 weeks failed to recoverconsciousness after her sixth ECT, and a 48-year-old manwith hypertensive cardiovascular disease whose right-sidedCVA preceded ECT by 2 years suffered cardiorespiratorystandstill during his first ECT.

Murray, Shea, and Conn (1986) reported more favorableresults in 14 patients, age 46 to 86 years, who received ECTfor poststroke depression, a sample selected from therecords of 193 patients with stroke and depression who weretreated at Massachusetts General Hospital from 1969 to1981. All strokes had been completed and none wereevolving at the time ECT was given. Twelve of the 14patients improved markedly with ECT, and none developednew neurologic findings or exhibited worsening of old ones,despite the fact that 4 patients were included whose strokepreceded ECT by less than 1 month. Of great clinicalimportance is the observation that of 6 patients whoexhibited cognitive impairment before ECT, 5 experiencedimprovement in this impairment (and in depression) as aresult of EC T. Kwentus et al. (1984) reported the case of adepressed, catatonic 52-year-old woman with a history of aleft -hemisphere CVA 2 years before who was successfullytreated with a course of 9 right-unilateral ECTs withoutneurologic complications. In fact, a coexisting neuroleptic-induced tardive dystonia also remitted with ECT andremained so at examination 7 months later. The shortestinterval between stroke and ECT was reported byAlexopolous et al. (1984a), who briefly mentioned theuneventful treatment of a patient 4 days after a cerebralinfarction documented by CT scanning.

Allman and Hawton (1987) used a combination of two beta-blockers â!”practolol and atenololâ!”given intravenously toinhibit completely the ECT-induced surge in blood pressure

P.92

in a 60-year-old normotensive woman who had suffered acerebral hemorrhage 3 years previously. Her depressionresponded to 9 ECTs, without associated complications.DeQuardo and Tandon (1988a) reported equally favorableresults in a 48-year-old depressed man who had suffered alarge right parietotemporal strokeâ!”secondary to atrialfibrillationâ!”15 months earlier. Eight right unilateral ECTsinduced a full remission of his depressive syndrome withoutany change in his neurologic or neuropsychological statuswhen he was examined 2 days post-ECT, despite the factthat he continued oral anticoagulation therapy with coumadinthroughout.

The only instance of a specific neurologic complication(Todd's palsy) secondary to ECT in a patient with a historyof stroke was reported by Strain and Bidder (1971), whogave four closely spaced seizures during a single treatmentsession of multiple-monitored ECT (MMECT) to a 62-year-oldwoman who, unknown to the authors, had had a CVA 7years earlier. Status epilepticus lasting 28 minutes occurredafter the fourth seizure, from which the patient awakenedwith a left hemiparesis that partially remitted over thefollowing week. The fact that this patient had received 2previous courses of 6 to 8 conventional ECTs since her CVAwithout neurologic sequelae illustrates the increasedneurotoxicity of MMECT.

Although it usually takes about 3 months for radioimagingevidence of cerebral damage to resolve (Jeffries et al.,1980), the data of Murray, Shea, and Conn (1986) suggestthat ECT given even as soon as 1 month after a CVA doesnot present a major risk to the patient. Indeed, the well-documented increases in cerebral blood flow and oxygenationfollowing

induced seizures described in Chapter 4 have been used astherapy for experimental strokes in dogs (Reed, Ciesel, andOwens, 1971) and in monkeys (Roberts et al., 1972). In thelatter study, bilateral ECTs enhanced by pentylenetetrazolinduced mean intracerebral pO increases of 74% lasting 40minutes or more in monkeys rendered hemiparetic by middlecerebral artery occlusion an hour earlier. Although bothexperimental and no-ECT control monkeys had substantialand equivalent recovery of neurologic function 3 months

poststroke, the authors noted that only 1 session of inducedseizures had been used and that seizures would have had tohave been repeated frequently to maintain elevated pOlevels. This study is cited not so much as a plea forcontrolled trials of ECT in human stroke, which would indeedbe a rational approach, but as a demonstration of theapparent safety of induced seizures in the presence of acuteor recent ischemic stroke.

Anticoagulation TherapyLoo, Cuche, and Benkelfat (1985) reviewed the safety ofadministering ECT to patients on anticoagulant therapy andreported 4 patients who successfully received thiscombination: 3 on heparin, 1 on a warfarin derivative.Tancer, Pedersen, and Evans (1987) found additionalsuccessful cases in their literature review, adding theexample of their own patient who received 2 full courses ofECT while on oral warfarin therapy (prothrombin times, 1.4to 2.1 times control). Subsequently, Tancer and Evans(1989) and Hay (1989) reported the uneventfuladministration of ECT in 4 additional geriatric patients onoral warfarin, bringing to 13 the total number of patientsreported in the literature to have safely received ECT whilereceiving anticoagulants.

Lupus CerebritisGuze (1967) described the successful use of ECT in threepatients with affective episodesâ!”including catatoniaâ!”secondary to lupus cerebritis, but did not mention whetherany adverse effects occurred as a result. Since then, severaladditional patients have been described (Allen and Pitts,1978; Doug las and Schwartz, 1982; Mac and Pardo, 1983;Kurokawa et al., 1989; Fricchione et al., 1990) whose lupus-induced organic affective syndromes with catatoniaresponded fully to ECT without exacerbating the underlyingdisorder. The two most recent reports document thegenerally dramatic responses that have been obtained inpatients unresponsive to high-dose corticosteroids andneuroleptic agents. Kurokawa et al. (1989) described a 30-year-old lupus patient with recurrent manic -depressiveattacks whose 13month history of mutism, incontinence, andclouded consciousness was abruptly terminated by EC T.

P.93

Fricchione et al. (1990) gave ECT to a 25-year-old womanwith a 3-year history of lupus who exhibited facialgrimacing, mutism, rigidity, and an elevated temperature.An intial course of ECT

was aborted because of a high seizure threshold, but asecond courseâ!”given in conjunction withcyclophosphamideâ!”resulted in full remission of herpsychiatric syndrome.

DementiaBecause of the transient, but occasionally pronounced,cognitive side effects of bilateral ECT, psychiatrists havebeen understandably cautious in administering any form ofECT to patients with preexisting cognitive impairment. In astudy of acute organic mental syndrome after bilateral ECT,Summers, Robins, and Reich (1979) reported that the only 2instances of markedly prolonged confusional states (lasting45 and 65 days post-ECT) occurred in the only 2 patientswith preexisting mild chronic dementia. Tsuang, Tidball, andGeller (1979) reported the development of profounddisorientation and urinary and fecal incontinence after 6ECTs (electrode placement unstated) in a 70-year-oldwoman with normal pressure hydrocephalus (NPH) and aright ventriculojugular shunt. (She had been stuporous,disoriented, and occasionally incontinent of urine just beforestarting ECT, but had become fully oriented and continent bythe fourth treatment.) Her cognitive func tioning graduallyimproved over 3 weeks post-ECT, and she was fully oriented at discharge.

Subsequent reports have been more sanguine. Demuth andRand (1980) gave 8 unilateral ECTs to a depressed 80-year-old man with documented severe primary degenerativedementia who achieved a full remission without any increasein confusion. Snow and Wells (1981) treated a woman withprobable Alzheimer's disease who also improved significantlywith ECT without evidence of any worsening of her organicstateâ!”in fact, her incontinence cleared during the courseof treatment. McAllister and Price (1982) described 2depressed patients with dementia (1 with NPH, 1 withCreuzfeldt-Jakob disease) who improved substantially withECT without any exacerbation of their cognitive deficits.

P.94

Unilateral ECT was used for 1 patient, but the other'streatment was not described. Perry (1983) reported the caseof a demented man in his 50s whose muteness and catatoniaresponded dramatically to ECT without any worsening of hisdementia; Dubovsky et al. (1985) successfully treated a 53-year-old depressed man with dementia and increasedintracranial pressure with a course of 9 ECTs (4 unilateraland 5 bilateral) that resulted in marked improvement in hismemory as well as in his depression. Not every depressivepatient improves with ECT, of course, and Young,Alexopoulos, and Shamoian (1985) reported the case of adepressed 73-year-old woman with Parkinson's disease anddementia who exhibited no relief from depression, but alsono persistent worsening of her dementia, after 7 unilateralECTs (there was, however, long-term improve ment in someof her parkinsonian symptoms).

Benbow (1987) described 5 patients in his ECT practice whohad received ECT for depressive illness complicated bydementia: 2 recovered

fully, 1 improved, and 2 did not respondâ!”4 of the 5 weredischarged after ECT to live in their own homes in thecommunity. Liang, Lam, and Aneill (1988) observedsubstantial improvement in affective and vegetative featuresin 2 elderly demented women but neither improvement norworsening of their cognitive dysfunction. In their extensiveliterature review of the results of ECT in 113 cases ofdepression occurring in the presence of organic dementia(mean age, 67 years), Price and McAllister (1989) found an83% overall positive therapeutic response to ECT, with 20%of patients experiencing significant cognitive or memory sideeffects, virtually all transient and reversible, and 15%showing improved cognition or memory consequent to ECT.In general, patients with subcortical dementias (e.g.,Huntington's or Parkinson's disease) showed moreimprovement in depression than those with corticaldementias (e.g., Alzheimer's or Pick's)â!”93% comparedwith 73%, respectivelyâ!”while exhibiting no cognitiveimprovement, compared with 30% cognitive improvement forpatients with cortical dementia.

In the largest study to date, Nelson and Rosenberg (1991)retrospectively analyzed a 4-year sample of 21 elderly

P.95

demented patients who had received primarily rightunilateral ECT for major depression. The method was safeand effective despite transitory increased confusion in somepatients, and the antidepressant efficacy of ECT was similarto that found in a group of nondemented elderly depressivespreviously studied in the same setting.

When one is prescribing ECT for patients with dementia, thegeneral recommendation for starting with a brief-pulsestimulus delivered through right-unilateral treatmentelectrodes is virtually obligatory. Only if improvement failsto occur after 6 ECTsâ!”or the severity of the patient'scondition dictates an earlier changeâ!”should the switch bemade to bilateral electrodes. Twice-weekly treatments arealso less likely to induce cognitive dysfunction than thosegiven 3 times a week, yet they are ultimately no lesseffectiveâ!”they just take longer. The improvement in thedementia syndrome noted after ECT by Price and McAllister(1989) does not suggest any ameliorative effect of thetreatment on the primary syndrome. Rather, the cognitivedeficits of melancholia (depressive pseudodementia) thathave been superimposed on the existing features ofdementia remit with success ful ECT just as they do innondemented patients.

Mental RetardationIn his review and case studies of affective disorders amongmentally retarded patients, Reid (1972) noted in passingthat â!œtricyclic antidepressants and ECT appeared to beeffective in most of the cases where they were prescribed.â! !There are several case reports that specifically address thisuse of EC T. Bates and Smeltzer (1982) reported the case ofa 25-year-old, severely mentally retarded man (full scale IQrange 21 to 25) whose persistent self-injurious, head-banging behavior resulted in a wide-based staggering gait,

loss of manual fine motor control, a Babinski sign, and othersymptoms of upper motor neuron damage. Insomnia,agitation, and weight loss were prominent. A course of 12bilateral ECTs induced a dramatic improvement in all areasof behavior, without any adverse consequences. Guze,Weinman, and Diamond (1987) reported a 21-year-oldbipolar man with spastic diplegia and mild mental

retardation (IQ 65) whose depressive symptoms were rapidlyand fully relieved by a course of 8 unilateral ECTs. Kearns(1987) successfully used ECT to treat nihilistic delusionalpsychosis (Cotard's syndrome) in a severely mentallyretarded man (IQ 35) who had not responded toantidepressant drugs, and Goldstein and Jensvold (1989)reported full recovery without adverse effect (described byhis internist as â!œa miracleâ! !) in a 68-year-old mildlyretarded (IQ 65) man whose relentlessly progressive mentaland physical decline had begun following orchiectomyseveral years earlier. Warren, Holroyd, and Folstein (1989)described an excellent response to ECT in 3 patients withtrisomy 21 (Down's syndrome) who were referred forevaluation of apparent dementia but were instead found tohave major depression. In each instance, ECT wasdramatically effective in restoring the patient to normalfunctioning despite failure to respond to antidepressantdrugs (1 patient returned to school, 1 to work in the familybusiness, and 1 remained well on lithium maintenance over a2-year follow-up interval).

EpilepsyThe anticonvulsant effects of ECT described elsewhere in thisvolume were well known to older clinicians who used them togood effect in the treatment of epileptic patients with andwithout psychiatric symptoms (Kalinowsky and Kennedy,1943; Caplan, 1946; Kalinowsky, Hippius, and Klein, 1982,p. 267). Sackeim et al. (1983a) reported using the seizurethreshold-raising effect of ECT to successfully treat a 19-year-old patient with lifelong, intractable, idiopathicsecondary generalized seizures. Viparelli and Viparelli (1992)successfully used bilateral ECT to treat a 19-year-old womanwith grand mal epilepsy who suffered from as many as 46partial seizures over a 12-hour period, and who wasunresponsive to diphenylhydantoin and diazepam. After thesecond ECT, all seizures ceased and she remained seizure-free over a 7-year follow-up interval, on very low doses ofcarbamazepine (plasma levels of 2.2 g/ml). ECT has alsobeen used effectively to control intractable epileptic seizuresin children (Griesemer et al., 1997).

Thus, ECT is far more likely to ameliorate than aggravate anepileptic disorder, and can be especially useful in patients

P.96

with the psychiatric manifestations of complex partialseizures (usually temporal lobe epilepsy). Epileptic patientsalready receiving anticonvulsants should continue to do soduring the course of ECT because abrupt discontinuationincreases the risk of status epilepticus (Hauser, 1983). Ifone considers the potent anticonvulsant properties of ECT,there seems little rationale for initiating anticonvulsantsbefore ECT in epileptic patients not already receiving them.In any

case, although higher-than-usual electrical doses may berequired, seizures of adequate length and therapeuticpotency can be obtained despite concomitant therapy withanticonvulsants (Weiner, 1981; Sackeim et al., 1986a;Kaufman, Finstead, and Kaufman, 1986; Cantor, 1986),although some manipulation of dose may be required, andan occasional patient may not develop a seizure at all(Roberts and Attah, 1988). Although Kaufman, Finstead, andKaufman (1986) found it easier to obtain seizures in thepresence of blood levels of carbamazepine than ofdiphenylhydantoin, Roberts and Attah (1988) could notobtain a seizure in their patient on carbamazepine despiteseveral double bilateral applications of 4 and 5 seconds of â!œstimu lating currentâ! ! (no further stimulus details wereprovided).

HydrocephalusIn addition to the cases of NPH described previously(Tsuang, Tidball, and Geller, 1979; McAllister and Price,1982), Karliner (1978) and Mansheim (1983) reported thesuccessful administration of ECT to a hydrocephalic patient.Mansheim's (1983) patient is particularly interesting becauseECT not only relieved his depression but also markedlyimproved some long-standing functional deficits despite thepresence of meningomyelocele, a ventriculogastric shunt,and epilepsy. Most recently, Cardno and Simpson (1991)safely and successfully used ECT to relieve depressive illnessin a 54-year-old woman with a ventriculoperitoneal shunt inplace for hydro cephalus secondary to Paget's disease of theskull.

It is unclear whether shunting reduces the likelihood ofsevere cognitive side effects of ECT in patients with NPH.

P.97

Improvement in depression but with marked post-ECTconfusion and memory loss occurred in a patient with NPH(Price and Tucker, 1977) who received ECT before shunting,but Tsuang, Tidball, and Geller (1979) and Levy and Levy(1987) had patients who received ECT after shunting whoalso developed severe disorientation (as well as incontinencein the former instance), both transient. Another NPH patienttreated with a shunt in place (McAllister and Price, 1982)recovered fully from depression without any unusualcognitive side effects, as did the patient of Cardno andSimpson (1991) described above.

Tardive DyskinesiaConsidering the clinical evidence presented below that ECTincreases postsynaptic dopamine receptor responsivity and,in particular, that ECT-emergent dyskinesias regularly occurin Parkinson's disease patients receiving this treatment fordepression (Rasmussen and Abrams, 1991), it might beexpected that ECT would aggravate tardive dyskinesia. It ispuzzling that the opposite is often the case, althoughpublished case reports are equally divided on whether ECTameliorates or aggravates this disorder.

Asnis and Leopold (1978) found that the frequency of oralmovements in 3 of 4 women with neuroleptic-induced,orofacial dyskinesia fell below baseline during ECT andremained so after the course in 2 of the women; the other 2showed increased abnormal movements after EC T. Price andLevin (1978) described a 49-year-old woman whosepronounced buccolingual dyskinesia improved suddenly anddramatically after her third ECT seizure, and Chacko andRoot (1983) reported 2 women, aged 62 and 63 years,whose prominent orofacial and buccolingual dyskinesias alsoimproved markedly after their third and fourth ECT seizures.One patient had no return of tardive dyskinesia (TD) overthe following 2 years; the other showed only an occasionaldyskinetic tongue movement 1 year later. Malek-Ahmadi andWeddige (1988) gave ECT to an elderly depressed womanwhose persistent tardive dyskinesia improved significantlywith remission of her depressive symptoms, and Gosek andWeller (1988) reported even more favorable results: Thedyskinetic movements in their 54-year-old woman improved

P.98

dramatically with ECT and remained so over a 14-monthfollow-up interval. In the only prospective study of ECT fortardive dyskinesia conducted to date, however, Yassa,Hoffman, and Canakis (1990) reported that only one patientof nine showed dramatic improvement.

Tardive dystonia, an entity less well known than tardivedyskinesia, is diagnosed in the presence of chronic torticollisor other dystonia in a patient exposed to neuroleptic drugsand without other known cause or family history of dystonia.Kwentus, Schulz, and Hart (1984) described a 52-year-oldwoman whose severely dystonic gait and posture completelyremitted by her fourth ECT and remained so at a 7-monthfollow-up examination. A similarly favorable, althoughtemporary, response to ECT was reported by Adityanjee etal. (1990), whose 30-year-old male patient enjoyed almostfull remission of pronounced sternocleidomastoid dystonia(anterocollis) for about 1 month after completing a course of11 ECTs for a severe, chronic hallucinatory psychosis.

There are some negative reports, however. Holcomb,Steinberg, and Heninger (1983) reported a 72-year-oldwoman with Parkinson's disease and bucco-lingual dyskinesiawhose formal ratings of tardive dyskinesia worsenedsubstantially during and after a course of ECT despiteremission of her depressive state and improvement in herparkinsonian symptoms. Roth, Mukherjee, and Sackeim(1988) described a manic patient whose preexistingparkinsonian-athetoid complex remitted with ECT, only to bereplaced by an emergent marked orofacial dyskinesia.Liberzon et al. (1992) described 3 depressed patients whodeveloped transient dyskinesias after right unilateral ECT,but responded well to ECT for their affective syndromes andwere free from ECT-induced dyskinetic movements at thetime of discharge.

LeukoencephalopathyThe advent of MRI has brought with it the radiologicdiagnosis of subcortical leukoencephalopathy (Coffey et al.,1988b, 1990a, 1991; Price and McAllister, 1989;

Pande et al., 1990), which is characterized by small areas ofincreased signal intensity in the subcortical white matter.Although this finding is reported in about two thirds of

elderly depressed patients referred for ECT, it neitherpredicts an unfavorable outcome nor is aggravated by ECT(Coffey et al., 1988b; Pande et al., 1990).

Neuromuscular-NeurodegenerativeDisordersAffective symptoms (depression or euphoria) occurfrequently in patients with multiple sclerosis and areoccasionally severe enough to require EC T. Savitsky andKarliner (1951) successfully treated two such patientsdiagnosed as having manic -depressive psychosis, withoutany significant worsening in neurologic status (indeed, onebedridden and incontinent woman regained bladder controland the ability to walk with a cane after treatment). In laterpapers, Savitsky and Karliner (1953) and Karliner (1978)reported several additional cases, although it is unclearwhether any were reported twice. Most patients showedimprovement in psychiatric status with no change or modestimprovement in neurologic symptoms. Savitsky and Karliner(1953), however, cite an additional case from the Germanliterature of a catatonic woman with multiple sclerosis whoimproved after a single, unmodified ECT but developedparaparesis with bilateral pyramidal tract signs and rightupper extremity weakness and hyper-reflexia that took 3months to resolve. In general, however, a pattern ofpsychiatic improvement with at least no neurologicworsening has been consistently observed by subsequentclinicians treating patients with multiple sclerosis (Gallinekand Kalinowsky, 1958; Hollender and Sleekier, 1972;Kwentus et al., 1986; Coffey et al., 1987), with theexception being a 38-year-old manic -depressive man withmultiple sclerosis in remission who developed a gaitdisturbance and dyscalculia requiring cessation of treatmentafter 8 ECTs, with only minor improvements in mental state(Regestein and Reich, 1985). A patient examined with brainMRI by Coffey et al. (1987b) before and after the ECTshowed no change in the pre-existing white matter lesions.

Electroconvulsive therapy has also been used successfully totreat psychiatric patients suffering from cerebral palsy(Lowinger and Huston, 1953; Guze, Weinman, and Diamond,1987), myasthenia gravis (Martin and Flegenheimer, 1971),muscular dystrophy (Zeidenberg et al., 1976), and

P.99

Friedreich's ataxia (Casey, 1991; Singh, Bryce, and Black,2001), all without any complications.

Anesthesia ConsiderationsAlthough many patients with multiple sclerosis have receivedECT with succinylcholine-induced muscle relaxation withoutill effect, Marco and Randels (1979) warned that thepotassium-releasing and muscle-depolarizing action ofsuccinylcholine might adversely affect patients withneuromuscular

disorders. For this reason, Hicks (1987) recommended usingatracurium, the curariform nondepolarizing muscle-blocker,as a muscle-relaxant instead of succinylcholine when givingECT to a patient with multiple sclerosis.

Because of the overlapping symptoms of neurolepticmalignant syndrome and malignant hyperthermia, a familialsyndrome induced by general anesthesia and succinylcholine,concern has also been raised that using this muscle-relaxantwhen giving ECT to patients with neuroleptic malignantsyndrome might reactivate their symptoms (Liskow, 1985).This author also chose atracurium instead of succinylcholinein treating a patient with neuroleptic malignant syndrome.However, Addonizio and Susman (1986) found no instancesof malignant hyperthermia in any of 13 patients withneuroleptic malignant syndrome who received ECT withsuccinylcholine muscle relaxation, and Hermesh et al. (1988)reported no symptoms or past or family history of malignanthyperthermia in 12 patients who had both a history ofneuroleptic malignant syndrome and treatment with a totalof 20 courses of ECT (146 exposures to succinylcholine). Inany case, the protective value of substituting anondepolarizing muscle-blocker for succinylcholine wasquestioned by Grigg (1988), who reported a patient,apparently recovering from neuroleptic malignant syndromewith coma that had occurred 2 days earlier, who againdeveloped fever, tachycardia, and marked elevation in serumcreatine phosphokinase several hours after receiving ECTadminis tered with pancuronium hydrochloride for musclerelaxation.

Other Risk Factors

P.100

PregnancySeveral early reviews of the effects of ECT during pregnancy(Boyd and Brown, 1948; Charatan and Oldham, 1954; Laird,1955; Forssman, 1955; Smith, 1956; Sobel, 1960) failed todemonstrate any increased risk of complications of labor anddelivery, fetal damage, or growth and development thatcould be attributed to the treatment. More recently, theCollaborative Perinatal Project did not find an excess ofmalformations in fetuses exposed to methohexital,succinylcholine, and atropine (Walker and Swartz, 1994).Doppler ultrasonography, external uterine tocodynamometry,ultrasonography, and continuous fetal heart electricmonitoring have revealed no significant alterations of fetalheart rate, fetal movement, or uterine tone during ECT (Wiseet al., 1984; Repke and Berger, 1984); both infants thatwere studied were normal at delivery and at follow-upexamination. In the light of these facts, there seems littlejustification to monitor routinely mother and fetus duringECT (Remick and Maurice, 1978; Wise et al., 1984; Repkeand Berger, 1984); rather, these procedures should bereserved for patients with high-risk pregnancies. The call forperforming ECT in pregnant women only under endotrachealintubation (Wise et al., 1984) seems particularly unwise,because this procedure requires substantially higher doses ofbarbiturate and

muscle-relaxant drugs and stimulates tracheal-laryngealreflexes that can increase pressor responses and theincidence of cardiac arrhythmias (Pitts, 1982).

Because of the release of oxytocin by ECT-induced seizuresâ!”and potential resultant stimulation of uterine contractionsand induction of laborâ!” Walker and Swartz (1994) havesuggested the potential usefulness of tocolytic therapy withritodrine or magnesium in patients who develop persistentuterine contractions accompanied by cervical changes duringor shortly after ECT, or in case vaginal bleeding occurs.

OsteoporosisWith unmodified ECT, the presence of osteoporosisapproximately doubles the incidence of vertebralcompression fracture (Lingley and Robbins, 1947; Dewald,

Margolis, and Weiner, 1954). The modern use of muscle-relaxant drugs, however, has eliminated the risk of any typeof fracture during ECT, along with the influence ofosteoporosis on that risk, so long as the cuffedlimb methodis avoided.

GlaucomaA transient rise in intraocular pressure during ECT isreported to occur in some patients, whereas othersexperience a reduction (Manning and Hollender, 1954;Epstein et al., 1975). Edwards et al. (1990) recentlycarefully defined the time-course of intraocular pressurechanges in 10 nonglaucomatous patients undergoing ECT.Applanation tonometry performed at baseline and every 15seconds poststimulus until return to baseline revealed thatsuccinylcholine significantly increased intraocular pressure,which was then further significantly increased during theinduced seizure by 60%; all values returned to baselinewithin 90 seconds of seizure termination. The authorscharacterized these changes as benign in allnonglaucomatous and many actively glaucomatous patientsand of potential importance only in those with severe orend-stage glaucoma, for whom they recommended ophthalmologic consultation before administration of ECT.

Old AgeIncreased age does not of itself increase the risk with ECT,and, as described in Chapter 2, some of the most rewardingresults with convulsive therapy are obtained in elderly,debilitated patients whose primary affective disordermasquerades as senile dementia.




> Table of Contents > Chapter 6 - The Electroconvulsive Therapy

Stimulus, Seizure Induction, and Seizure Quality

Chapter 6

The Electroconvulsive TherapyStimulus, Seizure Induction,and Seizure Quality

A few basic electrical concepts are required to understandthe properties of the ECT stimulus. Although most ECTdevices utilize alternating current, some of the principlesinvolved are more readily presented through the simplersituation that obtains for direct current. The primaryvariables of direct current electricity are voltage, current,and resistance, which are measured in units of volts,amperes, and ohms, respectively. Voltage is anelectromotive force that drives a current of electronsthrough a conductor just as hydraulic pressure drives waterthrough a pipe. The greater the resistance to the flow ofcurrent, the greater is the voltage required to maintain thesame current. Therefore, the flow of current through aconductor varies directly with the voltage and inversely withthe resistance, a relationship known as Ohm's law andexpressed by the following equation:

Current = voltage/resistance

EnergyEnergy (expressed in joules [J]) is the work required toovercome resistance. It is the time-integral of power, whichis simply the product of voltage and current:

Power = voltage Ã— current

Energy is therefore expressed as the product of voltage,current, and time (the duration of electron flow):

Energy (J) = voltage Ã— current Ã— time

Because voltage = current Ã— resistance (by Ohm's law),the equation for determining the energy of a stimulusbecomes:

Energy = current2 Ã— resistance Ã— time

P.102

ChargeCharge is the total quantity of electrons flowing through aconductor during a given period of time: It is the timeintegral of current. For a constant current, the charge isequal to the product of the current and its duration:

Charge (coulombs) = current (amperes) Ã— time (seconds)

and, therefore:

Energy = voltage Ã— charge

The main difference compared with direct current in theprinciples governing the behavior of the alternating currentsused to generate the ECT stimulus lies in the concept ofimpedance, which is analogous to direct current resistance.Impedance is a measure that combines resistance withcapacitance (the property of being able to accumulate acharge) and inductance (the property of being able to inducean electromotive force, e.g., an electromagnetic field).Measurable inductance has not been reported to occur duringECT, and although neural membranes exhibit the capacity tostore current, the magnitude of this effect is quite smallcompared to the impedance of the electrode-skin interface,which can be substantial (Sackeim et al., 1994). Untilfurther data become available on the nature of inductanceand capacitance during ECT, however, it is safe for practicalpurposes to assume that the impedance during ECT isprimarily attributable to resistance, and therefore subject toOhm's law as expressed above:

Current = voltage/impedance

Seizure InductionThe seizure of ECT is initiated when a quantity (charge) ofelectrons is driven through the brain with sufficient voltageto generate the energy required to depolarize cellmembranes. The seizure threshold is crossed when sufficientneurons have been depolarized to initiate corticalpropagation (spatial summation) of an excitatory processthat ultimately reaches the brain-stem, presumablyactivating a pacemaker area to begin firing rhythmically andrepeatedly, and through its projections, ultimately causingthe rest of the brain to discharge in unison.

With pulsed square-wave stimuli, this process is initiated byrepeatedly delivering tiny doses of current to the scalp,pulse by pulse, until a critical point is passed and ageneralized seizure ensues. The neurophysiologicmechanisms for this process are temporal and spatialsummation.

P.103

Temporal summation is a function of the interaction betweenstimulus frequency and amplitude (Malmivuo and Plonsey,1995); for any 2 consecutive pulses, the voltage required toreduce the membrane threshold is inversely

related to the interpulse interval. Spatial summation,equivalent to neuronal recruitment, proceeds vianeurotransmitter-induced excitatory post synaptic potentials.

Each individual pulse of an effective ECT stimulus deliverssufficient current for at least a portion of it to penetrate thescalp, skull, and brain coveringsâ!”were that not true, noseizure could occur. (When we say, for example, that 90%of a typical ECT stimulus is deflected by the skull, we arespecifically referring to 90% of the charge of each individualpulse. Assuming a stimulus current of 1 A and a pulsewidthof 0.5 ms, each pulse would deliver a charge of 0.5 mC, ofwhich perhaps only 10%, or 0.05 mC, reaches the brain.)

It is clear from the foregoing that, aside from the currentstrengthâ!” which is typically fixed close to 1 Aâ!”thepulsewidth, the rate at which pulses are delivered, and thetotal number of pulses, jointly determine whether and whena seizure is produced. The pulsewidth equals the time thatthe current is â!œon,â! ! or flowing: the wider the pulse, themore current is applied per pulse. The pulse frequencydetermines how many pulses are delivered each second, andthe stimulus train duration defines the total num ber ofpulses and thereby the total dose administered.

The Single-Neuron Action PotentialModelBecause temporal summation is the cumulative excitationproduced at the neuronal membrane by repeatedsubthreshold stimulation, we must be concerned with theneurophysiologic parameters of the single-neuron actionpotential model, which was worked out in virtually itspresent form almost a century ago.

Effect of PulsewidthFor a pulsed current to fire a neuron, it must reduce theresting membrane potential to the threshold level, thusinitiating the sequence of depolarization, reversepolarization, and repolarization (Geddes, 1987). The smallestelectrical dose that will initiate depolarization if administeredfor a long enough duration is designated the rheobase, andthe minimum duration that a dose of twice this strengthmust be applied in order to initiate depolarization is termedthe chronaxie, which is primarily a function of the time-constant of the neural membrane (Geddes, 1987): the timerequired for the membrane to discharge about two thirds of

P.104

its potential. Ranck (1975) estimates the chronaxie ofmammalian neurons to be about 70% of the membranetime-constant.

It is evident that if the stimulus is administered via a pulsedsquare wave, its pulsewidth should not exceed thechronaxie, because the portion of the stimulus thatcontinues after depolarization has already commenced wouldbe wasted, thereby reducing stimulus efficiency. Modernestimates of

the chronaxie of human cortical neurons range up to about0.2 ms (Geddes, 1987; Malmivuo and Plonsey, 1995), whichis consistent with an estimate of 0.15 ms derived from anearly animal study of ECT stimulus parameters (Liberson andWilcox, 1945). This means that pulsewidths for ECTsubstantially larger than 0.2 ms are inefficient to the extentthat they exceed the minimum duration needed to initiateneuronal depolarization at the threshold dose.

Effect of Stimulus FrequencyIn an earlier edition of this volume, I used the termstimulus crowding to describe the potential inefficiency ofsqueezing too many pulses into the same interval, therebystimulating neurons while they are still refractory (Sackeim,1992). My recommendation not to exceed a frequencygreater than about 140 pps to 160 pps (Abrams, 1994a;1997) was based on the approximately 7 ms elapsedbetween initiation of a 1 ms stimulus pulse and the return ofthe neuron to its fully repolarized state. This maximumfrequency range was consistent with an early report fromLiberson and Wilcox (1945) that the voltage required forseizure induction in rabbits using 0.3 ms ultrabrief pulsesfell sharply between about 70 pps and 120 pps and remainedconstant until about 150 pps before increasing again.

However, because the absolute refractory period is onlyabout 1 ms (Kandel, Schwartz, and Jessel, 1991),restimulation during the partially refractory period, althoughless efficient, is nevertheless capable of reducing themembrane potential to a measurable degree, therebycontributing to the process of temporal summation andseizure induction. Hyrman et al.'s (1985) study of 1 mspulses in man found that although the charge required topass the seizure threshold indeed dropped sharply from astimulus frequency of about 7 pps through 100 pps, it thenremained unchangedâ!”or even fell slightlyâ!”through themaximum study frequency of 300 pps, thus demonstratingcontinued stimulus efficiency throughout a much higherfrequency range. Similarly, a standard electronic neuronmodel shows that, for a 0.8 ms pulsewidth, a frequency ofabout 300 pps will indeed cause temporal summation if the

P.105

pulse amplitude is sufficient (Malmivuo and Plonsey, 1995).

Relation of Action-Potential Model toElectroconvulsive TherapyHow do these observations apply to constant currentstimulation of the brain via the scalp electrodes used forECT? Because the applied current is constant, the pulsepenetrates the skull with its parameters intactâ!”that is,assuming, as above, that only 10% of the applied currentpenetrates the skull, a single 1 amp square pulse of 0.5 msduration applied to the scalp should be recorded at thesurface of the cortex as a 0.1 amp square pulse of 0.5 msduration.

Because these parameters are in the range of those requiredto initiate depolarization of a single neuron (Malmivuo andPlonsey, 1995, p. 367), one or more neurons in closestproximity to the stimulating electrodes are depolarized byeach pulse, while others further away only have theirthresholds lowered. The cumulative lowering of individualneuronal membrane potentials by repeated subthresholdstimuli until depolarization occurs (temporal summation)contributes to, and is paralleled by, interneuronal spatialsummation (recruitment), thus generating seizure foci thatcoalesce and spread across the cortical surface and thenceto the subcortex, eventually activating a subcorticalgenerator to drive a generalized seizure.

Stimulus WaveformThe electrical stimulus can be delivered in an infinite varietyof forms, of which the two most common are the sine waveand the pulsed square wave. Sine-wave currents arecharacterized by a continuously changing stream ofelectrons, flowing alternately in opposite directions, at afrequency of 50 to 60 wave pairs (1 negative, 1 positive)per second, for which units of Hertz (Hz) are employed. Thisis the current waveform that is universally supplied by walloutlets and was the first type to be used for ECT (Cerlettiand Bini, 1938). The alternating slow rise-and-fall times ofthe sine wave current, and especially its great phase-width(8.3 ms), deliver substantial amounts of electricalstimulation below seizure threshold. Such inefficient below-threshold stimulation contributes little to seizure intensity orgeneralization and, there fore, to the therapeutic effect, butadversely affects memory functions (Ottosson, 1960).

The pulsed square wave, with its fast rise-time, wasrecognized early to be a more efficient and physiologicalstimulus for inducing seizures (Merritt and Putnam, 1938). Itrises and falls almost vertically, delivering all of its charge

P.106

in about 1/1000 of a second or less. The current is offduring most of the time that the stimulus is administered(e.g., current flows for only 0.14 second during a 1-secondbrief pulse stimulus of 140 pulses of 0.001 second each).The pulsed square wave therefore induces seizures withsubstantially less charge and energy than do sine-waves,thereby achieving the same therapeutic effects withsignificantly less memory loss and EEG abnormality (Weineret al., 1986a,b).

Attempts to reduce the neurotoxicity of the sine-wavestimulus by chopping or clipping a portion of it (e.g.,Siemens Konvulsator, Ectron Duopulse) were only partiallyeffective (McClelland and McAllister, 1988). Because sine-wave stimuli exhibit no therapeutic advantage over brief-pulse stimuli (Weiner et al., 1986a,b), they have long beenconsidered obsolete for ECT (Weaver and Williams, 1982;Royal College of Psychiatrists, 1989, 1995). In 1982 theBritish government ordered all sine-wave devices in itsNational Health Service hospitals replaced with brief-pulseinstruments (Department of Health and Social Security,1982). The Ontario Medical Association and The OntarioPsychiatric Association (1985),

Danish Psychiatric Association (Bolwig, 1987), and AmericanPsychiatric Association (1990, 2001) have made similarrecommendations.

Although a double-blind, controlled prospective comparisonof sine-wave and brief-pulse stimuli for bitemporal ECT inpatients with endogenous depression (Andrade et al, 1988a)reported that 93% of the sine-wave group compared with60% of the brief-pulse group exhibited depression ratingscale reductions of 75% or better after an average course ofabout 6 ECTs, mean seizure durations were borderlineâ!”27seconds in each group â!”and patients in the sine-wavegroup received more than a 3-fold higher mean stimuluscharge: 317 mC compared with only 97 mC for the brief-pulse patients. Thus, the brief-pulse group receivedsuboptimal treatment (Weiner and Coffey, 1989). The 97 mCmean dose used for brief pulse ECT was, in fact,substantially lower than the mean dose that Sackeim et al.(1987a) found to be without therapeutic effect for unilateralECT.

In an open retrospective comparison of the records of 197patients who received sine-wave ECT with those of 144 whoreceived brief-pulse ECT, Fox, Rosen, and Campbell (1989)found that the mean number of ECTs received by each groupwas about 7.5 and that both groups were equally improvedat time of discharge, according to the treating physician'sglobal impression.

P.107

Pulse Stimulus ParametersThe pulsed, square-wave stimulus can be described in termsof frequency, pulse-width, and number of pulses. Thestandard stimulus delivered by modern ECT devices is a trainof bidirectional square waves, with each â!œcycleâ! !consisting of one negative and one positive pulse.Manufacturers of brief pulse devices have, somewhatconfusingly, followed the tradition for describing sine-wavefrequency in Hz, so for the bidrectional pulsed stimulus, afrequency of 70 Hz actually delivers 140 pulses per second(pps). To avoid this confusion, the remainder of the text willuse only the pps unit.

For a pulse width of 1/1000 of a second (1 millisecond, ms),for example, one second of stimulation will deliver 140pulses of 1 ms each, or a total of 140 ms (0.14 seconds) ofstimulation. The charge of this stimulus delivered by aninstrument with a constant current of 0.9 amps is calculatedas follows:

(Ampere-seconds are known as coulombs, C, andmilliampere-seconds as millicoulombs, mC.)

The energy (in joules, J) of the same stimulus can becalculated only if the impedance is known or assumed, e.g.,220 ohms. The equation is:

Use of joules to describe a stimulus can be misleading.Should the patient's impedance double for the nexttreatment because the skin was not cleaned properly, thetotal stimulus energy would also double, suggesting to theunsophisticated operator that a greater stimulus had beendelivered to the brain. Almost twice as much variance in theseizure threshold can be accounted for in units of charge asunits of joules (Sackeim et al., 1987b, 1994; Coffey et al.,1995a). Moreover, units of joules were insensitive indetecting a rather strong sex difference in the seizurethreshold that was demonstrated with the unit of charge(men had a higher threshold than women), and the authorsconcluded, with others (Gordon, 1982; Gangadhar et al.,1985), that for brief-pulse, constant-current ECT, the unitof charge was superior to joules as a measure of electricaldose.

P.108

Ultrabrief StimuliThe term ultrabrief was originally introduced to describepulsed square wave stimuli of less than 1.0 ms duration(Weaver and Williams, 1982), but is now commonly used forpulsewidths of less than 0.5 ms. Earlier studies (Cronholmand Ottosson, 1963b; Robin and De Tissera, 1982) reportingreduced efficacy for the ultrabrief pulsed stimulus comparedwith other stimulus wave-forms are uninformative becausethey confounded pulsewidth and wave-form, did not employsquare waves, used irregular frequencies, failed to controlfor dosage, and only used bitemporal treatment electrodeplace ment (Hyrman, 1999; Sackeim, 1999b; Sackeim et al.,2001b).

Hyrman et al. (1985) were the first to systematically applyultrabrief pulsed square-wave stimuli in the modern era,employing 0.04-0.06 ms micropulses at frequencies of 100to 300 pps to successfully induce seizures in animals. As inthe earlier ultrabrief pulse animal studies of Liberson(1945), less energy was needed to produce a seizure athigher stimulus frequencies. Hyrman et al. (1985) found theoptimal frequency to be in the range of 125 pps to 300 pps,consistent with the 120 pps to 150 pps recommended byLiberson (1945). No report ever appeared of the use of suchmicropulses in man, however; instead, the originators of thismethod chose stimuli an order of magnitude larger for theirhuman trials (Pisvejc et al., 1998), well within the range ofpresent-day ECT devices. Using experimental equipment oftheir own design, these investigators administered unilateralECT with 0.2 ms-0.4 ms pulses at frequencies in the 200pps-300 pps range to patients with schizophrenia, andreported similar clinical and

cognitive results to those obtained with a commerciallyavailable device that delivered 1 ms pulses at a rate of 140pps.

Sackeim et al. (2001b) reported interim findings of a studyin progress comparing the clinical effects in patients withmajor depression of 0.3 ms ultrabrief pulses with 1.5 mspulses, both administered at two different dosage levels andtreatment electrode placements: unilateral ECT given at 6Ã—threshold, and bitemporal ECT at 2.5Ã— threshold. The studyemployed frequencies in the 40-80 pps range, and pulsetrain durations up to 8 s. Seizure thresholds, ictal EEGmeasurements, and cognitive side effects were all muchsmaller with the 0.3 ms than the 1.5 ms pulsewidth.Therapeutic effects varied with treatment electrodeplacement: response rates were in the 60% to 75% rangefor ultrabrief pulse unilateral ECT, and 1.5 ms pulsewidthunilateral and bitemporal ECT, but ultrabrief pulsebitemporal ECT yielded an unexpectedly low response rate of

P.109

only 20%.

This puzzling result may be due to the low stimulusfrequencies em ployed (40 pps-80 pps), presumably tomaximize stimulus duration and avoid stimulus crowding.

I expect that the weak efficacy of low-frequency ultrabriefpulse bitemporal ECT reported by Sackeim et al. (2001b) canbe augmented by employing higher stimulus frequencies,perhaps in the 150 pps-300 pps range. Of course, thereported increased efficiency of longer -duration stimulustrains may be lost by increasing stimulus frequency (due tothe inverse relation between stimulus frequency andstimulus-train duration when dose is held constant).However, it has never been determined that this increasedefficiency is due to the longer interval over which thestimulus is applied, rather than to the greater total numberof pulses thereby delivered. At this juncture, it seems morelikely that the total number of pulses is the more importantvariable, and that the clinical efficacy of ultrabrief pulsebitemporal ECT will be augmented by delivering it at higherpulse frequencies. (Two recent studies have not shown thatincreasing the frequency makes a stimulus more effective.Swartz and Manly (2000) found the same failure rate for a120 pps, 1 ms stimulus as for a 60 pps, 1 ms stimulus, andDevanand et al. (1998) did not find stimulus titration in thefrequency domain to be a particularly effective method forseizure induction. However, neither study employedultrabrief pulses or stimulus frequencies in the 150 pps-300pps range.)

Constant Current, Constant Voltage,or Constant Energy?An ECT stimulus can have a constant voltage or a constantcurrent, but Ohm's law enjoins it from having both. Constantcurrent stimulation is the more physiological method of thetwo for inducing neuronal depolarization and is more likelyto induce a seizure in the presence of a high impedanceâ!”for example, in the elderlyâ!”because of insufficientcurrent delivery with

constant voltage or constant energy devices (Weiner, 1980a;Weiner and Coffey, 1986, 1988; Sackeim et al., 1994).

Railton et al. (1987)â!”as corrected by Railton (1987) andMcClelland and McAllister (1988)â!”found that much moreenergy was delivered by constant-voltage than by constant-current devices, which were more consistent in stimulusdelivery. A constant current ensures stable delivery of thestimulus over a wide range of impedances, in contrast toconstant voltage or energy, which more readily induce briefor missed seizures when administered close to the patient's

P.110

threshold (McClelland and McAllister, 1988; Sackeim et al.,1994). In addition to the reports, cited above, the BritishDepartment of Health and Social Security (1982) and theAmerican Psychiatric Association (2001) also recommendedconstant-current instruments for ECT.

Impedance MeasurementsThe charge passing through the brain is related to theimpedance of the head in a complex fashion. Most of theimpedance is across the skull, estimated at 18,000ohms/cm, compared with about 200 ohms/cm across the skinor brain (Weaver, Williams, and Rush, 1976). Although thecharge with a constant current device does not vary withimpedance, its distribution among the 3 compartments ofscalp, skull, and brain does vary with the voltage. At lowvoltages there is insufficient electromotive force to driveenough current through the high-impedance skull to inducea seizure; most of it is shunted (short-circuited) betweenthe electrodes via the low-impedance scalp. As voltageincreases, more and more current penetrates the skull toenter the brain, increasing the likelihood of depolarizingenough neurons to exceed the threshold for a seizure.

An inverse relation for constant-current devices betweenseizure threshold (the charge required to induce a seizure ofspecified duration) and dynamic impedance was documentedby Sackeim et al. (1987a), and replicated by Coffey et al.(1995a). It results in the counterintuitive observation thatthe high-threshold patients in whom seizures are the mostdifficult to elicit are actually those with the lowestimpedances. This is probably due to greater shunting of thestimulating current through extracranial tissues, resulting ina lower dynamic impedance and less current entering thebrain (Sackeim et al., 1994). The finding of McCall et al.(1993a) that higher thresholds were associated with larger(and presumably, thicker) skulls is consistent with this view.

Brief-pulse devices deliver a constant current, so the voltagevaries directly with the dynamic impedance of the patient.Because extremely high impedances would drawcorrespondingly high voltages to maintain the same currentacross the electrodes, thus markedly increasing the energygenerated, brief-pulse devices also limit the maximumvoltage that can be applied to

about 500 volts (the point is moot, however, because inclinical practice a patient with 500 ohms' dynamic impedanceis virtually never encountered).

Impedance to the electrical stimulus during ECT is primarilyattributable to the patient, although corrosion may causesubstantial impedances to develop in the stimulus leads

delivering the current, and their connectors. During a giventreatment, the high impedance of the skull relative to theskin and subcutaneous tissues causes most of the stimuluscurrent to be shunted through the scalp (Weaver, Williams,and Rush, 1976; Gordon, 1981); the closer the treatmentelectrodes are placed to each other (e.g., as for bifrontal orunilateral ECT), the greater this shunt will be. The chargeentering the brain is then distributed along the paths ofleast impedance. With bitemporal ECT, current densities aregreatest in the frontal poles, diminishing in more remoteareas in proportion to the square root of the distancetraversed; with unilateral ECT, current density is greatest inthe pathway between the electrodes, across the surface ofthe brain (Weaver, Williams, and Rush, 1976).

Measurement of the patient's skin (static) impedance beforeadministering the electrical stimulus for ECT providesimportant information on the quality of the skin-to -electrodecontact: If the skin is oily, or if the electrodes are appliedloosely or with inadequate conductive gel, a high impedancewill be registered, informing the physician that his techniquerequires improvement. Such impedance testing is performedwith a high frequency, very low milliamperage current thatis undetectable by the patient. The static impedance is muchhigher than the dynamic impedance that is recorded duringthe actual passage of the treatment stimulus; the dynamicimpedance is function of the summed electrical properties ofthe skin, hair, scalp, subcutaneous tissues, periosteum,bone, dura and pia mater, brain, blood vessels, blood, andcerebrospinal fluid (Weaver, Williams, and Rush, 1976), andfalls dramatically during the passage of the treatmentstimulus (Maxwell, 1968).

Umlauf, Gunter, and Tunnicliffe (1951) measured the staticand dynamic impedances of the human head during ECTwhile systematically varying voltage. They found nocorrelation between static and dynamic impedance butobserved the latter to vary inversely with the voltageapplied (200 to 300 ohms over a range of 60 to 160 volts,with a mean impedance of about 200 ohms above 160 volts),a result later confirmed by other investigators (Maxwell,1968; Gordon, 1981; Gangadhar et al., 1985; Railton et al.,1987; Sackeim et al., 1987b), although often with greatvariability.

The Seizure ThresholdMany studies have now systematically employed an iterativestimulus titration procedure to determine, throughincrementally repeated subconvulsive stimulations, thesmallest electrical dosage for brief-pulse ECT to evoke agrand mal seizure of specified minimum durationâ!”generally25-30 sec by EEC criteria (Weaver, Williams, and Rush,

P.111

1976; Weiner, 1980a; Sackeim et al., 1987b,

1993; 2000; Letemendia et al., 1993; McCall et al., 1993a,1995, 2000; Rasmussen, Zorumski, and Jarvis, 1994; Bealeet al., 1994; Enns and Karvelas, 1995; Coffey et al., 1995a;Shapira et al., 1996; Isenberg et al., 1996; Bailine et al.,2000; Delva et al., 2000).

Across grouped patient samples, there is a great variabilityin the mean threshold values obtained for unilateral ECT, forexampleâ!”ranging from 13 mC (Beale et al., 1994) to 113mC (Sackeim et al., 1987a)â!”which reflects differences inpeak current, age, sex, treatment electrode placement,seizure duration criteria and measurement method, electricalstimulus parameters, and the strength of the initial andincremental dosages of the titration schedule. For example,Swartz (200 Ib) has pointed out the substantial reduction inseizure threshold variability that results from using a fixed,1 ms pulsewidth, rather than varying the pulsewidth as isdone in most titration studies. Even when using the identicalECT device, stimulus parameters, and titration procedure,the variability in threshold measurement between centerscan be great, as evidenced by the 60% range in thestandard deviation of the seizure thesholds obtained at 2different study sites reporting the same mean thresholdvalues for the same-sized samples (Boylan et al., 2000).

However, the resultant up to 4000% or 5000% individualvariability reported (McCall et al., 1993a; Sackeim,Devanand, and Prudic, 1991; Boylan et al., 2000) has beeninflated by inclusion of a few outliers; the data collected byBeale et al. (1994) revealed the distribution of thresholdvalues to be tightly clustered in the 10%-40% range(Kellner, 2001). Indeed, in a sample of the dozen mostrecent titration studies that I located in my files, more thantwo thirds of all reported seizure threshold values forunilateral or for bitemporal ECT fell within less than a 3-foldaverage range, repre senting a variability of about 200%.

Estimates of the seizure threshold obtained by varying oneof the parameters determining stimulus dosage (e.g., thestimulus frequency) are not inherently more valid thanestimates obtained by varying another parameter (e.g.,stimulus train duration). Thus, Swartz and Larson (1989)found that a 144 mC brief-pulse stimulus charge forbitemporal ECT was more effective in producing seizures ofat least 20 seconds' duration when administered with a 2-second stimulus train than with a 1-second stimulus train,voltage and energy remaining constant. Their study nicelydemonstrates the depen dence of seizure thresholdestimates on the stimulus parameters selected, focusing onstimulus-train length.

P.112

Swartz and Larson's (1989) finding has been confirmed andextended in 3 clinical studies using stimulus titration.Rasmussen, Zorumski, and Jarvis (1994) performed stimulustitration for right unilateral ECT using 3 iterative dosagelevels of 25.2 mC, 50.4 mC, and 75.6 mC, delivered via a0.5-ms pulse width, 60 pps frequency, and stimulus-traindurations of 0.93 second, 1.9 seconds, and 2.8 seconds,respectively (mean 1.8 seconds). However, no minimumseizure duration criterion was specified. The mean seizurethreshold for their sample was 48.9 mC, significantly lowerthan the mean threshold of 73.5 mC reported by Sackeim etal. (1993), using a 1-second

stimulus train duration. Even more to the point, the meanthreshold of Rasmussen Zorumski, and Jarvis (1994) wassignificantly lower than the 73.5-mC threshold found byMcCall et al. (1993a), who used the identical dosage andincrements for the first 3 stimuli as Rasmussen, Zorumski,and Jarvis (1994), but administered via a 60 pps, 1-msstimulus, and therefore with stimulus trains that were onlyhalf as long. (It is, of course, unclear whether halving thepulse width or doubling the stimulus train duration is thekey to the observed differences between the 2 studies; inclinical practice the two cannot be separated: Halving thepulse width at a given charge setting doubles the stimulusduration.)

The most recent study of the problem was conducted byDevanand et al. (1998), who used a custom-modified ECTdevice to compare the efficiency of altering pulse frequency(40 pps-280 pps) or stimulus train duration (0.5 sec-4 sec)in determining the seizure threshold to bitemporal ECT in 12patients who served as their own controls. At the firsttreatment, seizure threshold was determined by varyingeither pulse frequency or stimulus duration; the procedurewas reversed at the second treatment. All stimuli wereadministered at 0.8 A using a pulsewidth of 1 ms, and eachstep in the titration procedure used the same charge forfrequency-titration as for duration-titration. Duringfrequency titration the duration of the stimulus train wasmaintained at a fixed value of 0.75 s; during durationtitration, the frequency was kept constant at 60 pps. Acrossthe 12 patients, the seizure threshold averaged 21% lowerwith duration titration than frequency titration (90 mC vs.114 mC), thus providing the most specific confirmation todate of Swartz and Larson (1989).

The study of Swartz and Manly (2000) sheds light on therelative importance of stimulus pulsewidth and stimulusfrequency for seizure induction. In this study, 24 patientsreceiving left -anterior, right-temporal ECT (Swartz, 1994a)at a stimulus charge about 2.5 times age were randomly

P.113

assigned to a sequence of 4 treatments with stimuli ofpulsewidth either 0.5 ms or 1.0 ms, and frequency of either60 pps or 120 pps, in a balanced, repeated measures design:all subjects received all stimuli. Of 7 subjects with at least 1failed motor seizure, 71% failed with both 1 ms stimuli, andnone failed with both 0.5 ms stimuli (p < 0.005).Differences between 60 pps and 120 pps stimulusfrequencies were negligible, demonstrating the greaterefficiency of the 0.5 ms pulsewidth than the 1 mspulsewidth and the relative inefficency of varying frequencyin the 60 pps-120 pps range.

Isenberg et al. (1996) included by far the largest sample ofpatients treated (n = 403) and used the identical titrationdosages and increments as McCall et al. (1993a) andRasmussen, Zorumski, and Jarvis (1994), for patientsreceiving either unilateral or bitemporal ECT given under 2different conditions: long versus short stimulus trainduration. The long stimulus train was administered via a0.5-ms stimulus at the lowest possible frequency (60 pps),and the short stimulus train via 1-to 2-ms pulse widths atvarying frequencies. With the 0.5 ms pulsewidth, longstimulus train, 80% of patients

receiving unilateral ECT had seizures at 50 mC or less,compared with only 37% of patients receiving 1 to 2 mspulsewidth, short-duration stimuli. For bitemporal ECT,100% of patients receiving short pulsewidth, long-durationstimuli exhibited seizures at 100 mC or less, compared withonly 29% of those receiving long pulsewidth, short-durationstimuli. Seizure duration was not different among the 4groups; age and seizure threshold were positively correlated(r = 0.32).

This study provided the additional opportunity to comparethe seizure-inducing efficacy at matched charge dosages of2 different brief pulse ECT devices during stimulus titrationfor unilateral ECT: one with a maximum stimulus duration of2 sec and a minimum pulsewidth of 1 ms, the other with amaximum stimulus duration of 8 seconds and a minimumpulsewidth of 0.5 ms. At increasing stimulus titrationdosages across the study range of approximately 25 mC to100 mC, the greater efficiency of the shorter pulse-width,longer stimulus train device became apparent, approaching100% successful seizure induction at the 75 mC to 100 mCcharge settings, com pared with only 60% to 70% for thelonger pulsewidth, shorter stimulus train device.

Confirmation of these results was provided for the same 2devices by Chanpattana et al. (2000), who reported thattitrated seizure thresholds to bitemporal ECT weresignificantly lower with the shorter pulsewidth, longerstimulus train device in 79% of patients studied, averaging

P.114

61% lower overall despite careful matching of stimuli andpatient titration increments.

At the present state of knowledge, therefore, the seizurethreshold cannot be viewed as a fixed or absolute quantity,but rather as infinitely variable across an endless range ofstimulus combinationsâ!”in fact, more of a metaphor than ameasure. The best that can be said is that under the rightcircumstances, every human being is capable of experiencinga grand mal seizure, and that it is the study of thosecircumstances that informs our knowledge of themechanisms through which ECT exerts its diverse influences.

Moreover, the recommendation to use the seizure thresholdas a guide to ECT dosing (American Psychiatric Association,2001) is problematic, because although a seizure is inducedduring ECT, the literature provides no evidence for anassociation either between the duration or the threshold ofthat seizure and clinical improvementâ!”on the contrary,many studies have looked for such associations and foundnone. Nor is there experimental data to support the 25-30 sminimum EEG seizure duration universally recommended forthis purpose. The original choice of this particular value wasarbitraryâ!”it might just as well have been 10 seconds or 60secondsâ!” but whatever duration might have been chosen,the problem is that seizure duration does not correlatesignificantly with clinical response (Lalla and Milroy, 1996)

The facts that the numerical value for the seizure thresholdis always a function of the minimum seizure duration chosenfor the titration procedure, and that the physicalcharacteristics of the stimulus used to elicit the seizure

always determine the particular threshold value obtained,are not mere quibblesâ!”they raise practical clinicalquestions that are not answerable at today's level ofknowledge (Swartz, 2001b).

For example, a particular patient in one clinical settingwhere shortpulsewidth, long-duration stimuli are employedmight have a seizure threshold of say, 50 mC, but doublethat threshold in another setting where longpulsewidth, shortduration stimuli are used. The advice (Sackeim, 1994a) tosimply administer the same multiple of the threshold dose ineach instanceâ!”say, 250 mC in the first, and 500 mC in thesecond, as with 5 times threshold dosingâ!”is not based onany evidence demonstrating that those 2 very differentdosage levels would have the same therapeutic impact inthat particular patient, or the same side-effects on memory.

Were the extent of dosage above threshold indeed the maindeterminant of the clinical antidepressant response to ECT,as the theory states, then a dose 2.5 times threshold should

P.115

yield a greater antidepressant effect than a dose only 1.5times threshold. Yet a recent comprehensive study (Sackeimet al., 2000) found the same 30%-35% final response ratesfor both those dosage levels. If increasing the stimulus doseby one entire level above thresholdâ!”or about 60 mCâ!”doesn't increase the response rate, then the theoryrequires re-examination.

(A similar argument applies to the claim, assessed in moredetail in Chapter 10, that, like the clinical response to ECT,the cognitive effects are also determined by the degree towhich dosage exceeds threshold. In one study (Sackeim etal., 1993), a greater adverse cognitive effect of higher,relative to lower, doses was not found; in another study(Sackeim et al., 2000), time to reorientation after ECT, andretrograde amnesia for word recall and recognition, wereboth affected more with 1.5 Ã— threshold dosing than 2.5 Ã— threshold dosing, a result that is, in fact, diametricallyopposite to the theory.)

The seizure threshold typically increases across a course ofECT, but neither the absolute value of the threshold, nor themagnitude of its increase, are related to clinical outcome.

Initial reports of an increase in seizure threshold over atreatment course contributed to the development of ananticonvulsant theory of the mechanism of action of ECT(Sackeim, 1994a), which found indirect support in the factthat patients whose seizure thresholds did not increasesubstantially across a course of treatment did not have agood clinical outcome (Sackeim et al., 1987c; 1993;Sackeim, 1999).

The assertion (Boylan et al., 2000) that: â!œâ!"titrationremains the only accurate method to determine electricaldosage in right unilateral ECTâ! ! is notable not only becausethe concept of accuracy in ECT dosage has never been raisedbefore, but because of its circularity. The authors could notbe referring to accuracy in determining the antidepressantpotency of a given dose because equal or betterantidepressant results are routinely obtained using a fixedhigh dose or age-based dosing. Since there is no otherobvious meaning of accuracy with regard to ECT dosage theauthors must

be referring to measurement of the extent to which dosageexceeds threshold, which translates into the circular: â!œtitration remains the only method to determine thedegree to which dosage exceeds thresholdâ! ! (Equally self-evident is the insight offered by the Task Force on ECT ofthe American Psychiatric Association (2001) that â!œâ!"titration provides the most pre cise method for quantifyingseizure threshold.â! ! It is, in fact, the only method.)

P.116

At best, the titration method is capable of revealing thelowest dose for inducing a minimal or weak seizure (Swartz,2001b), which may be somewhat useful information forconventional bitemporal ECT, but is only marginally usefulfor unilateral ECT. Swartz (2001b) has cogently describedmore than a dozen shortcomings of the stimulus titrationmethod, and concluded that the scope and depth of itsinternal inconsistencies render it unsuitable for clinicalpurposes. I suggest that it is time to move on to theintensive study of alternate methods for determining ECTdosingâ!”partic-ularly for unilateral ECTâ!”with the aim ofreplacing seizure duration and seizure threshold withmeasures that vary with clinical efficacy (Abrams, 2001;Swartz, 2001b).

Stimulus Dose and Parameters inRelation to Efficacy

Stimulus DoseAvailable data are not sufficient to allow a definitiveassessment of the separate contributions of total dosage andstimulus parameters to therapeutic impact. Although dosageis arguably the variable most potently affecting clinicalefficacy, the efficiencyâ!”and thereby, presumably, theefficacyâ!”of every stimulus dose is affected by thecharacteristics of the stimulus parameters with which it isadministered. Moreover, because of an important interactionbetween stimulus dose and treatment electrode placementâ!”higher doses are required for unilateral than bitemporalECT in order to achieve equivalent clinical effectsâ!”it isuseful to examine dosage effects separately for eachmethod. Finally, dosing method must also be consideredâ!”fixed-dose, age-based, or titration-basedâ!”as each has aparticular relation to the dose ultimately admininstered.

Unilateral Electroconvulsive TherapyTable 6-1 includes all studies I could find of brief-pulseunilateral ECT that provided the mean dosages used andeither the response rate or the percent improvementobtained. In practice, response rates (calculated as theproportion of patients meeting specific improvement criteria)and percent improvement (calculated as the percentagereduction in depression scale score) are highly correlatedand often numerically close as well. For example, in our

comparison of high-dose unilateral and bitemporal ECT(Abrams, Swartz, and Vedak, 1991) the response rates were65% and 78%, respectively, com pared with depressionscale reductions of 68% and 79%.

Table 6-1 Brief-pulse unilateral ECTresponseby stimulus dose

Author

Mean

dose

Improvement

(%)

Response

rate (%)

Mean

#

ECTs

Dose

method

Sackeim etal. (1993)

86mC

17 9 1 Ã—

Letemendiaet al.(1993)

107mC

50 12 1 Ã—

Sackeim etal. (1987a)

113mC

38* 28 9 1 Ã—

McCall etal. (2000)

136mC

39 6 2.25 Ã—


139mC

35 10 1.5 Ã—

McCall etal. (1995)

151mC

65 8 2.25 Ã—


175mC

43 9 2.5 Ã—

Ng et al.(2000)

188mC

40 6 2.5 Ã—


195mC

45 9 2.5 Ã—

Abrams et

al. (1991)378mC

68 65 6 Fixed

Abrams,Swartz,and Vedak(1989)

378mC

70 6 Fixed

P.117

McCall etal. (1995)

403mC

69 6 Fixed

McCall etal. (2000)

403mC

67 6 Fixed


441mC

80 8 6 Ã—

Pettinati etal. (1990)

476mC

89 6 Age

Fixed = fixed dose; Age = age-based dose; (n ) Ã— = titration-based dose at (n) times threshold.* Calculated from published figure.

To facilitate comparison among the studies I have arrangedthem in order of increasing mean stimulus dose, whichnecessitated splitting up studies that examined more thanone level or type of dosing and placing them in differentparts of the table as if they were different studies. Wheremore than one post-ECT evaluation interval was reported, 1chose the one obtained immediately or 1â!”2 days after thetreatment course; in some studies, how ever, improvementwas determined a week later.

Inspection of Table 6-1 reveals a strong and consistentrelationship between dosage and efficacy: better and fasterunilateral ECT outcomes are obtained at higher dosagelevels. The big jump in efficacy (which I have indicated by adotted line) occurs between mean doses of 195 mC and 378mC: no study in the 195 mC or lower range obtained betterthan 65% efficacy, and every study in the 378 mC orgreater range yielded at least 65% efficacy. Because everystudy in the lower-dosage range employed titration-baseddosing, and every study but one in the higher-dosage rangeemployed fixed-or age-based dosing, the table isuninformative concerning the relative efficacies amongtitration-based, fixed-dose, and age-based dosing.

What is quite clear from the table, however, is that highstimulus dosing (e.g., in the 375 mC to 500 mC range) isquite effective for unilateral ECT, regardless of the methodemployed, yielding greater improvement and better responserates than low stimulus dosing, and with 2 fewer treatmentson average (mean ECTs = 6.8 for high-dose versus 8.9 for

low-dose).

In a recent retrospective comparison of age-based andtitration-based dosing for unilateral ECT, Frey et al. (2001)reported the clinical results obtained in 2 groups ofdepressed patients receiving either a stimulus dose in mCequal to 5 times the patient's age in years (mean dose atfirst ECT = 312 mC), or a dose titrated to 2.5 Ã— theseizure threshold (mean dose at first ECT = 92 mC). After acourse of 9 unilateral ECTs, 82% of patients in the age-based dosing group exhibited moderate-markedimprovement, compared with 36% of patients in the titrated-dose group. As expected from the more than 3-fold higherdoses administered, patients receiving age-based dosingconsistently had shorter seizures than those receivingtitration-based dosing. Following the course of unilateralECT, 64% of patients in the titration-based group had to beswitched to bitemporal ECT because of an inadequatetreatment response, compared with 9% of patients in theage-based group. There were no between-group cognitivedifferences following the course of unilateral ECT.

In addition to their greater absolute efficacy, higherstimulus intensities also accelerate the therapeutic responseto ECT (Ottosson, 1960; Robin and de Tissera, 1982;Sackeim, Devanand, and Prudic, 1991; Abrams, Swartz, andVedak, 1989, 1991; Pettinati et al., 1994; McCall et al.,1995).

Prudic et al. (1994) have expressed concern that with fixed,high-dose stimulation some patients receiving unilateral ECTwill receive dosages that are barely suprathreshold, therebyundermining the therapeutic effect. This argument overlooksthe fact that many fixed, high-dose studies reported haveachieved therapeutic results that are substantially betterthan those obtained with stimulus titration (Abrams, Swartz,and Vedak, 1989, 1991; Pettinati, 1994; Lamy, Bergsholm,and d'Elia, 1994; McCall et al., 1995, 2000).

To date, the studies of McCall et al. (1995, 2000) are theonly published prospective comparisons of the clinicalefficacy of titrated-dose versus fixed-dose unilateral ECT. Intheir initial study (McCall et al., 1995), elderly patients(mean age = 76), receiving ECT with a fixed 403-mC doseresponded faster and required fewer treatments than thosereceiving titrated, 2.25 Ã— threshold stimulation (mean dose151 mC). The high-dose patients also exhibited shorterseizures and greater regularity of EEG morphology (McCalland Farah, 1995). Strikingly, despite the almost threefoldhigher dosage received by the fixed-dose group, memoryself-ratings after ECT were not different for the two groups.A subsequent study (McCall et al., 2000) in a different,smaller, sample (n = 72) confirmed the original results:

P.118

after about 6 ECTs, 67% of the fixed high-dose patientswere ECT responders, compared with only 39% of thetitrated moderate-dose patients.

Cognitive Effects of High-DoseUnilateral ECTThe fear has often been expressed that because high-doseunilateral ECT without titration provides a dose that can beseveral multiples of the seizure threshold in some patients,excessive cognitive dysfunction might result (Sackeim,Devanand, and Prudic, 1991; Sackeim, 1994a; Beale et al.,1994; Prudic et al., 1994; Enns and Karvelas, 1995):

â!" the practice of using electricalintensities far in excess of that needed toproduce seizures undoubtedly contributes toadverse cognitive side ef fects. (Sackeim,Devanand, and Prudic, 1991)

Memory and cognitive effects of ECT are discussed in detailin Chapter 10, but it is worth noting here that, for aresearch area that has been driven by patient complaints forover 60 years, it is remarkable how many studies find that,regardless of treatment electrode placement, stimulus wave-form, or dosage level, patients tend to subjectively ratetheir memory as unchanged or improved after ECT(Cronholm and Ottosson, 1963a; Small, 1974; Johnstone etal., 1985; Shellenberger et al., 1982; Weiner et al., 1986b;Pettinati and Rosenberg, 1984; Mattes et al., 1990; Calev,Nigal, and Shapira, 1991; Sackeim et al., 1993, 2000;McCall et al., 1995, 2000; Coleman et al., 1996).

Prudic, Peyser, and Sackeim (2000) have recently reviewedthis data and concluded that ECT-induced improvement inaffective state is probably responsible for the generallybenign self-assessments of post-ECT memory function.Nevertheless, the fact remains that treatment -inducedmemory impairment is not considered a problem by mostpatients who receive ECT, and this includes many who havereceived the very high dosages delivered by sine-wavestimuli, which dosages are far greater than those usually delivered by conventional brief pulse devices.

Unlike the earlier study by McCall et al. (1995), the recentone (McCall et al., 2000) included objective cognitive testingat baseline and 1-2 days after the ECT course, assessingimpairments in global cognitive status and retrograde andanterograde memory, which will be fully discussed in Chapter10. This is the only published study comparing the memoryand cognitive effects of titrated-dose and fixed-dose ECT.

P.119

Examined on a variety of cognitive tests, patients in thefixed-high-dose group recalled significantly fewerautobiographical items from a baseline interview thanpatients receiving a 2.25Ã— threshold mean dose of 136 mC(characterized as â!œmoderateâ! !). Although Mini -MentalState Examination (MMSE) scores fell 2.6 points farther inthe fixed high-dose than the titrated moderate-dose group(26.7 to 23.4, compared to 25.7 to 25.0), the mean post-ECT MMSE score of 23.4 for the high-dose group remainedin the normal range for that test. Finally, although theauthors detected a significantly greater delay in post-ECTreorientation time for the high-dose patients whencomparing the first to the second ECTs, reorientation timewas not significantly different between the 2 groups afterthe first treatment, and was longer for the fixed high-dose

group after the second; for technical reasons, reorientationtime after sub sequent studies was not examined.

Moreover, these rather modest cognitive differences betweentitrated-dose and fixed-high-dose stimulation were obtainedin a study employing a titrated dose that was only 1/3 aslarge as the fixed high dose. Nobody today would (orshould) use 2.25 Ã— threshold dosing for unilateral ECTâ!”6Ã— threshold is now the titration standard. Had McCall etal. (2000) been prescient when they first began their study,they would have used a 6Ã— threshold dosing strategy toyield a stimulus charge that was 6/2.25 or 2.67 times higherthan the 136 mC dose they actually used: 363 mC. There islittle doubt that the cognitive differences they foundbetween 403 mC and 136 mC would not have been manifestin a comparison of 403 mC with 363 mC.

The Columbia studies (Sackeim et al., 1993, 2000) are theonly ones that have compared different threshold multiplesfor unilateral ECT. It is notable that Sackeim et al. (1993),although reporting significantly greater effects of moderate-dose (2.5Ã— threshold) than low-dose (1.5Ã— threshold)unilateral ECT on the right hemisphere-specificneuropsychological measures of nonsense shape recall andmemory for facial emotional expression, nevertheless foundno deleterious effect of the higher dosage on the moreclinically relevant measure of autobiographical retrogradeamnesia, the phe nomenon that is most troubling to patientsand their families.

Moreover, both of the high-dose versus low-doseneuropsychological differences found by Sackeim et al.(1993) immediately after a course of treatment were nolonger detectable a week later.

As mentioned above, the 2.5Ã— threshold dose for unilateralECT originally characterized by Sackeim et al. (1993) as â

P.120

!œhighâ! ! would today be considered only moderate and isat the very bottom of the officially recommended range(American Psychiatric Association, 2001). Because 6Ã—threshold dosing for unilateral ECT is the multiple that hasachieved the best clinical results in depressed patients (seeTable 6-1), it has become the standard for those employingtitration-based dosing. To date, Sackeim et al. (2000) arethe only investigators to have reported memory andcognitive data for 6Ã— threshold titrated-dose unilateralECT.

In a 2-7 day post-ECT comparison among 3 dosage levelsfor unilateral ECT (low, moderate, and high, withcorresponding mean doses of 132 mC, 173 mC, and 441 mC,respectively), these authors reported greater impairment forthe high-dose group than the 2 lower-dose groups on 2 outof more than 25 independently scored tests and subtestsstudied: delayed reproduction of a complex figure, anddelayed short story recall. These differences were no longerdetected at the first post-ECT follow-up assessment 2months later. Specifically, at no time were there detectableeffects of high-dose unilateral ECT on measures of memoryfor famous events, or autobiographical (personal) memory.Just as for the Sackeim et al. (1993) cognitive findingsreviewed above, these modest cognitive differences haveunknown, but prob ably minimal, clinical implications.

Taken together, then, the McCall et al. (2000) and Sackeimet al. (1993, 2000) studies amply confirm that fears ofundue or unacceptable cognitive consequences of high-doseright unilateral ECT are unwarranted. In view of this, therecent article of Prudic et al. (2001), who surveyed ECTpractices in New York State and assigned a â!œcognitivedeficit scoreâ! ! to each hospital depending on the method ofstimulus dosing used (not surprisingly, the authors' ownstimulus titration method received the most-favorable score,and fixed dosing, the least -favorable score), is not onlyunscientific, but lacks the patient examination data requiredto demonstrate either the existence or degree of theputative cognitive deficit, or its relation to the dosingmethod employed.

Cognitive Effects of Long Stimulus-Train DurationSome of the high-dose studies cited also employed longerstimulus trains (e.g., Pettinati et al., 1990; Abrams, Swartz,and Vedak, 1989, 1991) and Sackeim (1994a) has raised thespectre of long-duration stimuli causing excessive dysmnesiaby stimulating the brain after it is already in a hyperactivestateâ!”that is, while it is in seizure. In support of thisview, he introduced the concept, derived from animal data,

P.121

that the seizure threshold might act as a â!œfilterâ! ! for thecognitive side effects of ECT.

Reification of a metaphor aside, this view does not take intoaccount the fact that a 1 second stimulus dose that crossesthe seizure threshold by the end of that second is unlikelyto do so when its rate of administration is cut by 90% (e.g.,by decreasing the frequency to spread the same dose over 8seconds). Indeed, it is far more likely that reducing the rateof administration of the ECT dose will substantially retardseizure onset, thus reducing, rather than accentuating,cognitive side effects (analogous to the relation between thesteepness of the increase in plasma concentrations of a drugand the occurrence and severity of side effects). This is, infact, precisely what Sackeim et al. (200Ib) demonstrated intheir recent study of ultrabrief pulse ECT administered witha stimulus duration of up to 8 seconds: the least cognitiveimpairment for unilateral ECT yet reported from theirlaboratory, or indeed, in the literature.

Bitemporal ECT: The relation of stimulus dose to efficacy ismore straightforward for bitemporal than for unilateral ECT,as shown in Table 6-2.

Inspection of Table 6-2 reveals little evidence for dose-sensitivity of bitemporal ECT; only Letemendia et al. (1993),who gave the smallest dose, achieved results that could becorrectly characterized as subtherapeutic. Moreover, thereare no listed improvement or response rates for bitemporalECT below 50%, as occurred for many of the unilateral ECTstudies summarized in Table 6-1; the results for bitemporalECT are clustered in a much narrower range, mostly between70% and 90%. Thus, bitemporal ECT is far less dose-sensitive than unilateral ECT, and yields decent clinicalresults

at all but the lowest dosages. However, because bitemporalECT is quite capable of achieving therapeutic results in the80% to 90% range, this should serve as the benchmarktreatment goal for practitioners.

Table 6-2 Brief-pulse bitemporal ECT responseby stimulus dose

Author

Mean

dose Improvement

Response

rate (%)

Mean

#

ECTs

Dose

method

Letemendiaet al.

148mC

56 11.5 1 Ã—

(1993)

Petrides etal. (inpress)

187mC

73 87 7.8 1.5 Ã—


192mC

71* 70 9.4 1 Ã—

Lerer et al.(1995)***

212mC

70 9.9 1.5 Ã—


212mC

66* 70 10.6 1 Ã—

Sackeim etal (1993)

321mC

74* 70 9.3 2.5 Ã—

Bailine etal. (2000)

235mC**

82 5.4 1.5 Ã—

Abrams,Swartz,and Vedak(1991)

378mC

79 78 6.0 Fixed


441mC

80 8.3 2.5 Ã—

* Calculated from figure.** Estimated from reported value of 30 J.*** 3 Ã—/week ECT.

Relation of Stimulus Parameters toSeizure EfficacyAs already noted, both pulsewidth and stimulus trainduration affect stimulus efficiency, so it is reasonable toexpect they might thereby also affect seizure efficacy.Unfortunately, because several of the highest-dose studiesreviewed in Tables 6-1 and 6-2 also used the longeststimulus trains and shortest pulsewidths, it is not possibleto separate the effects of these latter variables on clinicaloutcome from those of dosage.

P.122

Stimulus Charge RateSwartz (1994b, 1995) has approached the problempragmatically by focusing on the rate of administration ofthe stimulus charge in relation to the efficiency of seizureinduction. He observed that lower charge rates were consistent with greater efficacy in seizure induction and betterquality seizures, and recommended a charge rate of about25 to 50 mC per second.

Of course, because current is fixed in modern brief-pulseECT devices, the charge rate is a function of pulse widthand frequency: reducing either will reduce the charge rateproportionately (other things being equal, reduc ing thepulse width and/or pulse frequency lowers charge rate byspreading the dose over a longer interval).

Swartz makes the additional important point that it isessential for efficiently low charge rates to be available atthe highest total stimulus charge settings for ECT devices.This is because it is the olderâ!”usually maleâ!” patientsthat have the greatest difficulty in obtaining good qualityseizures, sometimes failing to develop any seizure activity atmaximum device ca pacity. Whereas inefficient stimuli canreadily induce seizures in low-threshold patients, they oftencannot do so in high-threshold patients.

For any given ECT device, the lowest charge rates areobtained by setting pulsewidth to the minimum, and thenadjusting stimulus frequency to maximize stimulus trainduration. Stimulus programs that perform this functionautomatically have been incorporated into commerciallyavailable ECT devices.

Seizure DurationSeizure duration has been variably reported to decrease(Sackeim et al., 1987b, 1993; Shapira et al., 1996) orincrease (Di Michele et al., 1992; Scott and Boddy, 2000)across a course of ECT, sometimes with a concurrentincrease in the seizure threshold. Although several studieshave reported an inverse relation between seizure thresholdand duration (Sackeim et al., 1987b, 1993; Krueger et al.,1993; Coffey et al., 1995a), more evidence exists for adissociation (Sackeim, Devanand, and Prudic, 1991; McCallet al., 1993a; Shapira et al., 1996; Scott and Boddy, 2000),and even where such a relation obtains across a sample,almost half the subjects may fail to show it (Coffey et al.,1995b).

Only one of the many studies in the modern era that havelooked for a positive correlation between seizure duration

P.123

and therapeutic outcome has found one (Delva et al., 2001),thus reopening an issue that had long been consideredclosed. These investigators found about one third longerseizures in responders than nonresponders to brief-pulseright unilateral ECT. The reason for their success indetecting a relationship so many before them failed to findis not readily explainable, although it is notable that theirstimulus titration method was carefully designed to producethe lowest possible extent of dosage above threshold. Manymore studies, however, have reported that an inverserelation between seizure duration and outcome actuallyexists (Nobler et al., 1993; Krystal, Weiner, and Coffey,1995; McCall et al., 1995; Shapira et al., 1996)â!”theshorter the seizure, the better the treatment responseâ!”anassociation that may have an important interaction withdosage.

Moreover, several investigators (Robin, Binnie, and Copas,1985; Sackeim, Devanand, and Prudic, 1991; Krystal et al.,1993; Shapira et al., 1996; Frey et al., 2001) have alsoreported an inverse relation between stimulus dose andseizure durationâ!”the higher the dose, the shorter theseizureâ!” thereby confuting the frequently offered advice toincrease the stimulus dose if seizures are too short, andreduce it if they are too long.

EEG Seizure QualityThere exists a consensus of expert opinion that clinicallyeffective stimulation for ECT results in morphologically well-developed, symmetrical, synchronous, high-amplitude seizureactivity that is followed by marked postictal suppression(Fig. 6-1) and accompanied by a prominent tachycardiaresponseâ!”phenomena that all reflect increasedintracerebral seizure intensity or generalization (e.g., morerapid development and spread) and therefore more effective,seizures (Robin, Binnie, and Copas, 1985; Krystal et al.,1992 , 1993; Abrams, 1992, 1996; Swartz, 1993a,b , 1996 ,in press; Nobler et al., 1993; Nobler, Luber, and Moeller,2000; Krystal and Weiner, 1994; McCall and Farah, 1995;Krystal, Weiner, and Coffey, 1995; Suppes et al., 1996;Hrdlicka et al., 1996; Folkerts, 1996; Fink and Abrams,1998; Krystal, 1998).

Three ictal EEG variables in particular have received themost attention: postictal suppression, ictal amplitude orpower, and interhemispheric coherence. Krystal et al. (1992)were the first to suggest that this group of vari ables mightreflect the intensity and generalization of ECT-inducedseizure activity.

P.124

P.125

EEG Postictal SuppressionAs described in Chapter 4, the more or less abrupt andpronounced fall in EEG amplitude immediately followingseizure termination is termed postictal suppression. Severalearly investigations employing visual EEG analysis reported agreater degree of postictal suppression with bitemporal thanunilateral ECT (Small et al., 1970; Abrams, Volavka, andFink, 1973; Staton, Hass, and Brumback, 1981), aphenomenon confirmed and quantified by Krystal et al.(1992) using power spectral (quantitative digital) EEGanalysis. The general perception that bitemporal ECT wasmore effective than uni lateral ECT in treating depression(see Chapter 7) led naturally to studies of the relation ofpostictal suppression to clinical outcome.

Nobler et al. (1993) were the first to report an associationbetween the degree of EEG postictal suppression and clinicalimprovement in depression. Employing a visual EEG ratingscale, they found that greater unilateral ECT dosagesinduced greater postictal suppression. Moreover, the degreeof suppression measured across the entire patient sample(including both unilateral and bitemporal ECT) was positivelyassociated with outcome. A later study from this group(Nobler, Luber, and Moeller, 2000) using quantitative digitalEEG analysis confirmed these results.

Krystal et al. (1993) also found greater postictal suppressionwith higher stimulus doses, for both unilateral andbitemporal ECT. In a later study that used a global scale toscore severity of illness, Krystal, Weiner, and Coffey (1995)again found greater postictal suppression with higher doses,plus a positive association between the extent of postictalsuppression and global

outcome ratings. Multiple-symptom depression rating scalesthat had been obtained but not reported were subsequentlyanalyzed (Krystal et al., 1996) and found to confirm theresults of the global ratings.

Figure 6-1 Well-Developed EEG Seizure. (Adapted from Fink and Abrams,1998.)

In a study of 33 depressed patients receiving either brief-pulse or sine-wave, unilateral or bitemporal ECT, Suppes etal. (1996) also confirmed a significant relationship betweenthe degree of EEG postictal suppression and the percentreduction in depression scale scores.

A recent study from the 4-hospital CORE collaborative ECTgroup (Petrides et al., 2000) found that of 260 patients withmajor depression who completed a course of bitemporalbrief-pulse ECT, the average postictal suppression index forthe 85% who achieved remission was 87, significantly higherthan for the 15% who failed to achieve remission (althoughthe ab solute between-group difference was small).

EEG Ictal AmplitudeEEG amplitude (measured as voltage with constant-currentdevices and often expressed as power, or voltage squared)is a reflection of ictal intensity that has also been variouslyfound to be greater with higher-than lower-dose stimulation,and for bitemporal than unilateral ECT (Krystal et al., 1992;Nobler et al., 1993; Krystal, Weiner, and Coffey 1995;Nobler, Luber, and Moeller, 2000; Krystal et al., 2000a).Several of the studies cited earlier on the relation ofpostictal suppression to clinical response also reportedsimilar results for EEG amplitude: the greater the amplitude,the better the outcome (Nobler et al., 1993; Krystal et al.,1993; Krystal, Weiner, and Coffey, 1995; Nobler, Luber, andMoeller, 2000). In addition, Folkerts (1996) examined therelationship of several different visually rated EEG measuresto clinical response, and found higher spike-wave phase

P.126

(mid-ictal) amplitude associated with greater clinicalimprovement; similarly, Hrdlicka et al. (1996) demonstratedby visual analysis that more â!œintenseâ! ! EEG seizureactivity (char acterized especially by a high amplitude) wasassociated with a more favorable therapeutic outcome.

Broken down by frequency, the delta and theta bandwidths(that include frequencies from 0.5 to 6.5 Hz) typicallyaccount for the greatest amount of EEG power during thepolyspike and slow-wave (mid-ictal) phase that representsthe bulk of the seizure (Staton, Enderle, and Gerst, 1981;Staton, Hass, and Brumback, 1981; Nobler, Luber, andMoeller, 2000), although the latter study found main effectsof global power across all 4 frequency bands for both dosageand electrode placement conditions.

Interhemispheric CoherenceThis variable, which requires quantitative digital EEGtechniques for its assessment, is a frequency-specific, time-based, cross-correlation function

that measures the degree to which corresponding loci inboth hemispheres are generating EEG signals in unison.Specifically with regard to the seizure, coherence reflectsthe extent to which the hemispheres are dischargingsimultaneously (e.g., as in response to a centralpacemaker). In this sense, interhemispheric coherencemeasured during the seizure can be considered a reflection(perhaps the most important one) of seizure generalization.

Krystal et al. (1992) found greater interhemisphericcoherence for bitemporal than unilateral ECT during all ictalphases, a relationship that was reversed during theimmediate postictal phase. Considering that coherenceshould fall markedly at seizure termination, whencentrencephalic driving abruptly ends, this latter findingnicely complements that of greater postictal suppressionwith bitemporal than unilateral ECT. Coherence measuresobtained by this group in a subsequent pilot study (Krystalet al., 1993) were not subjected to statistical analysis;qualitatively, however, greater coherence was associatedboth with higher stimulus doses and bitemporal as comparedwith unilateral ECT.

Clinical Utility of Computer-DerivedIctal EEG MeasuresBecause unilateral ECT is far more dose-sensitive thanbitemporal ECT, it is primarily for unilateral ECT dosing thatictal EEG guidance is sought. Although much useful clinicalinformation can be gleaned from visual inspection of theictal EEG during ECT (Fink and Abrams, 1998), computer -

P.127

derived ictal measures provide the greatest potential clinicalvalue because of their precision, objectivity, reliability, andability to detect and quantify phenomena andinterhemispheric relationships not apparent to the nakedeye.

Although, as noted above, Nobler, Luber, and Moeller (2000)found superior clinical outcome with unilateral andbitemporal ECT to be associated with both greater ictalpower and greater postictal suppression, they characterizedthe relationship as modest and raised doubts concerning thepotential clinical utility of EEG algorithms based on thesefeatures to guide stimulus selection. This judgment ispremature, however, because it is based on very smallsample sizes (e.g., there were only 12 subjects in thehigher-dose uni lateral group, arguably the most importantof the 4 groups studied).

Moreover, technical limitations of the recording equipmentobscured the first 10 seconds of the seizure, correspondingto the tonic motor phase and manifested by high-voltagespikes, phenomena that may contain important informationrelevant to treatment response. Unilateral ECT dosages werelow (the mean â!œhighâ! ! 2.5Ã— threshold dose forunilateral ECT was only 151 mC), and the more robustseizures and therapeutic impact of the much higher ECTdoses now recommended are likely to exhibit strongerassociations. Finally, the stimulus parameters employed wereessentially limited to the inefficient combination of a 1.5 mspulsewidth and a 1 sec stimulus train (Boylan et al., 2000);the use of more efficient stimuliâ!”especially in combinationwith substantially higher dosagesâ!”should also yieldstronger associations between EEG features and clinicaloutcome.

Two commercially available ECT devices have incorporatedcomputer -based assessment of 1 or more of the ictal EEGmeasures described above â!”computed during the seizureand reported at its terminationâ!”thereby providing theclinician with information potentially relevant to the decisionwhether to restimulate the patient at a higher dose. To date,only 2 published studies have prospectively assessed thevalidity of such ECT device-based measures. Folkerts (1996)studied 40 patients receiving right unilateral ECT for majordepression, and defined treatment response as a drop of atleast 50% in depression scale score both for the entiretreatment course, and for the first 4 ECTs. The meanpostictal suppression index was higher in rapid respondersthan in slow or nonresponders, a difference that just failedto reach significance due to the small sample size. Petrideset al. (1998) reported a significant association between thepercent fall in depression scale scores and a measure of

P.128

postictal suppression in a sample of 260 depressivesreceiving moderate-dose bitemporal ECT. The ability todiscriminate between treatment responders andnonresponders was limited in both studies, however; theaverage postictal suppression index among responders wasonly modestly higher than among nonresponders.

In a retrospective study, Krystal et al. (2000a) combinedEEG ictal and postictal amplitudes in a stimulus-dosingmodel intended to make clinically relevant â!œseizureadequacyâ! ! distinctions among ECT treatments, usingclinical outcome obtained previously in 38 patients as thedependent variable, and comparing the model results withthose that had been obtained via titration-based, seizurethreshold dosing. The authors' stated rationale was todevelop a method to avoid seizure-based dosing (thestimulus titration method) because of the lack of evidencefor this method, and its failure to ensure seizure adequacy.

Although the model suggested a dosage increase in anumber of patients for whom no such increase had beengiven (based on titration against seizure duration), andidentifed several other patients who had received dosage increases that were not indicated by the model,methodological inadequacies render the model ofquestionable utility.

These inadequacies are the inclusion of multiple (up to 7)seizures per patientâ!”a statistically doubtful procedure thatcan substantially bias results â!”and the fact that the 2.25Ã— threshold dosing level chosen for unilateral ECT (whichaccounted for 2/3 of their sample), has been amplydemonstrated to be therapeutically inadequate, as detailedelsewhere in this chapter. Thus, rather than measuring thedegree of â!œseizure adequacy,â! ! the model appears onlyto measure degrees of seizure inadequacy, information of unknown clinical relevance.

Stimulus and DosingRecommendationsThe relations outlined above among stimulus parameters,stimulus dose, treatment electrode placement, ictal EEGcharacteristics, and treatment re sponse, form the basis forthe following approach to stimulus dosing.

Because bitemporal and bifrontal ECT are minimally dose-sensitive, any of the usual dosing methods will be effective.The simplest and most straightforward of these is the â!œhalf -ageâ! ! method of Petrides and Fink (1996), whichprovides a moderate stimulus dose that induces a seizure atthe first attempt in the majority of patients. To select thehalf -age dose using an ECT device with a continuously

variable stimulus dial calibrated according to the proportiondelivered of maximum device capacity, simply set the dial asclose as possible to one-half the patient's age in years. (Forother types of ECT devices just set the numerical value ofthe stimulus charge as close as possible to the patient's agemultiplied by 2.5.) The half -age method works much moreeffectively with a 0.5 ms pulsewidth than a 1 ms pulse-width(Swartz and Manly, 2000); for most patients, the shorterpulse also allows the remainder of treatments in the courseto be administered without increasing the dose. For stimulustitration, a 1.5Ã— to 2.5Ã— threshold dose should bechosen.

Unilateral ECT requires much higher doses than bitemporalECT to approach maximum efficacy. The simplest and mosteffective methods are the fixed-high-dose and age-basedtechniques. The fixed high dose is delivered by setting theECT device to deliver 75% to 100% of its maximumcapacity; the age-based dose by setting a continuouslyvariable stimulus dial to approximate the patient's age inyears (or selecting a stimulus charge in mC thatapproximates 5 times the patient's age). For stimulustitration, a 6Ã— to 8Ã— threshold dose should be chosen.

Stimulus parameters should be selected to maximizeefficiency. This means that, as a general rule, pulse widthshould not exceed 0.5 ms and stimulus train duration shouldbe the longest allowed for the dose selected; this will oftenbe as long as 6-8 seconds. With dose, pulsewidth, and duration thus determined, stimulus frequency falls where itmay. (See above discussion of lowest charge rate dosing.)

Examined visually, the monitored EEG recording shouldprogress clearly through all phases, manifesting thosequalities of amplitude, rhythm icity, symmetry, andtermination described earlier for a well-developed, ef fectiveseizure (Fig. 6-1).

For ECT devices that provide computer -derived measures ofictal EEG quality, it is best to administer the first seizure ofthe course with a very substantial stimulus doseâ!”regardless of treatment electrode placement or the dosingmethod chosen for subsequent treatmentsâ!”in order tomaximize the likelihood of inducing a high-quality seizure.For this purpose I prefer simply setting the ECT device todeliver the maximum dose. As suggested to me by Dr.Conrad Swartz, the EEG characteristics reported for thisinitial â!œbenchmarkâ! ! seizure can then serve as targetsfor adjusting the dose for subsequent seizures, with the aimof maintaining the computed ictal values at or near theirmaxima, thus titrating stimulus dose against seizure quality.

For example, if the first seizure given at 100% of maximumdose yields excellent postictal suppression, the second

P.129

treatment might be given at 80% of maximum; if postictalsuppression remains at or near the target level, the

3rd treatment might be given at 60% of maximum, and soforth, thus titrating the dose from treatment to treatmentagainst an ictal response initially obtained from the bestseizure a particular patient is capable of exhibiting.

Of course, as everywhere in medicine, clinical responseoverrides all other considerations: patients whose EEGpatterns reflect a high seizure quality at lower dosagelevels, but who are not showing clinical improve ment, needhigher doses; those who are enjoying a satisfactory responsedespite apparently poor-quality seizures require no doseadjustment.




> Table of Contents > Chapter 7 - Treatment Electrode Placement:

Bitemporal, Unilateral, Bifrontal

Chapter 7

Treatment Electrode Placement:

Bitemporal, Unilateral, Bifrontal

Efficacy has not, and has never been, theproblem with ECT. ECT remains,indisputably, the single most efficacioustreatment for serious depression.

The problem with ECT has been, and remains, the need todiminish ad verse cognitive effects. (Kellner, 2000). Thisquotation underlines an important point that is oftenoverlooked, which is that from the very outset, thepronounced amnestic effects of bitemporal sine-wave ECTdrove a substantial portion of the ECT research conducted,leading first to the introduction of the brief-pulse stimulusin the early 1940s, and to unilateral treatment electrodeplacement a few years later. Dissatis faction with theefficacy of unilateral ECT led, in turn, to the introduction ofbifrontal placement.

Unilateral ECT, especially when administered with a brief-pulse square-wave stimulus, represented an importanttechnical advance in the field of convulsive therapy. Moreclinical research was conducted since 1960 on the cognitiveand therapeutic effects of this modality than on any othersingle topic in the ECT literature; much of this research hasbeen directed primarily at resolving the controversyconcerning the precise clinical role of this method in relationto the older bitemporal ECT.

The early history of ECT research was largely characterized

P.131

by attempts to reduce the side effects of seizures inducedby the original sinusoidal currents delivered via bitemporalelectrodes. Only a year after the first English-languagepapers on ECT appeared (Kalinowsky, 1939; Fleming, Golla,and Walter, 1939; Shepley and McGregor, 1939), DouglasGoldman demonstrated the new treatment at the annualAmerican Psychiatric Association meeting, using a devicebuilt to his specifications by Offner (Fink, 1987). In view ofthe fact that Offner subsequently marketed the first commercial brief-pulse device 6 years later (Weiner, 1988), it islikely that Gold man used such a device in hisdemonstration.

Friedman (1942) and Friedman and Wilcox (1942) were thefirst to publish data on the use of modified sine-wavecurrents, which they administered via a left -sided unilateralelectrode placement. These authors' primary interest wastechnicalâ!”to investigate the amount of electric currentdeliv-eredâ!”and they made no mention of therapeuticeffects, memory loss, or

confusion. These studies were continued and expanded byProctor and Goodwin (1943), Liberson (1944)â!”whointroduced the term â!œbrief-stimulusâ! ! to describe the0.16-to 0.33-ms spikes employed and was also the first toreport right-sided unilateral ECT placementâ!”Liberson andWilcox (1945), Moriarty and Siemens (1947), and Liberson(1948). A variety of electrode placements were used (mostof them right unilateral), but although they were not appliedsystematically, these authors all reported reduced cognitiveside effects with their new techniques.

Goldman (1949) invented unilateral ECT as we know it: Hewas the first to specify that unilateral treatment electrodesshould be placed over the right hemisphere in order to avoidthe speech areas. In 112 patients, he observed clinicalimprovement equal to that produced by bitemporal ECT butwith a â!œmarked diminution and, at times, absence ofconfusion associated with the electric shock therapy.â! ! Hisattribution of the beneficial effect thus obtained to the brief-pulse, square-wave stimulus he used was only partlycorrectâ!”right unilateral electrode placement probablyplayed the more important role.

Workers into the mid-1950s (Bayles, Busse, and Ebaugh,

P.132

1950; Blaurock et al., 1950; Impastato, Berg S, Parella BL(1953); Liberson, 1953; Liberson et al., 1956) continued toreport reduced memory loss and confusion with unilateralplacements, particularly right-sided ones, but they neverattempted to separate the effects of stimulus type fromelectrode placement, or to further characterize the specificbenefits accruing from right unilateral placement.

The Argentine psychiatrist Thenon (1956) was the first todemonstrate the specific link between right unilateralelectrode placement and reduced memory loss andconfusion; he also observed an accentuation of post-ECTslow-wave EEG activity over the treated hemisphere. Hecalled his method monolateral electroshock. Two years later,Lancaster, Steinert, and Frost (1958), apparently entirelyunaware of the work of Goldman or Thenon, published thefirst English-language paper devoted to this technique,giving it its present name of unilateral ECT.

Their study was also notable for employing randomassignment to treatment and blind assessment of depressionand orientation using objective rating scales. In 21 patientsreceiving unilateral ECT, orientation and recall returnedsignificantly faster than in 15 controls who receivedbitemporal ECT. Moreover, these authors observed automaticbehavior, dazed expression, and restlessness to be lessprominent after unilateral ECT. They reported that 4bitemporal ECTs reduced depression scores by 71, comparedwith 54 for unilateral ECT, noted â!œslightly better andmore complete remissionâ! ! with bitemporal ECT, andtherefore recommended that it be preferentially given to themore severely depressed patients: involutional depressives,patients who were actively suicidal, depressed patients whofailed to show substantial improvement after 6 unilateralECTs, and catatonic schizophrenics who were â!œdangerously impulsiveâ! ! (these latter patients might wellreceive a diagnosis of acute mania today).

This report of Lancaster, Steinert, and Frost (1958) alreadycontained each of the elements of what later was to becomeâ!œthe unilateral ECT controversyâ! !: (1) sharply reducedmemory loss and confusion with unilateral ECT; (2) a formalequivalence for right unilateral and bitemporal ECT onobjective outcome measures; and (3) the authors' clinical

impression that bitemporal ECT was nonetheless moretherapeutically potent than right uni lateral ECT.

Over the next 30 years, dozens of comparisons of unilateraland bitemporal ECT were published, without providing eithera resolution of the controversy, or objectively derivedguidelines for practitioners as to precisely how to choosebetween the two methods for a particular patient (Abrams,1997). Many investigators, although finding the methodsequally effective on objective measures, neverthelessprovided clinical impressions to the effect that unilateral ECTwas less effective, or conducted further analyses showing anadvantage for unilateral ECT in a particular patientsubgroup.

Despite its thoroughly documented and large cognitiveadvantages over bitemporal ECT, numerous surveys revealedthat right unilateral ECT was used far less widely than mightbe expected (Heshe and Roeder, 1976; American PsychiatricAssociation, 1978; Asnis, Fink, and Saferstein, 1978;Fredericksen and d'Elia, 1979; Gill and Lambourn, 1979;Pippard and Ellam, 1981; Tancer et al., 1989; Farah andMcCall, 1993). The most recently conducted survey (Sackeimet al., 2000) revealed that, more than 40 years after theintroduction of unilateral ECT, about 80% of ECT units inNew York State, for example, still gave more bitemporalthan unilateral ECT.

It was only in the mid-1980s that the reason for the widelyperceived superior efficacy of bitemporal ECT was finallydemonstrated to be dose-related: unilateral ECT was farmore dose-sensitive than bitemporal ECT (Sackeim et al.,1987a). This important discovery was initially made byadministering for each electrode placement the lowestelectrical dose that would produce a seizure of at least 25-second duration by EEG criteria, using a method-of-limitsprocedure that titrated dosage against seizure duration toobtain an estimate of the seizure threshold. The just-above-threshold dose that worked well for bitemporal ECT wasineffective for unilateral ECT; the reduced efficacy ofunilateral ECT relative to bitemporal ECT was due to thefact that unilateral ECT was often being given at too low adose. All seizures were not equalâ!”as had been previouslyclaimed and widely acceptedâ!”effective seizures withunilateral ECT required higher dosages than for bitemporal

P.133

ECT.

Following this demonstrationâ!”and for well over a decadenowâ!”the focus of unilateral ECT research has shiftedtowards attempts to determine optimal dosing for thistreatment method, defined as approximating that ofbitemporal ECT, but without the latter method's occasionallypronounced adverse effects on memory and cognition.

Table 7-1 shows the painfully few modern clinical studiesdirectly comparing the clinical impact of unilateral andbitemporal ECT, by which I mean studies that employ abrief-pulse, square-wave stimulus. Most of these

studies come from a single research group that employstitration-based dosing, with the general results as alreadyoutlined for dosage levels in the preceding chapter (Tables6-1 and 6-2). Our aim here, however, is to directly comparethe relative therapeutic and cognitive results obtained withbrief-pulse, square-wave, unilateral and bitemporal ECT,which will require a more detailed examination of thestudies.

Table 7-1 Direct comparisons ofbrief-pulse unilateral and

bitemporal ECT

Treatment Results*

Study Unilateral Bitemporal Dosingmethod


17 70 1Ã—â!”1Ã—


43 70 2.5Ã—â!”1Ã—

Sackeim etal. (2000)**

80 (65) 80 (65) 6Ã—â!”2.5Ã—

Letemendiaet al.(1993)

50 56 1Ã—â!”1Ã—

Abrams,Swartz, andVedak(1991)

79 78 fixed

fixed, fixed dose; (n )Ã—, titration-based dose at(n ) times threshold.

* Response rate or % improvement.** Immediate post-ECT response (response 1 weeklater).

Following the initial demonstration (Sackeim et al., 1987a)that just-above-threshold (1 Ã— threshold) dosing yieldedonly a 17% response rate for unilateral ECT, compared with70% for bitemporal ECT, the study was repeated using 2unilateral dosage levels: I Ã— and 2.5Ã— threshold(Sackeim et al., 1993). Both unilateral dosage levels stillexhibited a smaller clinical response than bitemporal ECTgiven at I Ã— threshold: 28% and 43%, respectively, for theI Ã— and 2.5Ã— threshold levels for unilateral ECT,compared with 70% again for 1 Ã— threshold bitemporalECT. A third study was therefore undertaken using yethigher levels of dosage above threshold for bothplacements, and 3 dosage conditions for unilateral ECT(1.5Ã—, 2.5 X, and 6Ã— threshold), versus an increasedlevel of 2.5Ã— threshold for bitemporal ECT (Sackeim et al.,2000). At the completion of this study, 20 years since thefirst data were collected for the initial investigation of thisseries, 6Ã— threshold unilateral ECT was found toapproximate 2.5Ã— threshold bitemporal ECT in therapeutic

P.134

impact: the response rates were 80% for each group 1-2days post-ECT, falling to 65% several days later.

Only one other comparison exists of the relative clinicalefficacies of unilateral and bitemporal ECT that usedtitration-based dosing. Letemendia et al. (1993) studied 3treatment electrode placements using I Ã— threshold dosing:unilateral, bitemporal, and bifrontal. The results with thelatter method are described later in this chapter; thepercent improvement in depression scores obtained forunilateral and bitemporal ECT was 50% and 56%,respectively.

The only other clinical comparison of unilateral andbitemporal ECT is that of Abrams, Swartz, and Vedak(1991), who used a fixed, high dose of 378 mC (75% ofmaximum device output) for both unilateral and bitemporalbrief-pulse ECT and obtained immediate post-ECT responserates of 78% and 79%, respectively.

Inspection of Table 7-1 reveals that direct comparisons ofthe two methods confirm the observations made in Chapter6 that only the highest dosage levels for unilateral ECT (6Ã— threshold or 378 mC fixed) are capable of yieldingtherapeutic results equivalent to bitemporal ECT.

Left Unilateral ElectroconvulsiveTherapyLeft unilateral ECT has suffered from an undeservedly badreputation (Sackeim et al., 1982), due largely to a selectivereading of the handful of studies on the subject, all datingfrom before the introduction of the now-standard d'Elia(1970) unilateral ECT placement, and all employing the now-obsolete sine-wave stimulus (Abrams, 1997).

In an early sine-wave study of left unilateral ECT (Abramsand Taylor, 1974b) we found 1 left -unilateral plus 1 right-unilateral ECT per treatment session equal in antidepressantpotency to 2 right-unilateral ECTs per session over a total of4 sessions: 8 ECTs in all (Figure 7-1).

Years later, my associates and I conducted the only study ofleft unilateral ECT that employed a brief-pulse stimulus andthe d'Elia placement (Abrams, Swartz, and Vedak, 1989).

P.135

Our study was suggested by an initial observation of astriking instance of recovery from melancholia after 6 left

Unilateral, Bifrontal 135 unilateral ECTs in a fully dextralman (Leechuy and Abrams, 1987), which indicated that itwas premature to attribute the beneficial effects of ECT inmelancholia primarily to right hemisphere mechanisms. Wetherefore undertook a random-assignment, double-blind,controlled comparison of the antidepressant potencies offixed-high-dose (378 mC) right and left unilateral ECT in 30patients who satisfied criteria for melancholia, of whom 19re ceived right-sided and 11, left -sided, ECT. Depressionratings were blindly obtained immediately following the 3rdand 6th ECTs.

Figure 7-1 Antidepressant effects of right, then left,unilateral ECT.

Patients receiving left unilateral ECT showed an 85%improvement after 6 treatments, compared with only 70%for right unilateral ECT (Figure 7-2). Although this maintreatment effect difference was not significant, post-hoc

P.136

tests did show a significant advantage for left unilateral ECTfrom the 3rd to 6th treatments: left unilateral ECT workedfaster later in the course. (The only patient who was notfully dextral was randomly assigned to left unilateral ECTand exhibited 78% improvement after 6 ECTs.)

Two patients readmitted for second courses of ECT morethan 6 months after the the study were again randomlyassigned to treatment but excluded from further analyses.One received his first treatment course with right unilateralECT and his second with left unilateral ECT, improving 97%and 96%, respectively; the other patient received leftunilateral ECT the first time with a 64% improvement, andright unilateral ECT the second time with a 27%improvement. We did not compare the cognitive or memoryeffects of the 2 methods, but observed no apparentdifferences between them in day-to -day contact with thepatients.

Figure 7-2 Antidepressant effects of left vs. right unilateralECT. (Adapted from Swartz et al., 1994.)

Thus, our study showed that brief-pulse left unilateral ECT isat least as effective as brief-pulse right unilateral ECT, andperhaps more so. In fact, the 85% improvement we obtainedwith left unilateral ECT after only 6 treatments more closelyresembles the results reported for bitemporal ECT than forright unilateral ECT (e.g., Chapter 6, Tables 6-1 and 6-2).

It is striking that, more than a decade after our resultsappeared, and despite a recent favorable editorial on thesubject (Kellner, 1997), left unilateral ECT is rarely used asa primary treatment method. This seems likely to result fromthe fact that we did not perform cognitive testing, whereasmost of the early sine-wave studies reported morepronounced cognitive side-effects with left than rightunilateral ECT, even accounting for the expectedhemisphere-specific differences (Halliday et al., 1968; Cohenet al., 1968; Costello et al., 1970; Fleminger et al., 1970;Pratt, Warrington, and Halliday, 1971). There are tworeasons for not being unduly swayed by those reports,however. First, those same studies also reported that mostof the cognitive side-effects of left unilateral ECT weresignificantly less than for bitemporal ECT; and second, thecognitive side-effects of the brief-pulse stimulus employedtoday are substantially less than those of the sine wavestimulus used by those investigators 30 years ago.

In my view, therefore, left unilateral ECT provides not onlya viable alternate choice for musicians, artists, architects,and others who rely on unimpaired right-hemispherefunctioning, but has the potential for replacing rightunilateral ECT, because, like bifrontal ECT to be discussedbelow, left unilateral ECT yields a therapeutic effectequivalent to bitemporal ECT but with less cognitivedisturbance.

Bifrontal ECTAbrams and Fink (1972) first introduced bifrontal ECT bymoving standard bitemporal stimulus electrodes forward overthe forehead, spaced about 2 inches apart and 2 inchesabove the nasion. In an open clinical trial in 4 depressedpatients, we used a sine-wave device (mean dose = â!”700mC) to administer 4-6 seizures per treatment session for upto 3 sessions. Treatment response was variable andmoderate at bestâ!”no instances of dramatic improvement

P.137

were notedâ!”and 2 of the patients developed self-limitedcon fusional states. Because of the unfavorable risk/benefitratio for multiple bifrontal ECT we did not proceed furtherwith the method.

The following year, Abrams and Taylor (1973) reported anopen clinical trial of bifrontal ECT in 17 depressed patients,given at the rate of one seizure per session, 5 days perweek. Depression scale scores improved 46% and 69% after4 and 8 treatments, respectively. Memory testing in a sub-sample of 10 patients examined at baseline and 4-6 hoursafter the 8th treatment, revealed a slight improvement.Because we found no obvious advantage of daily bifrontalECT compared with results obtained earlier with daily orconventional-rate unilateral ECT (Abrams, 1967; Abrams etal., 1972),

and because several patients developed mild skin burns overthe forehead due to shunting of current secondary to thevery short interelectrode distance we used, we abandonedthe method.

Fortunately, however, Letemendia et al. (1993) resuscitatedbifrontal ECT 20 years later, improved by a greater distancebetween electrodes (placed 5 cm above the lateral angle ofeach orbit). In a prospective, random-assignment, double-blind, controlled comparison among 1Ã— threshold bi-frontal, bitemporal, and right unilateral ECT in 59 patientswith major depression, bifrontal ECT proved the mosteffective method. Hamilton depression scale scores improved61% after the first 6 treatments, compared with a 50%improvement with bitemporal ECT and a 45% reduction withright unilateral ECT.

One week after the completed treatment course (whichaveraged 10 ECTs for the bifrontal group, compared with 12ECTs and 16 ECTs for the bitemporal and unilateral groups,respectively), the bifrontal ECT patients were 74% improved,compared with 56% and 50% improvement for thebitemporal and unilateral groups, respectively). Moreover,bifrontal ECT also caused the least cognitive disturbanceamong the 3 methods, as mea sured in a subsample of 40patients (Lawson et al., 1990).

Titrated seizure thresholds were not significantly different

P.138

for the 2 bilateral methods (164 mC and 107 mC forbifrontal and bitemporal ECT, respectively), but both werehigher than the threshold for unilateral ECT (107 mC). The IÃ— threshold dosing used in the study strongly biased thetreatment response data of the study against unilateral ECTbecause that level of dosing is reported to be devoid ofclinical efficacy (Sackeim et al., 1987a). The finding of acognitive advantage for bifrontal over unilateral ECT,however, is substantially strenghtened by the 1Ã— thresholddoing strat egy, with which right unilateral ECT shouldexhibit minimal, if any, cog nitive impairment.

Bailine et al. (2000) conducted the most recent study ofbifrontal ECT (using the same placement as Letemendia etal., 1993), comparing it with bitemporal ECT in a sample of48 major depressives who were randomly assigned to the 2methods, administered with a brief-pulse stimulus and adose titrated to 1.5Ã— the seizure threshold (whichaveraged about 100 mC in each instance). The 2 treatmentmethods were equally effective in reducing depression scalescores (which fell 77% with bifrontal ECT and 82% withbitemporal ECT), but patients in the bifrontal group showedless ECT-induced cognitive impairment as measured by theMini-Mental State screen ing examination.

It is apparent that our early disappointment with bifrontalECT resulted from a very short interelectrode distance, theinefficiency of the sine-wave stimulus, and the atypicaltreatment schedules studied (multiple ECTs per session inthe first example, daily ECTs in the second). Modernbifrontal ECT as administered by Letemendia et al. (1993)and Bailine et al. (2000) appears to be a potential candidateto replace right unilateral ECT or bitemporal ECT, or both, inview of the facts that bifrontal ECT exhibits greater

antidepressant efficacy than right unilateral ECT, and fewercognitive side effects than either bitemporal ECT or rightunilateral ECT.

In my view, however, although these two studies provide afirm research base in support of bifrontal ECT, one or twoadditional confirmatory studies that include both clinical andcognitive assessmentsâ!”with follow-up data â!”will beneeded before bifrontal ECT can be considered as a standardtreatment option, let alone as a replacement for existing

methods.

Miscellaneous Treatment ElectrodePlacementsVarious modifications of the bitemporal, right unilateral, andbifrontal placements already described have been suggestedover the years: right frontoparietal (Muller, 1971; d'Elia andWidepalm, 1974; Erman, Welch, and Mandel, 1979;Alexopoulos, Young, and Shamoian, 1984), fronto-frontal(Widepalm, 1987), fronto-vertex (Pridmore and Ryback,1999), and left frontal-right temporal (Swartz, 1994a; Manlyand Swartz, 1994; Swartz and Evans, 1996). However, noneof these has been demonstrated to have either greaterclinical efficacy or fewer cognitive side-effects than any ofthe orig inal placements.

Electrode PlacementRecommendationsThe choice among treatment methods for ECT is based onthe same considerations that apply to all medicaltreatmentsâ!”the need to maximize clinical efficacy whileminimizing side effects. This clinical choice must always bemade flexibly, according to the particular needs andcircumstances of the patient, keeping in mind that the goalis to achieve the 80%-90% response rates that ECT iscapable of delivering in patients with major depression.

My own preference is to start every depressed patient onunilateral ECT, administering all treatments at maximumpresent-day standard device capacity (not the double-dosecapacity allowed in some countries), using the stimulusparameters recommended at the end of the previouschapter. In this way the patient will be assured of enjoyingthe maximum possible benefit from unilateral ECT at today'sstandards, with the fewest cognitive side effects for thedose administered. (Age-based dosing would provide a moreconservative stimulus, yet with good expected generalefficacy.) Those patients who do not improve sufficientlyafter a predetermined number of seizures with this methodcan be offered the option of switching to bitem poral ECT.

As already discussed in Chapter 6, concern about thememory and other cognitive side-effects of such high-dose

P.139

stimulation for unilateral ECT is misplacedâ!”the fact is,that there is no published research demonstrating that brief-pulse unilateral ECT administered at any dosage causesmemory disturbances that are still detectable even 2 weeksafter a course of treatment,

let alone 1, 2, or 6 months later, which might morerealistically provide cause for concern.

Equally viable is the alternate approach of starting allpatients on twice-weekly, brief-pulse bitemporal ECT, usingthe half -age dosing method and the stimulus parametersoutlined in the previous chapter (with the proviso that a 0.5ms pulsewidth be used because of the reported reducedefficacy of the 0.3 ms pulsewidth for bitemporal ECT). Ispecify twice-weekly treatment because it has been amplydemonstrated to reduce the amnestic effects of bitemporalECT without reducing its therapeutic impact (Lerer et al.,1995). In this way, the maximum possible benefit from ECTwill be assured with the fewest possible anesthesiainductions, and the cognitive side effects of bitemporal ECTwill be minimized.




> Table of Contents > Chapter 8 - Technique of Electroconvulsive

Therapy: Theory

Chapter 8

Technique of ElectroconvulsiveTherapy: Theory

Electroconvulsive therapy is not a trifling or inconsequentialprocedure to be delegated to a junior resident or casuallyadministered, unassisted, in the office. It is unique amongpsychiatric treatments: a significant medical interventionrequiring general anesthesia and entailing risks, however,small, of morbidity and mortality. The psychiatristadministering ECT adopts a role most like that of his medicaland surgical colleagues; to perform it well requires anintimate knowledge of the physiology and biochemistry ofinduced seizures, an understanding of the pharmacology ofanesthetic agents, familiarity with the physical properties ofthe electrical stimulus used, and the confidence and skill tolead a treatment team in the event of a medical emergency.

TrainingNo requirements yet exist for the training of psychiatrists ingiving ECT (Fink, 1986b); indeed, neither the AmericanCommittee on Graduate Medical Education and its ResidencyReview Committees nor the American Board of Psychiatryand Neurology provide any guidelines in this regard. It isclear, however, that training in ECT has been sadlyneglected by many institutions here and abroadâ!”thesurveys of Pippard and Ellam (1981) in Great Britain andLatey and Fahy (1985) in Ireland amply demonstrated howlittle attention has been paid to this subject in the BritishIsles; the Consensus Development Conference Statement(National Institutes of Health, 1985) from the United Stateslikewise calls for the increased training of medical studentsand residents in the use of ECT. A reasonable trainingprogram for psychiatric residents should provide didactic

P.141

course-work on ECT early in the first year of training(usually the inpatient year), including at least 3 hours oflecture and discussion on history, clinical indications,treatment response, side effects, precautions andcontraindications, medical physiology, cognitive effects, EEGeffects, the physical properties of the electrical stimulus,and the comparative and combined effects of psychotropicdrugs. To gain practical experience, the psychiatric residentshould personally administer at least 30 to 40 treatmentsunder the direct supervision

of a faculty member; this will usually require a 1-or 2-month rotation on an ECT service. The resident'sresponsibilities should include both the performance of ECTconsultations and assistance with the administration of ECT.After an initial week of observing all aspects of thetreatment procedure with full discussion by his supervisor,the resident in training should be responsible for treatingpatients under direct supervision. He should learn to insertan intravenous line, prepare and administer the anestheticagents, apply the treatment electrodes for unilateral andbitemporal ECT, deliver the electrical stimulus, and monitorthe induced seizure. He should observe patients in therecovery room awakening from their treatments and learn tomanage emergence delirium. He must become thoroughlyfamiliar with the requirements for informed consent for ECTand participate in obtaining such consent from new patients.

For residents enrolled in programs at facilities where ECT isunderutilized, educational videotapes can be used toaugment their training (Fink, 1986b). Such tapes areavailable (Frankel, 1986; Ries, 1987; Grunhaus, 1991;Alger, 1991) and should be used in conjunction withassigned readings and attendance at the training coursesconducted for credit during the year at various medicalcenters and at the American Psychiatric Association's annualmeeting.

The Electroconvulsive Therapy UnitThe ECT unit is an integral functioning part of thepsychiatric inpatient service and ought to be located nearby,not set apart in a surgical suite or other remote area of thehospital.

Physical Requirements

P.142

Electroconvulsive therapy should be given in pleasant, well-lit surroundings, air-conditioned in summer and heated inwinter, with ample room for staff and equipment and withwaiting and recovery areas designed to maximize privacyand minimize the apprehension engendered in patients byseeing or hearing others receiving or recovering fromtreatment. These points may seem self-evident, but thePippard and Ellam report (1981) revealed them, sadly, notto be so.

The treatment room should be large enough to comfortablyaccommodate a patient on a stretcher, all of the equipmentin the following list, and from 4 to 8 people, depending onwhether any observers are present. (The ECT unitappropriately serves as a training site for nursing andmedical students and should have adequate space for thisimportant function.) This will require from 225 to 400square feet of space with at least 4 grounded hospital -gradeoutlets. A telephone is needed for calling the patients' unitsand in the event that emergency assistance is required. Theroom should

have 2 doors: One for patient entry, the other leading to arecovery area. Recommended equipment is as follows:

ECT instrument and cart, preferably with integral EECand ECG monitor

6 rolling stretchers, operating room type, with wheellocks and intravenous pole holders EEG monitor (if notintegral to ECT instrument)

Defibrillator and cart

ECG machine (if not integral to ECT instrument ordefibrillator)

Oxygen tank with valve, flow meter, and positivepressure bag

Tracheal suction pump and cart

Refrigerator with lock

Wheeled intravenous pole and stand

Lockable cabinet for medication and supplies

Medication cart

Emergency medication tray (not lockable)

P.143

containing

atropine (20-mL vial, 0.4 mg/mL)

diazepam (2-mL ampules, 5 mg/mL)

diphenhydramine (30-mL vial, 50 mg/mL)

epinephrine (1-mL ampules, 1 mg/mL)

levarterenol (4-mL ampules, 1 mg/mL)

lidocaine (50-mL vial, 20 mg/mL)

methylprednisolone (125-mg vial, 62.5mg/mL)

esmolol (10-mL vial, 10 mg/mL)

Laryngoscope with three sizes of blades and assortedcuffed endotracheal tubes

The recovery area should be large enough to hold at least 3stretchers, separated from each other by curtains orscreens, and have its own tracheal suction apparatus,portable positive pressure ventilation device (e.g., Ambubag), and intravenous pole.

Staffing RequirementsTo maximize safety and efficiency, ECT should be given by ateam consisting of a psychiatrist, a registered nurse, ananesthetic specialist, and a licensed practical nurse ornursing assistant. In hospitals with residency trainingprograms, a resident should be assigned to assist, as well asany additional nursing staff who accompany their patients tothe ECT unit and remain to observe them in the recoveryroom until they are ready to return to their wards.

The psychiatrist directs the treatment; his overall medicalresponsibility for the procedure is analagous to thesurgeon's role in the operating suite.

He must ascertain how clinically appropriate ECT is for eachpatient referred, weigh the treatment risks against thepotential benefits, monitor the patient's treatment response,and decide (in consultation with the primary physician) whenmaximum benefit has been obtained. He selects theanesthetic agents and their dosage, determines thetreatment electrode placement, sets and subsequentlyadjusts the treatment stimulus parameters, administers the

electrical stimulus, and observes the patient throughout thetreatment course for the occurrence of side effects.

The ECT charge nurse has a large administrativeresponsibility in coordinating the unit, as well as substantialtraditional nursing functions. These responsibilities includethe verification that patients have been appropriatelyprepared for ECT and that all necessary paperwork andlaboratory examinations are complete, including consentforms. Medical supplies for the unit must be ordered andmaintained; records must be kept of the controlledsubstances used; and intravenous solutions and injectablemedications must be prepared freshly each treatment day.Patient flow from the waiting area to the treatment room tothe recovery room must be facilitated; vital signs must berecorded before, during, and after treatment; and nursingpersonnel must be supervised in the treatment and recoveryrooms.

The anesthetic specialist induces the anesthesia with theassistance of the psychiatrist, maintains the airway,ventilates the patient, determines when it is safe to movethe patient to the recovery area, and initiates any requiredresuscitative or corrective procedures (e.g., intubation,treatment of cardiac arrhythmias). The APA Task ForceReport on ECT (2001) specifies that the anesthetic specialistshould be privileged in anesthesia by the hospital, and thatthe psychiatrist administering the ECT should only â!œrarelyâ! ! also provide the anesthesia, without specifyingthe circumstances that might require such a dualresponsibility. I fail to see any possible justification for thepsychiatrist to assume both roles, and advise stronglyagainst it, despite the fact that the treatment is soinherently safe that thousands of consecutive seizures areroutinely induced without fatality, regardless of who isstanding at the head of the stretcher (Pearlman, Loper, andTillery, 1990). The fact is, that the increasing age andconcurrent medical morbidity of patients who receive ECT,coupled with the progressively complex pharmacologicmanipulations of hemodynamic status these patientsfrequently require (Regestein and Reich, 1985; Hay, 1989;Drop and Welch, 1989), have aleady driven virtually allpsychiatrists out of the practice of administering anes thesiaduring ECT.

It is noteworthy in this regard that Slawson's (1985) reviewof ECT malpractice claims against the American Psychiatric

P.144

Association's insurance program from 1972 to 1983,revealred that of only 2 claims of wrongful death, oneoccurred in a patient who developed laryngospasm andapnea during ECT; the psychiatrist, who was administeringthe anesthesia himself, had difficulty resuscitating thepatient and the case was settled out of court for $270,000,the largest settlement during the 11 -year period studied.

The nursing assistant ensures that patients have voidedtheir bladder and are properly attired, that they do notsmoke before or chew gum during treatment, and thatdentures, jewelry, and eyeglasses have been removed to asafe place. Other duties include wheeling patients to andfrom the treatment room, applying the ECG monitoringelectrodes, and maintaining the unit in clean and orderlycondition (by check-list) for the next treatment day.

The Pretherapy WorkupAs for any procedure conducted under general anesthesia, amedical history and physical (including neurologic)examination are prerequisite. No laboratory tests are specificto ECT; the purpose of requesting routine examinations ofthe blood and urine (CBC, SMA-6, SMA-12, urinalysis), achest film, and an ECG is simply to screen for medicalconditions that may complicate the procedure so that theymay be remedied or controlled beforehand. Someexaminations that traditionally have been obtained beforeECT or suggested by some authorities are not recommendedhere for the following reasons.

1. X-ray examination of the spine . Before the introductionin the mid-1950s of succinylcholine muscle relaxationfor ECT, up to 40% of the patients receiving thistreatment experienced compression fractures of thedorsal spine (DeWald, Margolis, and Weiner, 1954). Itwas de rigueur to obtain spinal films before treatmentfor medical -legal purposes. Although the need for suchx-rays no longer exists, some facilities continue torequire them, which adds not at all to the patient'ssecurity, but very considerably to his bill.

2 . Skull X-rays. Skull films are occasionally requested to â!œrule out brain tumorâ! ! before one gives ECT. Suchfilms, however, are insensitive to intracerebral lesions,regardless of cause, and together with spine films

P.145

needlessly increase the cost of treatment and add tothe patient's lifetime radiation exposure. Brain tumorslarge enough to cause psychiatric symptoms are likelyto manifest themselves during a careful behavioralneurologic examinationâ!”any clinical suspicion of sucha neoplasm should then be investigated with one of thenewer imaging techniques (CT scan, MRI).

3. Electroencephalograms. Conversely, the high sensitivityand low specificity of the EEG render it a poorscreening tool before ECT. One quarter to one third ofmelancholies exhibit EEG abnormalities (Abrams et al.,1970; Abrams and Taylor, 1979) that are usually in theform of nonspecific slowing, which is not infrequentlyasymmetrical. Such slowing does not predict a pooroutcome with ECT (Abrams et al., 1970) and may evendisappear after the acute EEG effects of ECT havepassed.

4. Pseudocholinesterase testing. Pseudocholinesterase isthe enzyme responsible for degrading succinylcholine,the muscle relaxant used in ECT.

Absence of this enzyme is transmitted as a rare geneticabnormality affecting fewer than 1 in 3000 (Lehmannand Liddell, 1969) and is responsible, along with liverdisease, polyphosphate insecticide poisoning, anemia,malnutrition, or the administration ofanticholinesterases, for the complication of prolongedapnea after ECT (Matthew and Constan, 1964;Packman, Meyer, and Verdun, 1978). Concomitanttherapy with lithium, certain antibiotics,aminoglycosides, magnesium salts, procainamide, andquinidine (Hill, Wong, and Hodges, 1976; Packman etal., 1978) may also prolong post-ECT apnea byenhancing the neuromuscular blockade induced bysuccinylcholine. A rapid screening test forpseudocholinesterase deficiency is available (Swift andLaDu, 1966), but when a sensitive test is used toscreen for a rare disorder, virtually all positiveresponses are false and therefore unhelpful (Galen andGambino, 1975).

Medical ConsultationThe mean age of patients receiving ECT has increased inrecent years, probably as a combined result of increased

P.146

longevity, the greater risk for depressive illness in later life,and insurance coverage under Medicare. More high-riskpatients are thus receiving ECT, and medical consultation intheir management will often be sought. The consultationprocess should not be viewed simply as â!œclearance forECTâ! !; no such clearance is really possible. Rather, whatthe referring psychiatrist needs is the consultant's opinionon the nature and severity of the medical disorder inquestion, its amenability to medical management, and thedegree of risk imposed by a grand mal seizure induced undercontrolled conditions of anesthesia, muscle relaxation, andoxygenation. For the consultant to provide a valid opinion onthese matters, he must have an understanding of themedical physiology of ECT, the risks of alternate therapies(e.g., tricyclic antidepressants), and the deleterious effectsof the untreated psychiatric illness itself. Because suchconsiderations are not routinely taught in most medicalresidencies, psychiatrists should seek out consultants withsome experience in ECT or provide the requisite informationthemselves through personal discussions and citations fromthe literature. (Some internists, not fully aware of the verybrief duration of ECT anesthesia or the precise physiologicaleffects of a medically controlled seizure, overestimate thestress of the procedure and reject some patients asinappropriate risks for whom ECT is actually a moreconservative treatment than pharmacotherapy.)

ConsentExcept for rare instances of judicially ordered treatment ortreatment given in a genuine emergency to â!œpreserve lifeor limb,â! ! patients may not receive involuntary ECT anymore than they may receive involuntary surgery. The

topic is treated in greater detail in Chapter 12, but theessential elements of informed consent always include:

1. A full explanation of the procedure in layman's terms

2. A presentation of the risks and potential benefits of thetreatment offered, as well as those of alternativeavailable therapies

3. A statement that the patient may withdraw his consentat any time and for any reason

An educational videotape is useful both to orient patients

and their families to the procedures for ECT and tounambiguously document the information that has beenpresented to the patient when obtaining informed consent(Barbour and Blumenkrantz, 1978). Such videotape aids areavail able commercially (Baxter and Listen, 1986; Ries,1987).

Seizure MonitoringBefore the introduction of succinylcholine muscle relaxationfor ECT, there was seldom any doubt as to whether or not apatient had a seizure. With succinylcholine-inducedattenuation of the motor convulsion, however, it may bedifficult or impossible to verify the occurrence and durationof the induced seizure through observation of muscle activityalone. The importance of such verification derives fromstudies suggesting that it is the induced cerebral seizure,more than any other aspect of the treatment, that isresponsible for the fully developed therapeutic effect of ECT(Ottosson, 1960), and from the risks of undetected,prolonged seizures (Scott and Riddle, 1989). Although directelectrical stimulation of the brain may itself haveantidepressant properties (see below), there is little doubtthat the cerebral seizure is central to the therapeuticprocess, especially in the more sever (e.g., melancholic)forms of depession.

Seizure monitoring is particularly important during unilateralECT. In a study employing EEG monitoring, Pettinati andNilsen (1985) reported significantly more missed seizureswith unilateral than with bitemporal ECT (63% comparedwith 29%). Because such missed seizures are not alwaysdetected clinically, the authors suggest that without EEGmonitoring, patients receiving unilateral ECT mayinadvertantly receive an inadequate treatment course.

Several manifestations of the seizure may be monitoredduring ECT: The motor convulsion, the electromyogram(EMG), the EEG, and the induced tachycardia. The motorconvulsion can simply be observed and timed but isoccasionally obscured by the effect of the succinylcholine. Tocircumvent this phenomenon, Addersley and Hamilton (1953)applied a blood pressure cuff to one limb and inflated it to10 mm Hg above systolic pressure before administration ofsuccinylcholine in order to occlude this drug from themuscles distal to the cuff and allow direct observation ofunmodified

P.147muscle activity. The procedure has become a standard partof ECT monitoring, regardless of the method used to recordthe motor component. Motor seizure duration estimates thusobtained are reliable and correlate highly with thoseobtained by EEG (Fink and Johnson, 1982; Larson, Swartz,and Abrams, 1984). The EMG can be monitored during ECTdespite the presence of neuromuscular blockade (Abrams,Volavka, and Fink, 1973; Ives, Weaver, and Williams, 1976;Sorensen et al., 1981; Couture et al., 1988a,b), althoughthe duration of the EMG response varies among musclegroups (Sorenson et al., 1981) in a pattern reflecting therostral to caudal spread and subsequent abatement ofsuccinylcholine-induced muscle depolarization. Moreover,there is not necessarily a close correspondence betweenEMG estimates of motor seizure duration in a partiallydepolarized muscle and an unaffected one: In one study, anestimate of motor seizure duration obtained from a foreheadEMG was 37% greater than a visual estimate in the cuffedarm (Couture et al., 1988b).

A microprocessor-controlled technique has been invented forincorporation into an ECT device to automatically measureand display the duration and end point of the EMGmanifestations of the ECT-induced seizure (Swartz andAbrams, 199la). Swartz et al. (1994a) demonstrated thevalidity of this method in 114 randomly selected recordingsfrom treatments of 52 patients at 2 different university-affiliated hospitals, reporting an interrater reliabilitycoefficient (intraclass r) of 0.86 between the computer -determined EMG seizure duration and the independentlyrated cuffed-arm motor seizure duration. Not included inthat article were the interrater correlations between the EMGend-point ratings of the two clinicians and the computer,which were 0.83 and 0.97.

Krystal, Weiner, and Coffey (1995) confirmed the highreliability of Abrams and Swartz's (1989) computer algorithmfor automatically determining EMG seizure duration, with 2clinician-raters achieving 0.89 and 0.90 agreement (byintraclass r) , respectively, with the computer -determineddurations.

Because the EEG directly measures the brain's electricalactivity, it remains the standard against which othertechniques must be measured. Two analog methods arepresently incorporated in ECT instruments for amplifying and

P.148

presenting unprocessed EEG activity during ECT. One uses achart-drive and penwriter to record the EEG signal onpaper; the resulting record is then read by the clinician (ora computer program) as it is generated to determine theoccurrence, duration, and end-point of the induced seizure.The second method provides an auditory representation ofthe EEG signal in the form of a tone that fluctuates with thefrequency of the seizure activity and becomes constant whenthe seizure ends. This method is as reliable as the first andcorrelates highly with it (Swartz and Abrams, 1986); it hasbeen used sucessfully to detect prolonged seizures requiringtermination with ben zodiazepines (Chen, Velamati, andStewart, 1990).

Greatâ!”and sometimes unacceptableâ!”variability has beenreported by some investigators in assessing seizure lengthduring ECT using the unprocessed

single-channel EEG (Fink and Johnson, 1982; Brumback,1983; Greenberg, 1985; Rich and Black, 1985; Zorumski etal., 1986; Ries, 1985; Guze et al., 1989). Others, however,report excellent interrater reliability (Larson, Swartz, andAbrams, 1984; Pettinati and Nilsen, 1985; Swartz andAbrams, 1986; Warmflash et al., 1987; Couture et al.,1988b; Kramer et al., 1989; Gilmore et al., 1991). Theexperience of the investigators and the uniformity of theirrecording techniques doubtless contribute more to thisvariability than do any deficiencies of a particular ECTdevice.

Electroencephalographic seizure duration derived fromgraphic displays of mean integrated EEG amplitude(averaged EEG), or power spectral analysis generated bycommercially available microprocessor-based EEG analyzers,correlate highly with visual estimates of the unprocessedEEG (Couture et al., 1988a,b; Gilmore et al., 1991). Amicroprocessor-controlled seizure-monitoring technique forincorporation in an ECT device has been invented thatautomatically detects and prints on the EEG strip theoccurrence, duration, and end point of the EEGmanifestations of the ECT-induced seizure (Abrams andSwartz, 1989). In the study by Swartz et al. (1994a) citedabove, computer -determined EEG seizure endpoint ratingswere also obtained and compared with the visual ratingsmade by the 2 clinicians from the paper records. The 2clinicians achieved a high degree of interrater reliability(intraclass r = 0.98) in their visual ratings of the EEG

P.149

seizure end point, and their equally high correlations withthe computer -determined EEG end point (intraclass r - 0.98and 0.99) amply demonstrated the validity of this method.

Krystal, Weiner, and Coffey (1995) likewise reported a highdegree of agreement for EEG seizure duration between the 2clinician-raters and Abrams and Swartz's (1989) computeralgorithm: intraclass r values of 0.83 and 0.86, respectively.It was further notable that the inability of the computer todetermine the seizure end point occurred only once despitethe fact that the majority of the seizures were rated ashaving a gradual end point.

Krystal, Weiner, and Coffey (1995) reported that thepresence of artifact, poor postictal suppression, and agradual seizure end point all reduced the agreementbetween clinician and computer, although it is impossible todetermine from their study whether the computer or theclinician exhibited greater accuracy in judging the end point.In the absence of these confounding variables, however,agreement between clinician and computer was nearlyperfect: 0.99 for both raters versus the computer. Moreover,regardless of other considerations, agreement betweenclinician and computer was also nearly perfect (98%) in theimportant determination of whether a seizure lasted longerthan 24 seconds. Similar results were obtained with theAbrams-Swartz algorithm by Rosenquist et al. (1998), whoreported a 0.98 correlation (by weighted kappa) betweenvisual and computer -determined EEG endpoints.

Heart rate may also be used to estimate ECT-inducedcerebral seizure duration because the point of maximaldecrease in the rate of the ECT-induced tachycardiaâ!”whichoccurs about 10 seconds before the EEG seizure

endsâ!”is highly correlated with both EEG and motormeasures of the induced seizure (Larson, Swartz, andAbranis, 1984; Swartz and Larson, 1986). Moreover, theheart rate is a sensitive enough reflection of intracerebralseizure generalization to differentiate right unilateral fromboth bitemporal ECT (Lane et al., 1989) and left unilateralECT (Swartz et al., 1994b). Electroencephalographic seizureduration can be predicted from the durations of the ECT-induced tachycardia and the motor seizure according to aformula derived by linear regression analysis (Larson,Swartz, and Abrams, 1984):

indicating that the durations of the ECT-induced tachycardiaand motor seizure are equally predictive of EEG seizureduration. This formula provides an objective estimate ofcerebral seizure duration that may be useful in the 10% ofseizures that are characterized by an indeterminate visual orauditory EEG end point.

The particular characteristics of seizures monitored withmotor, visual, and auditory techniques are as follows: Themotor seizure is observed in the limb distal to the blood-pressure cuff and consists of an initial sudden contraction ofthe muscles during the passage of the electrical stimulusand an equally abrupt relaxation as the stimulus terminates.The tonic phase is then ushered in by a gradually increasing,sustained, tetanic contraction of the muscles (arms flexed,legs extended), lasting from 10 to 15 seconds andcharacterized by rigidity and a fine tremor. This is graduallyreplaced by the increasingly rhythmical jerking movementsof the clonic phase, starting at a frequency of 10 to 12 persecond and gradually slowing over the next 20 to 45seconds to 3 to 4 per second at seizure termination, whichalways occurs abruptly in the muscles. It is always advisableto monitor the motor seizure as well as the EEG becausetermination of clonus occasionally provides the only estimateof seizure durationâ!”that is, when the EEG end point isindeterminate or if EEG paper runs out during the seizure(Weiner et al., 1980a).

Electroencephalographic monitoring consistently reveals aprogression through a series of characteristic patterns (Fig.8-1): Build-up, hypersynchronous polyspikes during tonus,and polyspike-and-slow-wave complexes during clonus thatterminate in suppression. The approaching end of theseizure is indicated by progressive slowing of the spike-and-wave bursts of clonus. A classical seizure end point occurswhen these are abruptly replaced by electrical silence. Adistinct end point is also signaled by sudden replacement ofparoxysmal clonic activity with lower amplitude, mixedfrequencies. The auditory EEG signal fluctuates intensely andrapidly during tonus and becomes increasingly irregularduring clonus, with staccato tones that accompany eachmuscular contraction; seizure termination is typically markedby a distinct change to a nearly steady tone (Abrams andSwartz, 1989).

P.150

Figure 8-1 The 4 EEG seizure phases (diagrammatic).

Abrams, Volavka, and Fink (1973) were the first to reportthat ECT-induced paroxysmal EEG activity typically persistedafter cessation of clonus, an observation subsequentlyconfirmed by numerous investigators (Weiner, 1980b; Finkand Johnson, 1982; Miller et al., 1985; Warmflash et al.,1987; Liston et al., 1988; Couture et al., 1988b; Gilmore etal., 1991). Liston et al. (1988) measured the duration ofEEG and motor seizure activity in 24 patients receiving rightunilateral ECT. In each patient, EEG seizure durationcontinued after cessation of motor activity in a cuffed limbby a mean of 12.4 seconds, a figure close to that of 13seconds reported earlier by Warmflash et al. (1987), butsubstantially shorter than those of 25 seconds reported byCouture et al. (1988b) and 22.5 seconds reported byGilmore et al. (1991). On average, motor seizures were only76% as long as EEG seizures, a figure that is about 65% for

P.151

all published studies comparing both monitoring methods.Because cerebral seizure activity sometimes continues formany minutes after the motor component ends (Greenberg,1985; Scott and Riddle, 1989; Chen, Velamati, and Stewart,1990) and may require termination by intravenous diazepamto prevent neuronal damage, routine EEG monitoring of theinduced seizure is mandatory in addition to observing themotor seizure duration.

Because the standard bifrontal EEG cannot distinguishgeneralized seizures from the partial or focal seizures thatoccasionally occur with unilateral ECT (Welch, 1982), it isadvisable to record from contralateral frontomastoid leadswhen administering this form of treatment. However,Brumback's (1987) assertion that single-channel bi-frontalEEG recordings during ECT are incapable of differentiatingmuscle artifact from cortical potentials is wrong because itignores the close correlations observed between such EEGrecordings and the duration of the ECT-induced tachycardia,and the fact that bifrontal EEG activity typically continuesfor many seconds after all muscle activity ceases.

Electrocardiographic MonitoringAlthough ECT is one of the safest procedures that is carriedout under general anesthesia, whatever risk the treatmentdoes entail falls primarily on the cardiovascular system andspecifically on the heart (see Chapter 4). This risk primarilytakes the form of cardiac arrhythmias engendered by theabrupt and massive autonomic stimulation resulting fromboth the electrical current and the induced seizure. Althoughsuch arrhythmias rarely require treatment, an awareness oftheir presence increases the likelihood of an efficientmedical response should corrective measures be required.For this reason, and because of the increased numbers ofhigh-risk patients receiving ECT, ECG monitoring of theprocedure is mandatory (American Psychiatric Association,2001). Modern ECT devices provide ECG monitoringcapability, although any oscilloscope monitor orelectrocardiograph is suitable for this purposeâ!”all that isgenerally required is a determination of the rhythm and adisplay of the QRS complex. Should evidence of myocardialdamage appear, however, it would also be useful to havefull limb and chest lead capability.

Benign neglect is a successful strategy for handling the

P.152

majority of transient ECT-induced cardiac arrhythmias. ECTwas given for tens of thousands of consecutive treatmentswithout fatal cardiac events long before ECG monitoring orbeta-adrenergic blockade were ever dreamed of (Kolb andVogel, 1942; Impastato and Almansi, 1942; Barker andBaker, 1959); ECT is remarkably benign even in thepresence of cardiac disease (Pitts, 1982). It would be ironicif the routine introduction of sophisticated moni toringtechniques and cardioactive agents into the procedureultimately served to increase, rather than reduce, its risk.

Protection of Teeth and TongueThe direct electrical stimulation of the temporalis musclesduring ECT causes the teeth to clamp shut powerfully,stressing them and risking a tongue bite. This is preventedby inserting a mouthguard between the teeth

just before delivery of the electrical stimulus, cushioning theforce of contraction and preventing the tongue fromprotruding between the teeth. The mouthguard should permitthe flow of oxygen and be small enough to fit under anoxygen mask; reusable (sterilizable) and disposable typesare both available. Under no circumstances should a Guedel-type plastic airway be in place when the stimulus isadministered to a patient with teeth because this exposesthe incisors to the full force of the bite on an unyieldingsurface, possibly fracturing them (Pollard and O'Leary,1981; Faber, 1983). Of course, edentulous patients requireno mouthguard.

Stimulus ElectrodesThe skin, with its oily secretions, presents the mainimpediment (other than the skull) to the flow of currentduring ECT (Weaver, Williams, and Rush, 1976), and specialcare should be taken to cleanse it thoroughly beforeapplying the stimulus electrodes. Stick -on, solid-gel,disposable stimulus electrodes have generally replaced theolder method of holding jelly -coated metal disks in placewith a rubber headstrap (Kellner et al., 1997).

Unilateral ECT and HemisphericDominanceUnilateral ECT has always been administered to the righthemisphere, following the practice of early investigators

P.153

(e.g., Goldman, 1949; Lancaster, Steinert, and Frost, 1958),who believed that the reduced cognitive side effects of thismethod derived from avoiding direct electrical stimulation ofthe speech and verbal memory centers of the dominanttemporal lobe.

Although the reduced overall cognitive effects of unilateralECT relative to bitemporal ECT are likely to be based moreon the fact that only one temporal lobe is stimulated, thanon the specific side of stimulation, right unilateral placementremains the standard.

However, because left unilateral ECT works at least as wellin depression as right unilateral ECT (see Chapter 7), leftunilateral ECT should, at the very minimum, be used inpatients who have right-sided strokes or skull defects, ormusicians, artists, architects, and others who need tominimize even transitory right hemisphere dysfunction.

For almost 30 years now, a temporo-parietal positioning ofunilateral ECT treatment electrodes has been standard(d'Elia and Raotma, 1975), with the parietal electrodeplaced adjacent to the vertex on the same side of the headas the temporal electrode.

Impedance TestingModern ECT instruments use minute currents, undetectableto the patient, to test the static (skin) impedance beforetreatment. This procedure provides

information about the quality of the electrode-to -skininterface that is especially critical when administeringunilateral ECT, because this method is much more likelythan bitemporal ECT to be vitiated by reduced stimuluslevels (Sackeim et al., 1987a). The static impedance cannotbe used to estimate the dynamic impedance to thetreatment current, nor is it possible to provide a specificfigure for the static impedance that is in the desirable rangebecause different ECT devices use different test stimuli andare therefore not directly comparable. If the staticimpedance tests above the range recommended by themanufacturer, however, it should be reduced by (1)increasing the pressure on the treatment electrodes; (2)applying Pre-Tacâ„¢ conductive solution; (3) repositioningthe treatment electrodes; or (4) gently abrading the skinunder the electrodes with Skin Prepâ„¢ tape (3M, St. Paul,MN) to remove the top layer of dead cells and sebum.

P.154

The specific risk incurred in treating a patient with anexcessively high static impedance is that of skin burns. Ifthe impedance cannot be brought into the desired range bythe preceding methods, then the decision must be madewhether the risk (which is actually rather modest) outweighsthe potential benefits of treatment; it usually does not. Ifthe impedance is too low (e.g., < 100 ohms) this meansthat moisture has short-circuited the stimulus, usuallybetween two ECT electrodes that have been placed tooclosely to gether. Treatment should not be given until thiscondition is remedied be cause no current will enter thebrain under these circumstances.

Stimulus SelectionA detailed description of the nature of the electrical stimulusis provided in Chapter 6. Suffice it to say here that thebrief-pulse, square-wave stimulus is the only appropriateone for modern ECT (American Psychiatric Association,2001). All brief-pulse ECT devices now deliver stimuli thathave similar pulse configurations and maximum energylevels; these devices differ mainly in the available range ofstimulus programs and parameters, moni toringconfigurations, and computer -derived ictal and othermeasures.

Age-Based DosingAlthough the seizure threshold is multi -determined, asignificant positive correlation with the patient's age hasbeen one of the most consistent research findings in thefield of ECT. In fact, age has been found to be adeterminant of seizure threshold in every study to date,accounting for anywhere from 10% to 60% of the variancein threshold (Watterson, 1945; Shankel, Dimassimo, andWhittier, 1960; Weaver, Ives, and Williams, 1982; Sackeimet al., 1987a, 1993 , 2000; Weiner et al., 1980a; McCall etal., 1993a, 1996 , 2000; Krueger et al., 1993; Beale et al.,1994; Enns and Karvelas, 1995;

Coffey et al., 1995b; Shapira et al., 1996; Colenda andMcCall, 1996; Delva et al., 2000; Ng et al., 2000).

Moreover, there is a reported close correspondence betweendosage values obtained by age-based methods and thoseobtained via stimulus titration (Beale et al., 1994; Petridesand Fink, 1996; Gangadhar et al., 1998). In a recent

P.155

comparison of the efficacy of titration-based and age-baseddosing for bitemporal ECT, Gangadhar et al. (1998) foundage-based dosing to yield high rates of successful seizureinduction, and recommended the method for routine clinicaluse.

Full-age dosing (Abrams and Swartz, 1985c) uses a formulathat provides a stimulus dose averaging about 2.5 timesthreshold (Weiner, 1980; Sackeim et al., 1987b; Beale etal., 1994; Enns and Karvelas, 1995):

Stimulus dose (mC) = 5 Ã— age

Thus, the initial dose for a 50-year-old patient would be 250mC, which is approximated with some ECT devices by settingthe stimulus dial to a figure closest to the patient's age inyears. This dosing method is useful for uni lateral ECT,which requires the highest dosage levels of any treatmentplace ment.

Half-age dosing (Petrides and Fink, 1996) is recommendedfor bitemporal and bifrontal ECT, which achieve their clinicalresults at lower dosage levels. The formula is: stimulus dose(mC) = 2.5 Ã— age, which is approximated by setting thestimulus dial of some instruments to a figure closest to onehalf the patient's age in years. The resultant dosages forbitemporal ECT, for example, are virtually indistinguishablefrom those obtained using threshold-based dosing (Petridesand Fink, 1996).

Fixed-Stimulus DosingChapter 6 contains a full discussion of the advantages andpotential disad vantages of fixed stimulus dosing, withdetailed recommendations for clin ical practice.

Threshold-Based DosingThis subject is discussed in detail in Chapter 6. Suffice it tosay here that routine clinical use of this seizure duration-based method is hard to justify in view of a) the absence ofa demonstrated association between either seizure thresholdor duration and the response to ECT (Kellner, 2001;Abrams, in press; Swartz, 200Ib), and b) the availability ofthe much simpler dosing procedures described above thataccomplish the same end.

OxygenationBefore the introduction of succinylcholine-induced musclerelaxation for ECT hemoglobin oxygen saturations routinelyfell to levels of around 40 gm% (in the old terminology) andpatients became profoundly cyanotic and frequently lostsphincter control before the seizure terminated. Whetherthis cerebral hypoxemia contributed to the occasionaloccurrence of late (tardive) seizures as a complication ofECT is not known; however, no more than 1 or 2 suchtardive seizures have been reported since oxygenation wasintro duced along with barbiturate anesthesia andsuccinylcholine for modified ECT in the 1950s.

It is therefore standard recommended procedure to initiateoxygenation by forced ventilation as soon as the patient isunresponsive from the barbiturate and to continue ituninterrupted throughout the treatment procedure until thereturn of spontaneous respirations (American PsychiatricAssociation, 2001). In patients with chronic obstructivepulmonary disease, 100% oxygen should be replaced byroom air or an oxygen and carbon dioxide mixture in ordernot to abolish the hypoxic drive to respiration. Pulseoximetry has long been incorporated into general anesthesiatechnique, and should always be used to monitoroxygenation during ECT (American Psychiatric Association,2001). Hyperventilation with oxygen will prolong the seizure(Holmberg, 1953a; McAndrew and Hauser, 1967), aphenomenon that has occasionally been recommended toincrease seizure length for therapeutic purposes. Such aprocedure is irrational, however, because there is norelation between seizure duration and the clinical efficacy ofECT.

Administration of Anesthetic AgentsPremedication with a parenteral anticholinergic agentprevents the slowing of the heart that results from thepowerful vagal outflow that occurs during and immediatelyafter administration of the treatment stimulus. Atropine, 0.4to 1.2 mg intramuscularly, is recommended for this purpose;glycopyrrolate, 0.2 to 0.4 mg intravenously, is a lesseffective agent with no advantages over atropine.

The anesthetic and muscle-relaxant are administeredintravenously immediately before seizure induction. Theultrashort-acting barbiturate methohexital is the anestheticagent of choice for ECT. Although the findings of Pitts et al.

P.156

(1965) showed that it induces substantially fewer cardiacarrhythmias than the older thiopental, it has been disputed(Selvin, 1987; Pearlman and Richmond, 1990); methohexitalnonetheless has the advantages of a shorter sleep time andless postanesthesia confusion (Egbert and Wolfe, 1960;Osborne, Tunakan, Barmore, 1963; Woodruff et al., 1968).The recommended initial dose of methohexital is 0.75 mg/kgof body weight (Pitts, 1982), given intravenously by rapidbolus push. The dosages for subsequent

treatments should be adjusted according to the patient's

response to the first injection.

Propofol is an alternative anesthetic agent to methohexitalthat is widely used because it is associated with a smallerhemodynamic response during ECT (Dwyer et al., 1988;Rouse, 1988; Rampton et al., 1989; Villalonga et al., 1993;Avramov, Husain, and White, 1995; Geretsegger et al.,1998). Although propofol has been associated with shorterseizures when given for ECT anesthesia (Halsall, Carr, andStewart, 1988; Dwyer et al., 1988; Rouse, 1988; Ramptonet al., 1989; Boey and Lai, 1990), the reduced seizuredurations have been associated neither with a smallertherapeutic effect in comparison with methohexitalanesthesia (Mitchell et al., 1991; Malsch et al., 1994; Fearet al., 1994; Martensson et al., 1994), nor with anyreduction in seizure quality as measured by a postictalsuppression index and mean integrated ictal amplitude(Geretsegger et al., 1998).

Propofol should be considered for ECT anesthesia in patientswho have problematic preexisting hypertension ortachycardia, or who exhibit an excessive hemodynamicresponse during ECT. Indeed, Farah, McCall, and Amundson(1996) describe the successful use of propofol anesthesiaalone to control the blood pressure surge in a patient whoreceived ECT 4 months after undergoing repair of a posteriorcerebral artery cerebral aneurysm.

Succinylcholine is the muscle-relaxant of choice for ECT andis given at a dosage of 0.6 mg/kg of body weight (Pitts,1982), also by rapid intra venous bolus push.

The SeizureAlthough electrical stimulation of the brain in the absence ofa seizure has well-documented therapeutic effects in someforms of depression (e.g., Klein et al., 1999; Post et al.,

P.157

1999; George et al., 2000), evidence for a direct therapeuticrole of the electrical stimulus during ECT has probably beenobscured by the much larger effect of the induced seizure(Abrams, 1997), which is therefore generally acknowledgedto be the primary therapeutic agent of ECT.

It is desirable that a fully developed, bilateral, grand malseizure is obtained during each treatment session, with ictalcharactersitcs as described above and in Chapter 6. Becauseno objective criteria have yet been established for specifyinga minimum requisite seizure duration, there is no rationalbasis for the widely adopted practice of restimulating thepatient at a higher dose if the EEG seizure fails to last 30seconds, for example, or the motor seizure terminatesbefore 25 seconds. The American Psychiatric AssociationTask Force Report on ECT (American Psychiatric Association,2001) has finessed the issue by establishing a new low cut -off of 15 seconds by EEG or motor criteria to define an â!œabortiveâ! ! seizure (compared with their previouslyspecified low cut -off of 20-30 seconds by EEG criteria). Butthere is no more evidence to support a 15-second criterionthan there

was to support a 20-30-second criterion; as the Reportcorrectly points out, even seizures shorter than 15 secondscan have a therapeutic impact if given with a high enoughstimulus dose.

As described in Chapter 6, however, the clinicanadministering ECT has only the characteristics of the seizureto guide him in assessing its quality (Fig. 6-1). He shouldlook for a synchronous EEG seizure pattern with highamplitude relative to baseline, well-developed, polyspike andspike-and-slow-wave phases, a clear ictal end-point withpronounced postictal suppression, and a substantialtachycardia response. He can also refer to the output ofcomputer programs that have been developed for integrationinto ECT devices that automatically measure and reportfeatures of the ictal EEG that have been most closelyassociated with seizure quality.

Because this important approach to seizure quality is rarelytaught in residency training programs, many clinicians obtainadditional training at one of the several postgraduate ECTcourses offered. Electronic training devices have alsorecently become available that generate good and poorquality EEG seizure patterns for playback through an ECT

P.158

device (Swartz and Abrams, 2001).

Postictal CareThe goal of the postictal phase is primarily that ofmaintaining an adequate airway until the return ofspontaneous respirations and, eventually, alertness. Theanesthetic specialist continues forced ventilation until thepatient is breathing on his own, at which time he istransferred to a recovery area under the observation oftrained staff until awakening.

Frequency and Number of TreatmentsIn the United States, ECT has generally been given 3 timesper week, whereas in the United Kingdom twice-weeklytreatment is the rule (Pippard and Ellam, 1981). In the firstcontrolled comparison of these 2 treatment schedules,McAllister et al. (1987) found that fewer treatments wererequired with twice-weekly (mean 6.5 ECTs) than thrice-weekly (mean 8.9 ECTs) unilateral ECT to achieve the sameantidepressant effects. More recently, Lerer et al. (1995)and Shapira et al. (1998) conclusively demonstrated througha genuine versus sham ECT design that a twice-weeklyschedule for bitemporal ECT yielded the same therapeuticoutcome as a thrice-weekly schedule, but with fewercognitive side effects. This is therefore the treatmentschedule I recommend for bitemporal ECT.

It is not unusual for psychiatrists to administer 2 treatmentsper session (Swartz and Mehta, 1986), usually for conditionsthat constitute a serious threat to the patient's physicalintegrity. Delirious mania, melancholia with intense suicidalsymptoms, and catatonic stupor may justify administering

double bitemporal ECTs in a single session, spaced 1 to 2minutes apart to allow for the refractory period following aseizure. Double unilateral ECTs, on the other hand, arefrequently given in an attempt to increase the thera peuticyield of this method.

Required Number of TreatmentsThe total number of treatments administered to a patient ina single treatment course is a function of the diagnosis,rapidity of response, response to any previous course ofECT, severity of illness, and the quality of the response totreatments already received. Clinicians are readily able to

P.159

weigh these variables in practice and treat accordingly aslong as their patients improve with treatment. It is moredifficult to specify the maximum number of treatments thatshould be given to a patient who is not showing theexpected treatment response. Although 6 to 8 treatmentsachieve the desired result in the majority of melancholiesreceiving bitemporal ECT, it is not unusual to give up to 12to a patient who has all the clinical features generallyassociated with a good response but has not yet achievedone or who exhibits at least a small incrementalimprovement with each additional treatment. It is only arare melancholic patient (perhaps fewer than 1 in 20) whorequires or substantially benefits from more than 12 ECTs ina single course; when this number has been given withoutsubstantial effect and the question arises whether or not tocontinue, it is generally prudent to withhold furthertreatment for several days while observing the patient andperhaps obtaining another opinion before proceeding withadditional treatments.

Manics may require more treatments than melancholies, with8 to 12 ECTs sufficing in most cases, and only a rare patientrequiring more than 16.

Catatonics typically show an initial dramatic response to thefirst few ECTs, only to relapse if treatment is terminated atthis point. It is advisable, therefore, to continue to 6 or 8treatments in catatonia, which is not infrequently amanifestation of melancholic stupor (Abrams and Taylor,1976a).

Multiple Monitored ElectroconuulsiueTherapy (MMECT)In 1966, Blachly and Gowing introduced the novel procedureof administering multiple seizures in a single treatmentsession, while monitoring the patient's ECG and EEG. Theiraim was to accelerate the treatment course and to reducethe required number of anesthesia inductions for a course oftreatments. Because these authors measured neither thememory loss nor the therapeutic effect of their method,however, their study was uninformative. Although thetherapeutic and cognitive effects of MMECT have never beenevaluated in a controlled study, the ECG and EEG monitoringthat Blachly and Gowing (1966) pioneered has become astandard part of modern ECT.

Subsequent open clinical trials of multiple seizures per ECTtreatment session have failed to document any advantagesover the conventional rate of admininstration (White, Shea,and Jonas, 1968; Bidder and Strain, 1970; Strain andBidder, 1971; Abrams and Fink, 1972; Bridenbaugh, Drake,and O'Regan, 1972; Maletzky, 1986). Because the procedureis accompanied by a substantially increased risk ofprolonged seizures and severe post-ictal confusional states,it was designated â!œnot recommendedâ! ! in the recent APATask Force Report on ECT (2001).

Record-KeepingTwo permanent records should be kept of each treatment,one in the patient's chart and the other in the ECT unit. Thelatter should be maintained in a bound ledger thatsequentially records the essential data for each patient eachtreatment day, including the date, name, age, sex, hospitalnumber and unit, ordinal treatment number, methohexitaldose, succinylcholine dose, treatment electrode placement,stimulus setting(s), seizure type and duration, and pertinentcomments concerning future adjustments in drug dosages orstimulus or individual peculiarities (e.g., â!œslow circulationtimeâ! ! or â!œdevelops emergence deliriumâ! !). The sameinformation should be entered in the patient's chart, whichcan most conveniently be done with a rubber stamp that hasinformation headings followed by blanks to be filled in bythe attending physician or his delegate.

Maintenance ElectroconvulsiveTherapyBecause few illnesses are permanently relieved by a briefexposure to a therapeutic agent, most medical treatmentsconsist of an acute phase followed by a maintenance phase.No one prescribing antidepressants for melancholia, forexample, would consider terminating therapy immediatelyafter the patient had improved or recovered (the usualcourse of maintenance anntidepressant treatment continuesfor at least 6 months, and often indefinitely), yet this isprecisely how patients are frequently treated with ECT.Relapse rates under these circumstances (e.g., on placebo)are quite high, ranging up to 70% over 6 months;fortunately, maintenance drug therapy with lithium ortricyclic antidepressants after a successful course of ECTsubstantially reduces these relapse rates (Bourgon andKellner, 2000).

P.160

The recent report of a 60% relapse rate during post-ECTnortriptyline maintenance therapy, and an 84% relapse onplacebo (Sackeim et al., 2001a) is scarcely representative ofthe general efficacy of either ECT or maintenance drugtherapy, because 90% of the patients entering the trialinitially received a demonstrably inadequate form oftreatment: 2.5Ã— threshold unilateral ECT (Sackeim et al.,2000). Moreover, although nortriptyline is often consideredto have equal efficacy with its parent compound,amitriptyline,

this is not, in fact, the case (Barbui and Hotopf, 2001),although it is cer tainly associated with a lower side-effectprofile.

These results of Sackeim et al. (2001a,b) contrast ratherunfavorably with those reported by Lauritzen et al. (1996)who obtained post-ECT relapse rates of only 30% onimipramine, and 10% on paroxetine; indeed, the 65%relapse rate reported by these authors for placebomaintenance is hardly different from that obtained bySackeim, Devanand, and Prudic (1991) for nortriptylinemaintenance.

Maintenance ECT is an outpatient procedure for patients whohave already exhibited satisfactory improvement with aconventional course of ECT and who have previously failedor do not tolerate maintenance drug therapy (Fink et al.,1996). The goal of continued treatment is to maintain thepatient in remission by administering additional ECT at afrequency sufficient to prevent relapse without incurringcumulative memory loss. The ideal vehicles for this purposeare unilateral or bifrontal ECT, which should be given aninitial trial in the maintenance phase regardless of whichmethod originally induced remission. The advantage of thesemethods for outpatient therapy lies in their reducedcognitive side-effects relative to bitemporal ECT.

The widespread introduction and general efficacy of lithiumprophylaxis for both bipolar and unipolar illness doubtlessreduced the use of maintenance ECT, as suggested by thevirtual absence of articles on the topic between 1965 and1990. Subsequently, however, several uncontrolled reportstestify to the continued need for maintenance ECT in the10% to 15% of patients with affective disorder whose post-ECT recovery is not maintained by any type of drug therapy(Kramer, 1987, 1990 , 1999b; Loo et al., 1988, 1991; Dubin

P.161

et al., 1989; Thornton et al., 1990; Grunhaus, Pande, andHaskett, 1990; Jaffe et al., 1990a; Thienhaus, Margletta,and Bennett, 1990; Fink et al., 1996). Although favorableresults are almost invariably presented in these case reportsand retrospective chart and literature reviews, all agree thatprospective trials are sorely needed and that concurrentpharmacother apy confounds the ability to draw firmconclusions on the efficacy of the procedure.

A disturbing note is heard, however, in the mention(Kramer, 1987) of one practitioner who gave approximately2400 maintenance ECTs to a single patient whom thepractitioner alleged was â!œstill receiving them withoutproblems,â! ! a highly doubtful assertion. The potential forabuse of maintenance ECT remains an unresolved problem;maintenance ECT, like psychoanalysis, should not beinterminably prolonged. Such a practice is unwarranted andif performed with bitemporal placements may producesevere, continuous, cognitive deficits (Regestein et al.,1975). In my view, a second opinion should be soughtbefore continuing maintenance ECT for more than 1 year or12 treatments, whichever comes first.

A typical schedule for maintenance ECT provides a treatment1 week after the initial course is successfully completed, asecond in 2 weeks, a third in 3 weeks, and the fourth andsubsequent treatments at monthly intervals

for up to 6 months. Some patients may not remain well onmonthly interval maintenance ECT and will requiretreatments at 3-week intervals or, rarely, biweekly. Thislatter spacing should only be given with unilateral ECT, for 2to 3 consecutive treatments, before again attempting todecrease the seizure frequency. Maintenance ECT patientsare included in the treatment schedule together withinpatients. They should receive written instructions not tohave breakfast, and although they may come to the hospitalalone, they should leave with a responsible adult. Manypatients on maintenance ECT find that they are able to go towork later that morning, especially if they have receivedunilateral ECT or monthly interval bitemporal ECT.Regardless of who accompanies them from the hospital, eachpatient should be cleared for release that morning by aphysician or nurse. The usual records are maintained in theECT unit and in the patient's outpatient chart. Laboratorytests other than those already obtained for the originaltreatment course are unnecessary.

P.162

Ambulatory TherapyThis phrase refers to outpatient administration of the entirecourse of ECT: The patient is never hospitalized. Although itis frequently performed in the United Kingdom, ambulatoryECT remains underutilized in the United States. This balanceshould change as free-standing outpatient treatmentfacilities (e.g., Surgicenters) flourish in response touniversal cost-containment pressures on medical practice.ECT is ideally suited to the outpatient setting. It is brief,safe, and well tolerated (far more so than the variety ofendoscopic, plastic, and dental surgical procedures nowroutinely performed on outpatients). Ambulatory ECT is onlyunsuitable for patients whose illness severity andconsequent risk mandate inpatient observation and care(e.g., suicidal, agitated, or delusional melancholies;catatonics; acute hmanics).

Treatment Complications and TheirManagement

Neurological PhenomenaTransient neurologic abnormalities, including aphasias,apraxias, and agnosias, which were considered normalaccompaniments of ECT rather than complications, werenoted by early clinicians to occur during the immediatepostictal phase following bitemporal ECT (Hemphill, 1940;Kalinowsky, 1945; Gallinek, 1952b; Kane, 1963), but werenever systematically investigated. Jargon aphasia has beenreported after left -unilateral ECT (Gottlieb and Wilson,1965), as have other dysphasias (Pratt, Warrington, andHalli day, 1971; Annet, Hudson, and Turner, 1974; Clyma,1975), all generally resolving within 30 minutes aftertreatment.

The only systematic study was done by Kriss et al. (1978),who performed neurologic examinations on 29 dextralpatients before and immediately

after each of 62 left -or right-unilateral ECTs. Asymmetricalmotor responses observed during the induced seizure usuallyconsisted of more intense clonic movements of themusculature contralateral to the stimulated hemisphere,despite the induction of a generalized, bilateral seizure.Following the seizure, and before recovery of consciousness,upper limb reflexes ipsilateral to the treated hemisphere

generally returned first. Limb strength tested after thereturn of consciousness revealed upper limb weakness in80% of the observations, with a gradual return to normalover the ensuing 15 minutes. Motor and visual inattentioncontralateral to the treated side also occurred, as well ascorresponding tactile inattention. All patients receiving left -sided ECT showed signs of dysphasia (dysnomia)immediately afterward. Overall, patients took longer torespond and to open their eyes after left -compared withright-unilateral ECT, supporting claims for the major roleplayed by the dominant hemisphere in the maintenance andmanifestation of consciousness. Anosognosia (unawarenessor denial of impairment) was profound and striking afterright-unilateral ECT, even after patients had become fullyalert and cooperative.

Electroconvulsive Therapy-EmergentDyskinesiasDyskinetic movements appearing during ECT take severalforms, the most ubiquitous of which are the typical postictalchewing and lip-smacking automatisms that physicians whoadminister ECT have long observed during the postictalphase and which Liberzon et al. (1992) characterized as â!œmild bilateral orobuccolingual dyskinetic movementslasting 1 to 3 min.â! ! The other oral and limb dyskinesiasthese authors observed during the immediate postictal periodin 3 patients generally resolved within minutes, but lasted24 hours in one instance. Their report is difficult to interpretbecause 2 of their patients were already neurologicallydamaged; 1 exhibited preexisting facial grimacing andcarried a diagnosis of choreoathetotic cerebral palsy, andthe other was a Wilson's disease carrier whom childhoodmeningoencephalitis had left with hemispheric cerebralatrophy. Of longer duration were the 3 instances describedby Flaherty, Naidu, and Dysken (1984), in which ECT-emergent involuntary facial -bucco-lingual movements tookweeks or months to resolve. Although each patient had ahistory of neuroleptic drug administration, it was too far inthe past for neuroleptic withdrawal to have caused thedyskinesias. The most likely explanation for these cases isan interaction between neuroleptic-sensitized dopaminereceptors and a dopamine-receptor stimulating effect of ECT.

Different still are the ECT-emergent dyskinesias reported byDouyon et al. (1989) that materialized in patients withParkinson's disease who continued to receive levodopa

P.163

during ECT; the dyskinesias disappeared with a reduction inthe dose of levodopa. The clinical data reviewed elsewhere inthis volume suggests that Liberzon et al. (1992) are correctin their attribution

of the ECT-emergent dyskinesias to ECT-induced increases inpost synaptic dopamine receptor sensitivity.

Failed Seizures and Poor SeizureQualitySeizures may occasionally be difficult or impossible to obtaineven at maximum electrical dosage (Krystal, Dean, andWeiner, 2000), especially in older men during the latter partof their treatment course, or in patients receiving long half -life benzodiazepines. This is not the case, however, for shorthalf -life benzodiazepines such as oxazepam (Olesen, Lolk,and Christensen, 1989). Seizure quality may also diminishas the course of ECT progresses, often with a parallelreduction in seizure duration. In the past, parenteraladministration of the methylxanthine, caffeine, wasrecommended for increasing seizure duration, but therationale for this procedure is obscure because of the lack ofcorrelation between seizure duration and clinical response invirtually every study of the subject. Moreover, caffeine wasnever a reasonable choice for patients who failed to obtain aseizure at maximum device capacity because caffeine doesnot lower the seizure threshold (Shapira et al., 1987; McCallet al., 1993b). Most importantly, Rosenquist et al. (1994)examined the effects of caffeine pretreatment on severalmeasures of seizure quality and impact (posticalsuppression, ictal EEG regularity, and heart rate response),none of which differed significantly from those recorded inpatients receiving pretreatment with placebo.

Moreover, caffeine augmentation is not without risk.Tachycardia, hypertension, multifocal VPCs, bigeminy, atrialectopy, junctional rhythms, fusion beats, ventricular andsupraventricualr tachycardias, olfactory hallucinations, andcardiac arrest have all been reported with caffeinepretreatment (Acevedo and Smith, 1988; Jaffe et al., 1990b;Beale et al., 1994; Kellner and Bachman, 1992; Liebowitzand El-Mallakh, 1993; Solomons, Holliday, and Illing, 1998).To that list must now be added the risk of status epilepticus,which Solomons, Holliday, and Illing (1998) reported in apatient who had received 500 mg oral caffeine 1 hour pre-ECT. Further concern is raised by the report of Enns,

P.164

Peeling, and Sutherland (1996), who found hippocampal andstriatal damage in electroshocked rats pretreated withcaffeine, but not in control rats that received placebo. Theauthors attributed this difference to methylxanthine-inducedblockade of the neuro protective effects of endogenousadenosine.

The related methylxanthine, theophylline, also lengthensseizures (Swartz and Lewis, 1991; Stern et al., 1999) andhas caused status epilepticus in patients who were alreadyreceiving this drug at the time they received ECT (Peters,Wochos, and Peterson, 1984; Devanand et al., 1988a; Criderand Hansen-Grant, 1995). I am aware through my legalconsulting work of 3 additional, unpublished, instances ofdeath or permanent brain damage occurring under similarcircumstances.

The American Psychiatric Association Task Force on ECT,which had recommended the use of caffeine in its 2ndReport (American Psychiatric Association, 1990), withdrewthis recommendation in its most recent Report (AmericanPsychiatric Association, 2001). Although the risks of caffeineand theophylline are presumably lower in patients with apreexisting high seizure threshold (Swartz, 1997), I see noclinical justification for their continued use.

Increasing the stimulus dose over a course of ECT is thepreferred method for maintaining seizure quality at thedesired level. Before trying this simple and usually effectiveexpedient, however, it is wise to ensure that ECT stimulusparameters heve been optimized as described in Chapter 6(e.g., by selecting short-pulsewidth, long-duration stimuli).If they are not, selecting a pulsewidth no longer than 0.5ms and maximizing stimulus duration will often achieve thedesired level of quality. It is the patients who exhibit poorseizure qualityâ!”or who fail to seize at allâ!”despitestimulation at maximum device capacity that present themost difficult problem. The fastest and most effectivesolution available at the time of this writing is toimmediately restimulate at the maximum dose.

High-dose stimulus configurations have long been anavailable option on US-manufactured ECT devices sold inother countries (Abrams, 2000); however, the US Food andDrug Administration (FDA) presently limits the capacity ofdevices sold in this country to 100 joules at 220 ohmsimpedance. Krystal, Dean, Weiner et al. (2000) reported that

P.165

one out of 6 patients receiving ECT required the maximumdose. Although only 5% of all patients treated failed toobtain a therapeutic seizure at this dose, the study resultssuggest a much higher failure rate in precisely thosepatients for whom high-dose devices were first introduced:old men. Moreover, Krystal, Dean, Weiner et al. (2000) useda very conservative dosing strategy for unilateral ECT: 2.25Ã— threshold. This value is far below the presentlyrecommended minimum of 6Ã— threshold (Sackeim et al.,2000), which would certainly have resulted in a much higherfailure rate because of the large number of patients whocould not receive this dosage level due to the FDA dosagelimitations. For these reasons, several experts have calledfor an increase in the maximum-allowable dosage of ECTdevices sold in the United States (Sackeim, 1991b; Lisanbyet al., 1996; Krystal, Dean, Weiner et al., 2000; Abrams,2000).

Prolonged SeizuresAlthough no specific guidelines exist for determining when agiven seizure has become excessively long, O'Connell et al.(1988) showed that neuronal lesions could be produced inrats after as little as 10 minutes of continuous seizureactivity induced by mercaptopropionic acid, a neurotoxicchemical. The relevance of this to ECT is unclear, but sincemost ECT-induced seizures last less than 90 seconds by EEGcriteria, and no relation has been established

betweeen seizure duration and therapeutic impact, it seemsprudent to terminate seizures exceeding 120 seconds(Abrams, 1990). Intravenous diazepam, 10 to 15 mg, haslong been the drug of choice for terminating prolongedseizures, although administration of additional barbiturateanes thetic (e.g., methohexital, 25 to 50 mg) or midazolam,1 to 3 mg intrave nously, is also effective.

Most instances of status epilepticus or other paroxysmal EEGabnormalities developing during or immediately after ECTcan be attributed to some unusual aspect of the patient orthe treatment: Mental deficiency secondary to brain damageat birth (Roith, 1959); a preexisting paroxysmal EEGabnormality (Weiner et al., 1980a; Kaufman, Finstead, andKaufman, 1986); old stroke (Strain and Bidder, 1971); theadministration of MMECT (Strain and Bidder, 1971;Bridenbaugh, Drake, and O'Regan, 1972; Maletzky, 1978,1981); the coadministration of lithium (Ray, 1975; Small et

P.166

al., 1980; Weiner et al., 1980b), caffeine (Solomons,Holliday, and Illing, 1998), theophylline (Peters, Wochos,Peterson, 1984; Devanand et al., 1988a; Crider and Hansen-Grant, 1995) or trazodone (Kaufman, Finstead, andKaufman, 1986); or hyponatremia due to self-induced waterintoxication (Finlayson et al., 1989). The only instance ofECT inducing status epilepticus in a patient known to haveepilepsy occurred after the administration of MMECT (Maletzky, 1981).

It is difficult to know what to make of the study by Mayur etal. (1999) who found prolonged seizures by EEG monitoringin 16% of their patients, an incidence 3 times larger thanthat reported by Greenberg (1985), the previous record-holder, whose results had already been consideredaberrantly high (Abrams, 1997). Using custom-madeequipment, Mayur et al. (1999) recorded only the firstseizure of each ECT course, obtained during stimulustitration. First seizures are always longestâ!”which is whythey are generally excluded from studies of EEGmorphologyâ!”and most prolonged seizures occur during thefirst ECT (Glenisson et al., 1998). Even so, no such resulthas been reported before or since by any of theinvestigators that monitored the EEG in many hundreds ofpatients receiving ECT with stimulus titration. Mayur et al.(1999) also defined a prolonged seizure as > 120 sec, inaccordance with a recommendation of the Royal College ofPsychiatrists (1995). Although I support such a definition forclinical purposes, most other investigators, as well as theAmerican Psychiatric Association (2001), do not define aseizure as prolonged until it exceeds 180 seconds EEGduration. No description is provided of pre-existing riskfactors in the patients with prolonged seizures, who werealso receiving uncontrolled and undocumented concurrentdrug administration. The most recent estimates of theincidence of prolonged seizures during ECT are in the l%-2%range (Glenisson et al., 1998; Scott and McCreadie, 1999),which coincides with my experience.

Nonconvulsive Status EpilepticusOver the years, several case reports and letters have usedthe diagnosis of nonconvulsive generalized status epilepticusto characterize the prolonged

clouded state with EEG abnormalities that can occurfollowing an ECT-induced seizure (Varma and Lee, 1992;

Rao, Gangadhar, and Janakiramaiah, 1993; Crider andHansen-Grant, 1995; Grogan et al., 1995; Solomons,Holliday, and Tiling, 1998). No specific inclusion or exclusioncriteria for this entity have been provided, and there are nodiagnostic or specific EEG findings. Most instances occurredin patients with preexisting brain abnormalities or who werereceiving medications with known epileptogenic orencephalopathic properties, all of which predispose patientsto simple post ictal delirium.

Although the articles of Weiner et al. (1980a,b) are oftencited as key references on the subject, Weiner et al.(1980b) actually reports a patient who received a total of 5ECTs with only mild confusion and memory impairment, andwho first exhibited increasing confusion, disorientation, andunresponsiveness beginning 2 days after the end of the ECTcourse, after post-ECT maintenance lithium therapy had beenstarted. (This, despite the assertion in the title and textthat the syndrome occurred with â!œconcurrentâ! ! ECT-lithium treatment â!œduring the latter portion of a course ofECT.â! !) Moreover, an EEG for assessment of this syndromewas obtained for the first time 5 days post-ECT. Nowhere inthe article is the term â!œnonconvulsive status epilepticusâ! ! mentioned, nor do the authors directly or indirectly referto the elements or concept of this entity; the same holdstrue for Weiner et al. (1980a).

Until and unless a series of cases is reported defining thevalidity of the diagnosis of ECT-induced nonconvulsive statusepilepticus as an entity separate from classical statusepilepticus and postictal or drug-induced delirium, it willhave to be considered merely a variant of these latter syndromes, which generally resolve spontaneously or followingthe intravenous administration of benzodiazepines (seebelow).

Prolonged ApneaThere is no antidote for succinylcholine and no specifictreatment to reverse prolonged apnea. Assisted respirationis simply continued for as long as it takes the patient's ownlimited pseudocholinesterase activity to metabolize thesuccinylcholine (usually 30 to 60 minutes). Intubation is notrequired as long as good pulmonary exchange documentedby oximetry is achieved by face mask. If it seems that apneamay be prolonged for more than an hour, considerationshould be given to the administration of a unit of typed and

P.167

cross-matched fresh whole blood or plasma to supply anexogenous source of pseudocholinesterase (Matthew andConstan, 1964). Although it is possible to continue to usesuccinylcholine for subsequent treatments, albeit at a muchlower dose (Impastato, 1966; Hickey, O'Connor, and Donati,1987), a more rational approach is to switch to atracuriumfor further treatments, because it is not dependent on serumpseudocholinesterase for its metabolism (Stack, Abernethy,and Thacker, 1988; Kramer and Afrasiabi, 1991).

Emergence DeliriumAbout 10% of patients develop a self-limited delirium oracute confusional state during the immediate postictalphase, characterized by all of the fol lowing clinical featuresoccurring in concert:

1. Restless agitation

2. Disorientation

3. Clouded consciousness

4. Repetitive stereotyped movements

5. Impaired comprehension

6. Failure to respond to commands

7. Subsequent amnesia for the episode

The anesthesia, electrical stimulation, and seizure eachpresumably contribute their part to causing the syndrome, inwhich the dazed, restless patients mutter incoherently whilefumbling with the bedclothes, rubbing and pulling at theirskin, moaning loudly, flopping about, and even trying toclimb off the stretcher. The physical restraint required toprevent this behavior only seems to make things worse. Thedelirium generally lasts from 10 to 45 minutes untreated andresembles nothing so much as a psychomotor seizure. It isreadily terminated by intravenous benzodiazepines orbarbiturates if a vein can be found and the patient held stilllong enough. Because patients who develop emergencedelirium manifest it during more than a third of theirtreatments (Devanand, Briscoe, and Sackeim, 1989), it isstandard practice to prevent recurrent episodes byadministering diazepam, 5 to 15 mg (or midazolam, 1 to 3mg) intravenously, as soon as the induced seizure

P.168

terminates.

Devanand, Briscoe, and Sackeim (1989) were unable toidentify predictors of emergence delirium from among thevariables of age, pre-ECT agitation or excitement, number ofECTs administered, barbiturate anesthetic or succinylcholinedosage, or mean seizure duration. Among 5 patientsswitched from unilateral to bitemporal ECT for lack of atreatment response, 3 developed emergence delirium onlyafter bitemporal ECT. It is clear, however, that thesyndrome occurs with equal frequency after bitemporal, rightunilateral, and left unilateral ECT (Leechuy, Abrams, andKohlhaas, 1988; Listen and Sones, 1990).

In the rare patient who fails to respond to benzodiazepineinhibition or prophylaxis, the intravenous line can be left inplace following the treatment and a 2% solution ofmethohexital infused at a rate sufficient to prevent thedelirium from emerging. This procedure should be directlysupervised by a physician or registered nurse.

Mania and Organic EuphoriaYears ago, Kalinowsky (1945) described the emergence oforganic psychotic states during the course of bitemporalECT; additional ECTs would typically

attenuate the syndrome, which might then transientlyreappear during the post-ECT convalescence. As notedelsewhere in this volume, Fink and Kahn (1961) described aeuphoric -hypomanic response to ECT that they consideredhighly favorable. Devanand et al. (1988b) described 3patients who developed maniform states while undergoingright unilateral or bitemporal ECT, 2 of whom exhibited noconcurrent cognitive impairment and were thereforediagnosed as having true hypomanic or manic syndromes;the third exhibited significant disorientation and impairedcognitive examination scores and was designated to besuffering from an organic-euphoric state. Andrade et al.(1988c) reported that 4 out of 32 endogenous depressivesdeveloped transient, nonorganic, self-limited manicsyndromes during the course of bitemporal ECT, followed byrecovery from their depressions. A fifth patient (Andrade,Gangadhar, and Channabasavanna, 1990) fared sim ilarly atfirst, but then relapsed and required 2 additional courses ofECT, with concurrent pharmacotherapy, to achieve asustained remission.

P.169

In my experience, the occurrence of a maniform syndromeduring ECT â!”regardless of the associated cognitive statusâ!”is favorable, and an indication to withhold furthertreatment while closely observing the patient. The majoritygo on to enjoy full remission of all symptoms, includingcognitive impairment; the few who slip back into depressionor remain in a maniform state can then be treated eitherwith additional ECT or pharmacotherapyâ!” the rationale forcombining the two, however, is obscure.

Aspiration PneumonitisSince the introduction of muscle relaxants, no cases ofaspiration pneumonitis were reported until Zibrak, Jensen,and Bloomingdale (1988) described 2 patients withgastroparesisone of whom had not eaten for 12 hoursâ!”whosuffered aspiration of gastric contents during ECT, followedby adult respiratory distress syndrome. (I had a similar casein my practice of a depressed woman in her 70s who hadnot eaten for over 12 hours prior to ECT, but neverthelessvomited copious amounts of gastric contents immediatelyfollowing termination of her first seizure and developedaspiration pneumonitis that took almost 2 weeks to resolve.)The authors recommend that ECT candidates with concurrentdisorders judged to put them at high risk for gastroparesis(e.g., diabetes mellitus, hypothyroidism, amyloidosis,scleroderma) should have a careful gastrointestinal historytaken, followed by a GI series or radionucleotide gastricemptying study, if indicated. The only sure way to preventvomitingâ!”and aspiration of gastric contentsâ!”in patientswith documented gastroparesis is to remove the stomachcontents by nasogastric tube before induction of anesthesia.

Ruptured BladderIrving and Drayson (1984) reported rupture of the urinarybladder during the 10th ECT in a 74-year-old man with ahistory of prostatism. The only

other case in the literature is that of O'Brien and Morgan(1991), who described a 55-year-old man on amitriptyline,150 mg/ day, who had failed to void before treatment andwho sustained a 3-cm tear in the bladder fundus during theextremely vigorous muscular contractions of an apparentlyun modified seizure. These cases amply support the standardrecommendation for patients to void their bladder beforecoming to ECT.

P.170

Nausea or VomitingThese are infrequent after ECT and can be prevented bydimenhydrinate, 50 mg intramuscularly, given at the end ofthe seizure.

HeadacheHeadache occurs in about one third of all patients after ECTand usually responds to aspirin or, if severe, ibuprofen.Acting on the theory that electrically induced temporalismuscle spasm contributes to the development of post-ECTheadache, Swartz (personal communication) has successfullypre scribed heat and massage for this symptom.

Drug Interactions

Psychopharmacologic AgentsTo date, there are no prospective, controlled, double-blindtrials demonstrating that any psychopharmacologic agenteither augments the therapeutic impact of ECT or reducesthe total number of treatments required (AmericanPsychiatric Association, 2001). The primary rationale forcombining such medications with ECT is to prepare for thepatient's discharge by initiating combined therapy during thefinal week of the treatment course. When a patient improvesor recovers on combined therapy initiated at the outset ofthe ECT treatment course, it is never possible to determinewhich of the two agentsâ!”drug or ECTâ!”was responsiblefor the improvement. The patient is thus committed to theincreased cost and potenial risk of combined therapy atevery subsequent treatment course; in my view, the patientis better served, and future therapy more rationallyadministered, when the physician is able to determine thetherapeutic impact of ECT given alone.

NeurolepticsReserpine is contraindicated in patients receiving ECTbecause several deaths and near deaths due tocardiovascular collapse or respiratory depression have beenreported with the combination (Foster and Gayle, 1955;

Kalinowsky, 1956a; Bracha and Hess, 1956; Gaitz, Pokorny,and Mills, 1956; Bross, 1957). Although the combination ofchlorpromazine and ECT is believed to be safer (Berg,

Gabriel, and Impastato, 1959), a number of deaths and life-threatening incidents have also occurred with suchcoadministration (Weiss, 1955; Kalinowsky, 1956a,b; Gaitz,Pokorny, and Mills, 1956; Grinspoon and Greenblatt, 1963),which must therefore be considered contraindicated and, inany case, has no justification from controlled trials. If aneuroleptic must be combined with ECT (as in a manicpatient who is early in the treatment course and has not yetresponded adequately), neither fluphenazine nor haloperidolhas been associated with ECT-related hypotensive reactions,although the latter compound seems inordinatelyrepresented among neuroleptics precipitating neurolepticmalignant syndrome (NMS).

LithiumAs noted earlier, lithium prolongs the neuromuscularblockade of succinylcholine (Hill, Wong, and Hodges, 1976)and has also been implicated in the causation of acute,reversible, delirious states after ECT, variously characterizedby confusion, disorientation, lethargy, EEG slowing, andseizures (Jephcott and Kerry, 1974; Hoenig and Chaulk,1977; Mandel, Miller, and Baldessarini, 1980; Small et al.,1980; el -Mallakh, 1988; Penney et al., 1990).

The fact that some patients have safely received thiscombination (Martin and Kramer, 1982; DeQuardo andTandon, 1988a; Penney et al., 1990) or that nodemonstrable pharmacokinetic or drug-interaction factors areknown to preclude prescribing it (Rudorfer, Linnoila, andPotter, 1987), proves only that the combination is notcontraindicated.

AntidepressantsAntidepressants may be safer than lithium or neurolepticdrugs in combination with ECT, although Freeman andKendell (1980) noted that the only 2 deaths in a sample of243 patients who received ECT occurred in patients withpreexisting cardiac disease who were taking tricyclicantidepressants. Despite concerns frequently expressed overthe safety of administering ECT to patients receiving MAOIs,the literature provides no evidence for an increased risk ofsuch combined therapy (el -Ganzouri et al., 1985; Freese,1985). As already noted for all other psychopharmcologicagents, however, no therapeutic advantage accrues to thecombination.

P.171

TheophyllineAs described above in the section on prolonged seizures,several severe interactions have occurred in patientsreceiving ECT while taking theophylline for respiratory orcardiovascular conditions. Virtually all of these instanceswere in patients who were not known to be takingtheophylline by

the physician administering ECT, and did not havedemonstrably high seizure thresholds prior to the adverseinteraction. Despite the fact that the patients of Rasmussenand Zorumski (1993) were lucky enough to have receivedECT without any adverse reaction while taking theophyllinefor medical conditions, it remains my view that ECT shouldnot be administered to a patient with significant theophyllineblood levels without unequivocal evidence for a pre-existinghigh seizure threshold, and a specific medical need fortheophylline.




> Table of Contents > Chapter 9 - Technique of Electroconvulsive

Therapy: Praxis

Chapter 9

Technique of Electroconvulsive

Therapy: Praxis

The following is a specific treatment sequence that I finduseful for administering ECT. It assumes that the patienthas been properly prepared for treatment as outlined earlierand is lying on a stretcher in the treatment room, that vitalsigns have been recorded, and that all necessary staffpersonnel are in attendance.

At the beginning of the treatment day, each patient'sindividual doses of intravenous medications and salineshould be drawn up into separate syringes, labeled with thepatient's name, date, type and dose of medication, andplaced together in a small container marked with thepatient's name. One can identify each syringe with stick-on,color-coded dots (e.g., green for anticholinergic, yellow foranesthetic, red for muscle-relaxant, and white for salineflush). To further facilitate identification, use a particularsize sy ringe for each medication: 3 cc for anticholinergic,10 cc for anesthetic, 5 cc for muscle-relaxant, and 30 cc forsaline flush.

1. Apply blood pressure cuff and record baseline bloodpressure. The same cuff will later be reinflated justbefore administration of succinylcholine to block thisdrug from the muscles distal to the cuff and permitsafe observation of the unmodified seizure. For thisreason, if unilateral ECT is contemplated, the cuffshould be applied initially to the arm ipsilateral to theplacement of the unilateral treatment electrodes in

P.173

order to document that a generalized, rather than afocal contralateral, seizure has occurred (Welch, 1982).Avoid the cuffed-limb method in the presence of severeosteoporosis to obviate the possibility of unmodifiedmuscle contraction causing fracture (Levy, 1988).

2. Apply ECG electrodes. Self-stick ECG recordingelectrodes are applied precordially above and below theheart, with a third applied to the shoul der as aground. It saves time to apply them while the patientis waiting for treatment. The appropriate ECG leads arethen connected and a baseline rhythm strip obtained.

3. Apply EEG electrodes. Disposable, pregelled, stick-onelectrodes are now available in a small size especiallyfor EEG monitoring. Frontal-to -mastoid electrodes onthe same side of the head are often preferred for EEGmonitoring during ECT because they produce a high-voltage,

easily read record (despite occasional ECG artifact) andcan be placed contralateral to the treatment electrodesduring unilateral ECT to verify that a generalizedseizure has occurred. The same ground electrodeapplied for ECG monitoring can be used for the EEG.

Patients should shampoo their hair the night beforeand wash their face on treatment mornings. Carefulrubbing of the skin over the recording sites with analcohol-soaked swab followed by drying with a gauzepad is the only other preparation necessary to removethe oily residues and provide artifact-free recordings.The EEG leads are then connected; ECT instrumentswith integral EEG automatically initiate monitoring whenthe stimulus is administered at the time of treatment.

4. Start intravenous line. An intravenous line is mostconveniently started with a 21-gauge, thin-walled â!œbutterflyâ! ! needle assembly attached to a saline-filled, 20-mL syringe. When blood flows into thesyringe receptacle of the butterfly assembly, connectthe saline-containing syringe, flush the intravenous linewith a few milliliters of saline, and apply a plasticclamp to the tubing while keeping pressure on thesyringe plunger. Be sure not to start the intravenous

P.174

line in the arm with the blood pressure cuff, because itwill infiltrate when the cuff is inflated.

5. Apply treatment electrodes. For bitemporal ECT, theelectrode sites are on the bare skin over both temples.For unilateral ECT the lower site is over one templeand the upper site is on the same side of the head justadjacent to the vertex. (If treating a patient with askull defect, be sure to place the treatment electrodesfar from the defect to avoid the possibility ofexcessively high current density at the brain surface.)For bifrontal ECT, the stimulus electrodes are placed onthe forehead 5 cm above the lateral angle of eachorbit.

The electrode sites should be carefully cleansed anddried. To maximize contact, extra care must be takento part the hair and expose the scalp when applyingthe para-vertex electrode for unilateral ECT.

6. Test impedance. The static impedance reflects thequality of the skin-to -electrode contact. An impedanceof 0 suggests a short circuit between the 2 electrodes,sometimes formed by wet hair. This is corrected bydrying the hair and scalp. A high impedance (e.g., themaximum value reported by the device) can be reducedby gently abrading the skin at the electrode site withSkin Prepâ„¢ tape (3M, St. Paul, MN) or abrasive gelsuch as Omniprepâ„¢ (Weaver Co., Aurora, CO). If theimpedance remains high, treatment should only begiven if the patient is so ill that the (small) risk of skinburn is of less importance than that of the un treateddisorder.

7. Administer atropine, 0.4 to 1.2 mg intravenously, byrapid bolus push. Patients with myocardial ischemiamay benefit from a few minutes of oxygenation beforeanesthesia induction with the mask held slightly awayfrom the nose and mouth to avoid a claustrophobicresponse. Pulse

oximetry is now routinely recommended for allprocedures carried out under general anesthesia.

8. Administer methohexital. While the patient countsaloud from 1 to 100, and after determining that the

needle is still patent and in the vein by gentlyaspirating the saline syringe and observing backflow ofblood, replace the saline syringe with one containing aninitial dose of 0.75 mg/kg methohexital (about 40 to70 mg), which is then given by rapid bolus push. Themethohexital dose may require adjustment forsubsequent treatments depending on the patient'sresponse to the initial injection (e.g., substanceabusers require larger doses; patients with prolongedcirculation times may take longer to fall asleep). Assoon as the patient has stopped counting and isunresponsive to questions, the empty methohexitalsyringe is replaced by one containing an initial dose of0.6 mg/kg succinylcholine.

9. Inflate blood pressure cuff to 10 mm Hg above thesystolic pressure to occlude the succinylcholine to beadministered next from reaching the distal muscles(Fink and Johnson, 1982).

10. Administer succinylcholine by rapid bolus push. Theempty syringe is then replaced with the saline syringe,and the tubing is flushed and clamped again for lateravailability in the event that additional intra venoustherapy is required. The dose of succinylcholine mayalso re quire subsequent adjustment.

11. Insert mouthguard and administer oxygen. As soon asthe succinylcholine has been given, a mouthguard isinserted between the teeth and 100% oxygen isadministered by positive pressure and continuedthroughout the treatment (including the seizure) untilspontaneous res pirations have returned.

12. Observe muscular fasciculations of the first(depolarization) phase of succinylcholine. These willappear first in the muscles of the head, neck, andupper chest and spread to those of the trunk and limbsbefore reaching the small muscles of the feet andhands. When the fasciculations have died down in thesmall muscles of the feet (generally about 1 minuteafter the succinylcholine injection), the patient is readyto be treated. A nerve-muscle stimulator documents thedegree of relaxation; set it to provide intermittentpulses over a peripheral nerve (e.g., radial) at a rate of

P.175

1 or 2 per second. When the resultant musclecontraction response is abolished, treatment may begiven.

13. Administer the treatment stimulus. When adequatemuscle relaxation has been achieved and the stimulusset and treatment electrodes placed according to theprinciples outlined in Chapters 6 and 7, oxygenation istemporarily interrupted, the patient's head and neckare hyperextended with the jaw held tightly shut (aproperly inserted mouthguard prevents the tongue fromprotruding between the teeth), and the stimulus is administered.

14. Observe the seizure. The tonic and clonic musclecontractions observed in the cuffed limb or EMGrecording support the occurrence of a cerebral

seizure. If a clearly discernible tonic-clonic motorprogression is not apparent (and this is judged not tobe the result of an excessive dose of succinylcholine),restimulation should immediately be performed at thedevice maximum. If the EEG does not exhibitqualitative evidence of good seizure intensity andgeneralization as described in Chapter 6, restimulationshould be performed at maximum dose after waitingabout a minute.

Perhaps the most important reason to monitor the EEGduring ECT is to assure that the seizure endscompletely, because paroxysmal brain electrical activitycommonly continues after motor activity has ended.The approaching end of the seizure is heralded byprogressive slowing of the spike-and-wave bursts ofclonus; a classical seizure end point occurs when theseare abruptly replaced by electrical silence or when theparoxysmal clonic activity is suddenly replaced by loweramplitude mixed frequencies. About 10% to 15% ofECTs do not have a clear-cut end point on the EEG.The paroxysmal activity gradually wanes, blendingindistinguishably into low-amplitude postictal activity.When this occurs, both the end of the motor seizureand the sudden decline in the ECT-induced heart rateelevation can aid in estimating seizure duration. It isprudent to be sure that all seizure activity has ended

before stopping EEG monitoring.

15. Initiate routine postictal care. In medically unstablepatients or those who have experienced emergenceagitation, the intravenous line should be left in placeuntil the patient awakens, in case medications need tobe given rapidly. A full-length arm board can preservethe intravenous line; just before the patient is taken tothe recovery area, the tubing is flushed with saline andclamped, and the clamp and saline syringe are taped tothe armboard. In the recovery area, the staff shouldkeep an eye on the intravenous line. After spontaneousrespirations have returned and full ventilatory exchangehas been established, the patient is wheeled on astretcher to a recovery area to be observed by trainedstaff until alert, oriented, and able to walk withoutassistance. The patient should be turned on his sideand carefully observed for any respiratory obstructionor distress. Stertorous breathing can be alleviated byhyperextending the neck with the jaw held tightly shut;excessive secretions may require suctioning, and theappropriate device should be maintained in readiness atall times in the recovery area. A manual-assistedrespiration bag and mask (e.g., Ambu) should also beavailable.




> Table of Contents > Chapter 10 - Memory and Cognitive

Functioning after Electroconvulsive Therapy

Chapter 10

Memory and CognitiveFunctioning afterElectroconvulsive Therapy

Several aspects of cerebral functioning can be affected forvarying intervals by ECT, depending on the number,frequency, and duration of the induced seizures, theanatomical placement of the stimulating electrodes, and thestimulus wave-form, parameters, and charge. Depending onwhich particular combination of these variables they receive,patients during the immediate postictal period mayexperience confusion and neurological dysfunction thatoccasionally progresses to delirium. After the postictalconfusion has cleared, two specific types of memoryimpairment can be demonstrated: retrograde amnesia forevents preceding the seizure and anterograde amnesia(rapid forgetting) for events following it. Non-memorycognitive impairments may also be detectable in theinterictal period using various modern neuropsychologicalprocedures. Finally, patients may experience subjectivememory dysfunction that may or may not be objectivelyconfirmed. Studies of the various aspects of ECT-induceddisruption of functioning naturally divide themselves intothose conducted during the immediate postictal period beforereorientation has occurred, those conducted after fullalertness and ori entation have been reestablished, andthose conducted days, weeks, or months later.

Prior to the early 1980s, virtually all memory and cognitivestudies of ECT conducted examined the consequences ofsine-wave stimulation. Although this wave-form is still used

P.177

in areas of the world where, for social or economic reasons,the practice of ECT is not up to modern standards, there islittle to be gained today from a detailed explication of eitherthe therapeutic or side-effects of sine-wave ECT, except forstudies directly comparing it with brief-pulse ECT. For thisreason, this chapter concerns itself primarily with thememory and cognitive effects of brief-pulse ECT,summarizing only those results obtained with the sine-wavestimulus that can reasonably be generalized to modernpractice. (For the remainder of this chapter the short-handphrase â!œmemory effectsâ! ! will be used for â!œmemoryand cognitive effectsâ! ! except where the two variables areaffected differ ently.)

The earliest study of the relative roles of the electricalcharge and the induced seizure in causing the memorydisturbance of ECT is that of Ottosson (1960), whoadministered bitemporal, partial (quarter-wave) sinewave

ECT and compared the amnestic effects of a very highstimulus dose, a moderately suprathreshold dose, and thelatter dose in which the seizure was shortened bypretreatment with intravenous lidocaine (Figure 10-1). Hisstriking and influential conclusion that electrical dose, ratherthan seizure duration, determines the amnestic effects ofECT has not been fully confirmed by studies in the modernera, doubtless because, although the quarter -sine-wavestimulus lies somewhere between the full sine-wave and themod ern brief pulse, its phase-width is an order ofmagnitude larger than today's pulse-width standard of 0.5ms.

Thus, Weiner et al. (1986b) did not find a significantcorrelation between stimulus dose and autobiographicalamnesia with brief-pulse ECT; Sackeim et al. (1986b) foundno relation between electrical dosage and memory loss forbrief-pulse, right unilateral or bitemporal ECT; and Coffey etal. (1990b) found no relation between electrical dosage andeither time to reorientation or Wechsler Memory Scalescores for brief-pulse, right unilateral ECT administered atvarying stimulus doses. Conversely, Miller et al. (1985)found a significant correlation between seizure duration andthe amnestic effects of brief-pulse right unilateral ECT;Sackeim et al. (1986b) found the duration of postictal

P.178

disorientation directly related to seizure duration with brief-pulse right unilateral ECT; and Calev et al. (1991b) reporteda significant correlation between seizure duration and theduration of post ictal disorientation after brief-pulsebitemporal ECT.

An exclusive role for the electrical stimulus in causing theamnesia of ECT is also inconsistent with the substantialmemory impairment reported for non-electrical methods ofinductionâ!”pentylenetetrazol and flurothyl

(Fink, 1979)â!”and by the extensive literaturedemonstrating that the amnestic effects of bitemporal ECTcan be sharply reduced simply by repositioning thetreatment electrodes anteriorly (as in bifrontal ECT), or overthe right hemisphere (as in right unilateral ECT), an effectthat is independent of stimulus dose (e.g., Weiner et al.,1986b; Squire and Zouzounis, 1986; Lawson et al., 1990;Sackeim et al., 1993). The well-known occurrence ofmemory disturbances associated with the seizures of primarygeneralized epilepsy, and the absence of significant memoryloss with nonconvulsive brain electrical stimulation inducedby magnetic fields, are also inconsistent with Ottoson's(1960) results.

Figure 10-1 ECT-induced forgetting: seizure versus stimuls.(Adapted from Ottosson, 1960.)

Just as the unmodified term â!œECTâ! ! was unhelpful indiscussing therapeutic efficacy, so an understanding of thenature of ECT-induced disorientation, dysmnesia, andimpaired cognition is contingent on full knowledge ofstimulus wave-form and parameters, dosing method, andelectrode place ment.

Confusion and DisorientationAlthough the imprecise term confusion as generally appliedto ECT subsumes mainly disorientation, a patient recoveringconsciousness after ECT might understandably exhibitmultiform abnormalities of all aspects of thinking, feeling,and behaving, including disturbed memory, impairedcomprehension, automatic movements, a dazed facialexpression, and motor restlessness. The term disorientationis also misleading because its â!œtime, place, and personâ! !components are actually memories, some recent (e.g., ageand date) and some remote (e.g., name). True orientation,which is the ability to properly locate one's self in space andtime solely by environmental cues, has not been studiedwith regard to ECT.

Sine-Wave StudiesA highly consistent finding of the early sine-wave bitemporalECT studies that has been replicated for brief-pulse ECT isthe existence of a temporal gradient of retrograde amnesia,known as Ribot's law after the 19th-century investigator whofirst demonstrated that the susceptibility of a memory todisruption is inversely proportional to its age (Lunn andTrolle, 1949; Wilcox, 1955; Mowbray, 1959; Lancaster,Steinert, and Frost, 1958; Rochford and Williams, 1962).

Relation to Treatment ElectrodePlacementAnother almost universal finding of the early comparisons ofsine-wave bitemporal and unilateral ECT that holds up verywell for brief-pulse ECT is the finding of less postictal

P.179

confusion (shorter reorientation time) with unilateral

than with bitemporal placement (Blaurock et al., 1950;Impastato and Pacella, 1952; Thenon, 1956; Lancaster,Steinert, and Frost, 1958; Cannicott, 1962; Impastato andKarliner, 1966; Valentine, Keddie, and Dunne, 1968;Halliday et al., 1968; Sutherland, Oliver, and Knight, 1969;d'Elia, 1970; Fraser and Glass, 1980). The same holds forthe reduced short-term retrograde amnesia following sine-wave unilateral, relative to bitemporal, ECT (Lancaster,Steinert, and Frost, 1958; Cannicott and Waggoner, 1967;Val entine, Keddie, and Dunne, 1968).

Brief-Pulse StudiesIn addition to assigning patients randomly to unilateral orbitemporal electrode placements, 2 groups of investigatorsalso split assignment by stimulus type: sine-wave or brief-pulse (Valentine, Keddie, and Dunne, 1968; Daniel andCrovitz, 1986). The method of statistical analysis employedby Valentine, Keddie, and Dunne (1968) does not permit aprecise separation of the effects on postictal reorientation ofthe 2 treatment variables; however, reorientation alwaysoccurred earlier after brief-pulse than after sine-wavestimulation. Daniel and Crovitz (1986) used analysis ofvariance to separate treatment effects and found that fullorientation was regained significantly earlier after brief-pulsethan after sine-wave stimulation.

In a series of comparisons of unilateral and bitemporal ECTadministered at varying multiples of the seizure threshold,Sackeim et al. (1986b, 1993 , 2000) consistently found thatrecovery of orientation took longest with bitemporal ECTregardless of dose. However, in a study described in moredetail below, when these investigators used an ultra-brief(0.3 ms) pulse for unilateral ECT (Sackeim et al., 2001b)they unexpectedly recorded a shorter time to reorientationthan with bitemporal ECT given via a 1.5 ms pulse.

Effects of Stimulus DoseCalev et al. (1991b) examined the effects of the stimuluscharge on postictal and interictal disorientation in 37 majordepressives who received brief-pulse, bitemporal ECTadministered at 1.5Ã— threshold: none of the correlations

P.180

between stimulus charge and disorientation reachedsignificance. However, in the studies of Sackeim et al.(1993, 2000) that varied stimulus dose within treatmentelectrode placement groups, higher dosage levelsâ!”substantially higher than used by Calev et al. (1991b)â!”were strongly associated with longer times toreorientation. The authors' assertion that the extent ofdosage above the seizure thresholdâ!”rather than absolutedosage â!”accounts for their findings has little relevance toclinical practice, where higher stimulus dosages areinvariably associated with greater reorientation delaysregardless of the dosing method used.

McCall et al. (2000) also found a dosage effect on post-ECTreorientation for the brief-pulse stimulus: fixed, high-dose(403 mC) unilateral ECT

caused a greater reorientation delay than titrated low-dose(mean =136 mC) unilateral ECT.

Pulsewldth EffectsSackeim et al. (2001b) performed the only systematic studyof the cognitive effects of brief-pulse ECT administered atdifferent pulsewidths. In a comparison of the cognitiveeffects of 6Ã— threshold right unilateral with 2.5 Ã—threshold bitemporal ECT, each given with pulsewidths ofeither 1.5 ms or 0.3 ms (the latter is designated an ultra-brief pulse), these authors found that ultra-brief pulse rightunilateral ECT yielded the shortest time to recovery oforientation, and 1.5 ms pulsewidth bitemporal ECT thelongest. The times to reorientation for the two intermediateforms (ultra-brief bitemporal, and 1.5 ms right unilateral)were consistent with a greater cognitive-sparing effect ofpulsewidth than electrode placement: ultra-brief pulsebitemporal ECT was associated with faster recovery oforientation than 1.5 ms right unilateral ECT (all between-group differences were significant), the first time that anyform of unilateral ECT has exhibited greater cognitive effectsthan bitemporal ECT.

Seizure DurationIn the study of Calev et al. (1991b) described above, longer

P.181

seizures were associated with longer post-ECT disorientationtimes, a relationship that was independent of stimuluscharge.

DeliriumAlthough postictal (emergence) delirium regularly occurs inpatients receiving ECT (Fink, 1979; Abrams and Essman,1982), it has received scant attention in the literature.Sackeim et al. (1983b) described 2 patients who manifestedtransient postictal delirium after bitemporal and rightunilateral ECT (but not after left unilateral ECT), exhibitingagitation, restlessness, clouded sensorium, disorientation,and failure to respond to commands. Based on these casesand on a proposed similarity to acute confusional statesoccurring after right middle cerebral artery infarction, theseauthors hypoth esized that postictal delirium reflected aprimary disruption of right-sided cerebral systems withresultant increased neurometabolic activity.

Daniel (1985) reported the contradictory case of a patientwho developed postictal delirium after bitemporal but notafter right unilateral ECT and claimed that the syndromewas nonspecific. An initial report of postictal deliriumoccurring in a fully dextral man who received left unilateralECT (Leechuy and Abrams, 1987) suggested that it wasindeed premature to attribute this syndrome exclusively toright hemisphere mechanisms. This caveat was subsequentlyconfirmed in 2 reports that ECT-induced postictal

delirium was a random effect that occurred with equalfrequency after bi temporal, right unilateral, or leftunilateral ECT (Leechuy, Abrams, and Kohlhaas, 1988;Listen and Sones, 1990).

As discussed elsewhere in this volume, postictal delirium is ashort-lived state, rarely lasting more than 30 minutes.Prolonged delirium can also occur after ECT, usually inassociation with some unusual aspect of treatment, such aslithium co-administration or induction of multiple seizuresper treatment session; however, no specific relation tostimulus wave-form, dos age, or electrode placement hasbeen positited.

Effects on MemoryThe side effects of ECT principally responsible for itscontinued lack of full acceptance among laity andprofessionals alike are the disturbances in memoryfunctioning caused most prominently by sine-wave,bitemporal ECT. The history of ECT research has largelybeen characterized by technical modifications introduced inattempts to reduce its undesirable memory ef fects withoutattenuating its therapeutic impact.

This research spans 3 broad eras of about 20 years each.The first, extending roughly from the introduction of sine-wave, bitemporal ECT in 1938 to the introduction of rightunilateral ECT (Lancaster, Steinert, and Frost, 1958),attempted to characterize the amnestic effects of sine-wave,bitemporal ECT, often (but by no means always) using tasksand methodologies that were conceptually simple by today'sstandards. The second era, extending until the generaladoption of the brief-pulse stimulus into clinical practicearound 1980, concentrated mainly on demonstrating thedifferential effects on memory of bitemporal, right unilateral,and left unilateral sine-wave ECT, employing more rigorousmethodology (e.g., random assignment, blind assessment,and control groups) and increasingly sophisticated andprecise neuropsychological measures that had beendeveloped for the study of lateralized hemispheric processes.The third era, extending from the early 1980s to thepresent, has applied the same sophisticatedneuropsychological measures to the study of the interactionamong the brief-pulse stimulus, treatment electrodeplacement, and stimulus dose in their effects on memory.

The results of the the first era of investigation have beensummarized or reviewed by several authors (Campbell,1960; Williams, 1966; Cronholm, 1969; Dornbush, 1972;Dornbush and Williams, 1974; Fink, 1979) and will generallybe referred to here only in abbreviated form or for historicalclarification. Because the sine-wave stimulus has long beenabandoned for modern ECT, the second era of investigationis also now mainly of historical interest; severalcomprehensive reviews are available for those desiring moredetail than will be presented in these pages (e.g., Squire,1982; Abrams, 1988b; Sackeim, 1992), which will focusprimarily on the brief-pulse ECT studies of the third era.

P.182In assessing the literature on ECT-induced memory loss, adistinction must be made between learning and retention(Cronholm and Ottosson, 1961a; Harper and Wiens, 1975).This is because the depressive syndrome (specifically,melancholia), either through attentional -motivational deficitsor some more integral biological dysfunction, generallyimpairs the ability to acquire new information, and the reliefof this syndrome by ECT tends to reverse this impairment.Thus, a standard â!œmemoryâ! ! test, such as the WechslerMemory Scale (Wechsler, 1945) is variously estimated tocontain only 13% to 22% of memory-specific variance(Cannicott and Waggoner, 1967; Zung, Rogers, andKrugman, 1968; Harper and Wiens, 1975) and thereforemeasures predominantly new learning, or acquisition.

Investigators studying the effects of ECT using the WechslerMemory Scale (and other similar instruments such as theBabcock or Gresham inventories) may incorrectly concludesimply that ECT improves memory. The same holds true forpaired-associate learning tasks, also widely used to studyECT-induced amnesia. The real memory variable of interestin the context of ECT is, of course, retention: The ability torecognize, relearn, or recall (in order of increasing difficulty)previously learned material (Ottosson, 1968; Dornbush andWilliams, 1974). Impaired retention for material learnedbefore a disruptive event (in this case, ECT) constitutesretrograde amnesia; impaired retention of material learnedafter a disruptive event constitutes anterograde amnesia.Anterograde amnesia is tested under conditions in whichpatients are required to learn new material (immediatememory) and then to recognize or recall it after a timedinterval has elapsed (delayed memory).

The difference between immediate and delayed memoryconstitutes the hypothetical variable forgetting (Cronholmand Ottosson, 196la), and it is precisely differences inforgetting induced by ECT that are of primary interest toinvestigators. Naturally, forgetting (or memory decay) isinherent to all memories, and it is therefore necessary todesign studies that control for normal forgetting, or baselinedecay, when investigating the effects of ECT on retention.Unfortunately, only a few investigators (e.g., Zinkin andBirtchnell, 1968; d'Elia, 1970) have done so.

P.183

Retrograde AmnesiaStudies of this phenomenon are usually divided into those ofshort-term effectsâ!”e.g., failure to recall or recognizematerial learned immediately prior to ECTâ!”and long-termeffects on the recall or recognition of well-establishedmemories for past public and personal life events,established days, weeks, months, or years prior to ECT.

Short-Term Effects

GeneralCronholm and Lagergren (1959) tested the ability ofendogenous depressives (among others) to recall a numberlearned at varying intervals before a single

partial sine-wave bitemporal ECT; recall was better thegreater the interval between learning and ECT, clearlydemonstrating a temporal gradient of retrograde amnesiaand supporting the consolidation hypothesis of memoryformation. In a more complex design employing verbal andnonverbal paired associate tasks to test learning and thenretention, Cronholm and Molander (1961) found performance6 hours after learning to be much worse following theinterposition of ECT, consistent with a significant adverseeffect of par tial sine-wave bitemporal ECT on retention ofrecently learned material. Miller (1970) confirmed theseresults for sine-wave bitemporal ECT.

Daniel and Crovitz (1983a,b) analyzed the raw data from 9published studies of ECT-induced retrograde amnesia (mostlyafter sine-wave stimulation) and found that materialpresented up to 1 hour before ECT was better recalled thanthat presented up to 10 minutes before ECT, once again confirming Ribot's law.

Although conducted in the modern era, the study of Perettiet al. (1996) is included here because it examined theretrograde memory effects of sine-wave, bitemporal ECT.Compared to drug-treated control depressives, patients whohad received a course of 12 ECTs exhibited significantlymore retrograde amnesia for the subjective experiences ofpast melancholic epi sodes, when tested 1 week after

P.184

treatment with free-recall, cued-recall and recognition tasks.

Electrode Placement EffectsThe sine-wave and partial sine-wave ECT studies of themiddle era consistently showed greater short-termretrograde amnesia with bitemporal than with unilateral ECT(Lancaster, Steinert, and Frost, 1958; Cannicott andWaggonner, 1967; Valentine, Keddie, and Dunne, 1968;Zinkin and Birtch nell, 1968; Sutherland, Oliver, and Knight,1969; Costello et al., 1970; d'Elia, 1970; Fleminger et al.,1970b).

Sackeim et al. (1986b, 1987c) employed a variety of recalland recognition tests for words, shapes, and faces that hadto be learned 15 minutes prior to brief-pulse, low-dose (I Ã— threshold) unilateral or bitemporal ECT; patients wereretested postictally after orientation had returned.Performance of the right-unilateral group was superior tothat of the bilateral group for the recall and recognition ofwords, and the recognition of both nonsense and geometricshapes; the 2 groups performed similarly on a neutral face-recognition test. In striking contrast to the middle-erastudies reviewed by Daniel and Crovitz (1983a), there wereno cumulative retrograde impairment effects for eithergroup; indeed, the bilateral group showed a cumulativeimprovement on the nonsense shape recognition task.

In a subsequent comparison of brief-pulse right unilateraland bitemporal ECT given at 2 dosage levels: 1Ã— thresholdand 2.5Ã— threshold, Sackeim et al. (1993) confirmed thegreater retrograde amnesic effects of bitemporal ECT oneach of the variables studied above, and the lack ofdifference on a neutral -face recognition test; for affectivelyexpressive faces, however, recall was worse after bitemporalECT. In this study, although 2 retrograde

dosage effects were found that were independent ofelectrode placement (higher dosage levels adversely affectedthe recall of both nonsense shapes and affectivelyexpressive faces), electrode placement had moreâ!”andmore importantâ!”retrograde amnesic effects than dosage.

The most recent study of this series (Sackeim et al., 2000)examined the cognitive effects of brief-pulse right unilateral

ECT given at 3 dosage levels (1.5Ã—, 2.5Ã—, and 6Ã—threshold), with high-dose (2.5Ã— threshold) bitemporalECT. The addition of the 6Ã— threshold unilateral ECT groupin this study allowed for a more precise delineation ofdosage effects than in Sackeim et al. (1993). Patientsreceiving high-dose bitemporal ECT performed worse on aretrograde word recall task only when compared withunilateral ECT patients receiving the lowest dosage.Otherwise, for tests of word recall and recognition, andshape recognition, the bitemporal ECT group performedworse than any of the 3 unilateral groups (which did notdiffer among themselves). There were no differences amongany of the groups for recognition of either neutral or affect-laden faces. Overall, then, retrograde electrode placementeffects were more pronounced than dosage effects, with thenotable exception that word recall was most impaired underthe two highest dosage conditions, regardless of electrodeplacement.

Two results of this study do not lend support to the authors'assertion that the extent of dosage above thresholddetermines the cognitive side-effects of ECT. The mean timeto reorientation was longer after 1.5Ã— threshold unilateralECT than after 2.5Ã— threshold unilateral ECT (18.7 minversus 17.1 min, respectively), and retrograde amnesia forword recall and recognition was greater for the 1.5Ã—threshold than for the 2.5Ã— threshold group (39.3 versus35.7, respectively). It is difficult to imagine how worseperformance for lower, relative to higher, doses abovethreshold for the same electrode placement can beintegrated into the concept that the extent of dosage abovethreshold is the determining factor for cognitive impairment.

Effects of Stimulus Wave-FormDaniel, Weiner, and Crovitz (1983) studied retrogradeamnesia in a sample of depressives who had been randomlyassigned to right unilateral or bitemporal ECT administeredwith either brief-pulse or sine-wave stimuli. Twenty-fourhours after the fifth ECT, retrograde amnesia for eventsoccurring 30 minutes before treatment was significantlygreater with sine-wave than with brief-pulse ECT, regardlessof stimulus dose, suggesting that the reduced retrogradeamnestic effects of the brief-pulse, square-wave are

P.185

inherent to its physical properties (e.g., fast rise-time, shortpulse-width) rather than the smaller charge it deliversrealtive to the sine-wave.

Effects of Seizure DurationAlthough Daniel, Weiner, and Crovitz (1983) failed to find acorrelation between seizure duration and short-termretrograde amnesia, they did report a modest but significantcorrelation between seizure duration and retrograde amnesiafor block designs learned 20 minutes prior to brief-pulseunilateral

ECT. Sackeim et al. (1987c) found that longer seizuredurations with low-dose, right unilateral ECT (but not low-dose, bitemporal ECT) were associated with impairedperformance on 3 of 5 retrograde memory measures,supporting the suggestion made above that seizure-durationeffects on memory should be easier to detect at lowerstimulus doses.

Autobiographical (Personal) MemoryAlthough the studies just described of very recently learnedmaterial are of importance in precisely defining the natureof ECT-induced retrograde amnesia, they do not evaluateretention of memories acquired some time before ECT, andit is disturbed memory for such past events that is of majorconcern to patients, their families, and their physicians.

Sine-Wave StudiesThe first ever long-term study of retrograde amnesia forpersonal memories found that, after an average course of 17sine-wave bitemporal ECTs, every patient exhibited deficitsfor some previously reported memories; deficits rarelyoccurred in control subjects (Janis, 1950a,b). Five of theECT patients followed over a 10-to 14-week period continuedto show such deficits.

Almost 20 years elapsed before any formal attempt wasmade to study the problem again, with contradictory resultsreported from different centers. Strain et al. (1968) foundretrograde amnesia for personal memories with both sine-wave unilateral and bitemporal ECT 3 days after a treatment

P.186

course, worse for bitemporal than for unilateral ECT; 7 dayslater, both groups still showed significant impairment,without any between-treatment differences.

In contrast, Weeks, Freeman, and Kendell (1980), using thepersonal memory schedule devised by Strain et al. (1968),detected no such retrograde amnesia 1 week after a courseof treatment; moreover, there were no differences betweenECT patients and a matched no-ECT control group for thenumber of personal memory items recalled 4 or 7 monthslater. Likewise, although Williams et al. (1990) foundimpaired autobiographical memory 2 weeks after sine-wave,right unilateral ECT, this was no longer present 6 monthslater.

Shortly after bitemporal, sine-wave ECT, Squire, Slater, andMiller (1981) reported that patients recalled substantiallyfewer autobiographical facts than at baseline. Seven monthslater, there was no difference between ECT and controlsubjects for the total number of items correctly recalled.When the items were broken down by the age of theparticular memory, however, ECT patients performedsubstantially worse than controls on the more recent items(especially those pertaining to the day of admission), andwere less likely than controls to recognize omittedinformation when re minded of it. Half of the ECT patients,but none of the controls, also failed to recognize somematerial regarding remote events.

Brief-Pulse StudiesSince the sine-wave bitemporal ECT study of Janis (1950),the only unqualified claim of persistent retrograde amnesiafor long-term memories after ECT, regardless of stimuluswave-form or electrode placement, appears in theconference proceedings of a 1985 meeting at the New YorkAcademy of Sciences (Weiner et al., 1986b). These authorsemployed a personal memory questionnaire thatconcentrated on the several years before testing. Thequestionnaire was administered before ECT, immediatelyfollowing the treatment course, and 6 months later, to asample of depressed patients who had been randomlyassigned to receive unilateral or bitemporal ECT with eitherbrief-pulse or sine-wave stimulation.

Shortly after ECT, all patients except those receiving brief-pulse unilateral ECT were significantly impaired in the recallof these personal memory items, but only those who hadreceived bitemporal ECT (regardless of stimulus wave-form)still exhibited significant deficits 6 months later.

However, the results of this study cannot be considereddefinitive because it was not subject to the usual andcustomary peer-review process for establishing the validityof research methods and conclusions prior to publication.Moreover, important methodological details are omitted thatwould allow the reader to judge these points independently.For example, the authors state neither the method used todetermine stimulus parameters and dosages, nor the actualdoses administered, offering only the opaque comment thatâ!œspecific initial stimulus parameters for each device werechosen to be relatively equivalent with respect to seizurethreshold.â! ! (The reader is referred to other publishedarticles for more information, but those articles also lack thesought-for data.)

In any case, Calev, Nigal, and Shapira (1991) were unableto confirm the results reported by Weiner et al. (1986b),using the same personal memory questionnaire. In a studyof titrated-dose (1.5Ã— threshold), brief-pulse, bitemporalECT, significant impairment in personal event recall wasrecorded immediately following a course of ECT; 1 monthlater this measure had returned to the pre-ECT level, whichit then exceeded at the 6-month follow-up, results that areopposite to those reported by Weiner et al. (1986b).

The study of Sackeim et al. (1993) described above assessedautobiographical memory prior to brief-pulse bitemporal orunilateral ECT and found no decrement relative to baselineat a 2-months post-ECT follow-up examination, regardless oftreatment condition (e.g., unilateral, bitemporal, high-dose,low dose).

The subsequent article of Sackeim et al. (2000) wasconfirmatory: 2months post-ECT there was no differencebetween the effects of unilateral and bitemporal ECT onautobiographical memory, and neither group examined byitself demonstrated significant impairment relative tobaseline. The same holds true for the related article ofLisanby et al. (2000), which analyzed a somewhat different

P.187

subset of autobiographical data collected during the courseof the Sackeim et al. (2000) study: significant deficits in

autobiographical memory relative to baseline were absent 2months post-ECT, regardless of dosage or electrodeplacement.

In summary, subsequent investigators have been unable toconfirm the report of Weiner et al. (1986b) that brief-pulsebitemporal ECT is capable of causing retrograde amnesia forautobiographical events relative to baseline that is stilldetectable 6 months post-ECT. In fact, no study sinceWeiner et al. (1986b) has detected persistence of suchdeficits with any form of brief-pulse ECT even as early as 1-2 months post-treatment, regardless of stimulus dosage(Calev, Nigal, and Shapira, 1991; Sackeim et al., 1993,2000; Lisanby et al., 2000).

Memory for Public (Impersonal) Events

Sine-Wave StudiesUsing a test of memory for past television programs, Squire,Slater, and Chace (1975) found that sine-wave, bitemporalECT affected recall more for the names of programsbroadcast recently than those from the more remote past, inaccordance with Ribot's law. The memory gaps for recentlybroadcast television programs, as well as impairedrecognition and recall for public events during anoverlapping time period, gradually subsided during theweeks after treatment and were no longer detectable on re-examination 7 months post-ECT (Squire, Slater and Miller,1981). No adverse effects of unilateral sine-wave ECT weredetected for public events memory.

In their comparison of sine-wave bitemporal and rightunilateral ECT, Weeks, Freeman, and Kendell (1980) includeda task that required subjects to identify names of famous orobscure personalities from up to 30 years earlier. Althoughno significant change in scores on this task relative tobaseline was observed even 1 week after ECT, the ECTgroup as a whole scored significantly lower than no-ECTcontrols at 4 months, but not at 7 months, post-ECT.

P.188

Brief-Pulse StudiesThe follow-up study of brief-pulse bitemporal ECT by Calev,Nigal, and Shipira (1991) also included a famous eventsquestionnaire (Squire, Slater, and Miller, 1981); just as forautobiographical events, patients scored above their pre-ECTbaseline at the 6-month follow-up testing.

In the study of Sackeim et al. (2000), although a between-groups comparison 2 months post-ECT showed better recallof famous events after unilateral than after bitemporal ECT,neither group by itself demonstrated a significant decrementfrom pre-ECT baseline performance, thereby failing tosupport claims for persistent or permanent retrogradememory effects of either unilateral or bitemporal ECT,regardless of dosage.

In the separately published comparison mentioned above ofsome time-limited autobiographical and public event datacollected during the course of the Sackeim et al. (2000)study, Lisanby et al. (2000) repeatedly refer to

persistent deficits in memory for public events 2 monthspost-ECT. However, these claims are based solely onnonsignificant â!œtrendsâ! ! (p > .05). Interpreted accordingto universally accepted statistical standards, however, thedata only show that 2 months post-ECT neither personal norpublic memory performance differed significantly from pre-ECT baseline, regardless of stimulus dose or electrodeplacement.

Anterograde Amnesia

Short-Term Effects

Sine-Wave Studies

General Considerations

Our basic understanding of the anterograde effects of ECTon memory begins with the Karolinska Institute studies ofCronholm and associates on partial sine-wave bitemporalECT (Cronholm and Molander, 1957, 1961 , 1964; Cronholmand Blomquist, 1959; Cronholm and Ottosson, 1960, 1961a).

Utilizing a procedure requiring patients to recall verbal and

P.189

nonverbal materials immediately after presentation(learning) and then again after a delay of 3 hours(retention) these investigators found that endogenousdepression was associated with an impairment of learningability, but not retention, and that the learning impairmentwas reversed by a course of ECT, which, in turn, impairedretention. This retention deficit was no longer de tectable 1month later.

In an equally sophisticated study, Korin, Fink, andKwalwasser (1956) examined the ability of patients over acourse of 12 sine-wave, bitemporal ECTs to learn and thenrecall word lists after either an interpolated nonsense-syllable learning task or a 10-minute rest period filled by anundemanding attentional task. Compared with untreatedcontrols, patients showed sharp and significant retentiondecrements throughout the treatment course under bothdelayed-recall conditions, but more markedly so after theinterpolated learning task. Scores under both conditionsreturned to baseline 1 week post-ECT and surpassed it 2weeks later.

The severity of the anterograde amnesia induced by sine-wave, bitemporal ECT is a function of how long ECT precedesthe new learning, and the length and contents of theinterposed retention interval. Retention is better for materiallearned 3 hours after ECT than for that learned 1 hour afterECT (Squire, Slater, and Chace, 1976), and short retentionintervals are associated with better performance than longerones (Squire and Miller, 1974). However, when learningoccurs 6 to 10 hours after sine-wave bitemporal ECT,immediate retention can be normal, but delayed retentiontested 24 hours later can be markedly impaired (Squire andChace, 1975). Thus, patients who have recovered the abilityto retain newly learned material and therefore appearnormal on casual observation, may actually be markedly

amnestic at the time as demonstrated by their inability toreproduce that material the next day (Squire, 1982).

In their study of brief-pulse bitemporal ECT, Calev, Nigal,and Shapira (1991c) found anterograde verbal andvisuospatial tasks (paired associates learning andrecognition, complex figure copying and reproduction) to be

impaired shortly after the treatment course, relative tobaseline.

Electrode Placement Effects

Often included in reviews of ECT-induced anterogradeamnesia (e.g., Price, 1982a; Daniel and Crovitz, 1983b), butactually more appropriately considered studies of learning(i.e., non-memory cognition) are the many investigations ofthe effects of sine-wave unilateral or bitemporal ECT thatemploy the Wechsler Memory Scale (or similar memoryinventories), or paired-associate learning tasks without adelayed-recall condition (Martin et al., 1965; Zamora andKaelbling, 1965; Cohen et al., 1968; Levy, 1968;Sutherland, Oliver, and Knight, 1969; Cronin et al., 1970;Fleminger et al., 1970; Fromholt, Christensen, andStromgren, 1973). Without exception, these studies showgreater impairment in verbal learning after bitemporal or leftunilateral ECT than after right unilateral ECT. Moreover,several studies using these and other verbal tasks actuallyfound an improvement in verbal learning after rightunilateral ECT (Martin et al., 1965; Zamora and Kaelbling,1965; Sutherland, Oliver, and Knight, 1969; Costello et al.,1970; Stromgren et al., 1976).

Most studies specifically examining anterograde amnesiaafter sine-wave stimulation found greater impairment ofverbal memory with bitemporal than with right unilateralECT (Zinkin and Birtchnell, 1968; d'Elia, 1970; Dornbush,Abrams, and Fink, 1971; Squire and Slater, 1978; Robertsonand Inglis, 1978; Weeks et al., 1980). In contrast,nonverbal performance differences were usually not found(Dornbush, Abrams and Fink, 1971; Robertson and Inglis,1978; Ashton and Hess, 1976). When a left -unilateral groupwas included, verbal memory was more impaired than afterbitemporal ECT (Halliday et al., 1968). Several studies alsoshowed that right unilateral ECT improved nonverbal taskperformance relative to bitemporal ECT (Dornbush, Abramsand Fink, 1971; Squire and Slater, 1978), while othersshowed that right unilateral ECT improved verbalperformance relative to bitemporal ECT (Martin et al., 1965;Zamora and Kaelbling, 1965; Suther land et al., 1969;Costello et al., 1970; Stromgren et al., 1976).

The demonstration by Horan, Ashton, and Minto (1980) that

P.190

sine-wave right unilateral, but not bitemporal, ECT,significantly improved performance on a sequentialprocessing task, led the authors to conclude that rightunilateral ECT suppression of right hemisphere interferenceallowed the left hemisphere's sequential processing mode tooperate more efficiently. Such reasoning can help explain thereports cited above that right unilateral ECT improves verbalmemory.

Brief-Pulse Studies

General Considerations

Although Sackeim et al. (1986b) included measures ofdelayed verbal, figural, and facial recall to test anterograde

effects of low-dose (I Ã— threshold) brief-pulse ECT, theycombined unilateral and bitemporal groups for analysis; anormal control group was included for comparison. Beforethe first ECT, after the seventh ECT, and 4 days after thelast ECT of their course, patients were tested for learningand delayed recall. Patients performed significantly worsethan controls before ECT on immediate recognition memory(learning), but did not exhibit a greater rate of forgettingover the 4-hour delay, confirming the work of earlierinvestigators. Although immediate memory was reduced afterthe seventh ECT, this effect had disappeared by 4 days afterthe treatment course. In contrast, delayed memory scoreswere reduced following both the seventh and finaltreatments, likewise supporting the claim of earlier work ersthat ECT had different effects on the acquisition andretention of infor mation.

Electrode Placement Effects

The study of Sackeim et al. (1993) included paired-associateword and face recognition tests administered after 7 ECTs,and after the treatment course: each test provided animmediate learning score and, 4 hours later, a recall score;a recall score after a 2-hour delay was also obtained for aselective reminding task. Bitemporal ECT regardless of dosecaused greater impairment than right unilateral ECT on bothlearning and recognition of verbal paired associates after theseventh and final treatments.

P.191

Several measures of anterograde learning and memoryeffects of ECT were also included in Sackeim et al. (2000),including complex figure reproduction, a variety of measuresof learning and retention of verbal, facial, and pictorialpaired-associates, and recall measures on a selectivereminding task. After the treatment course, bitemporal ECTcaused greater anterograde performance decrements thanunilateral ECT (regardless of dosage condition) on bothdelayed recall aspects of the selective reminding test, andon the delayed paired-word and paired-picture recallportions of the Randt An terograde Memory test.

Dosage Effects

Dosage effects separate from the effects of electrodeplacement were minimally expressed in the data of Sackeimet al. (1993): the expected greater adverse cognitive effectsof higher, relative to lower, doses did not materialize. Asnoted above in the discussion of the time-to -reorientationdata presented in Sackeim et al. (2000), such a resultprovides little support for the claim that the extent ofdosage above threshold deter mines the cognitive sideeffects of ECT.

Dosage effects were more prominent in Sackeim et al.(2000), doubtless because of the higher and more varieddosage levels administered. Thus, adverse anterogradeeffects of dosage, regardless of electrode placement, werefound for delayed reproduction of a complex figure, anddelayed story recall: the high-dose unilateral and high-dosebitemporal groups both scored worse than the remaininggroups. However, the absence of significant dosage effectsamong the low-, moderate-, and high-dose unilateral groupson several

important anterograde tasks (e.g., delayed word recall,delayed picture recall, delayed free recall and reacquisitionon the selective reminding test) falsifies the proposed directrelation between the extent of dosage above threshold andcognitive impairment (which is, after all, of potential clinicalimportance primarily for unilateral ECT). Failure to find theexpected dosage effect is the more striking in this studybecause of the 4-fold difference between the low-and high-dose unilateral groups relative to the seizure threshold, and

because the cognitive tests employed were sufficiently sensitive to detect electrode placement effects on the samemeasures.

Effects of Seizure Duration

The study of Miller et al. (1985) of brief-pulse rightunilateral ECT described earlier in the retrograde amnesiasection also included a delayed recall condition in whichpatients learned verbal and nonverbal material 4 hours post-ECT and were tested for recall 20 hours later. Anterogradeamnesia for block designs, but not for verbal paired-associates, significantly correlated with seizure duration: thelonger the seizure, the greater the forgetting.

Long-Term Effects

Sine-Wave StudiesSquire and Chace (1975) found no objective evidence forpersistent anterograde memory effects on any of the teststhey performed on patients who had received courses ofsine-wave, right unilateral or bitemporal ECT 6 to 9 monthsearlier. Likewise, the impairment in delayed recall found byWeeks et al. (1980) shortly after sine-wave, bitemporal, butnot unilateral, ECT was no longer detectable either 4 or 7months post-ECT.

Brief-Pulse StudiesThe impairment in verbal paired-associate recall found, byCalev et al. (1991c) shortly after bitemporal, brief-pulse ECTand 1 month later, was no longer present 6 months post-ECT. Similarly, none of the anterograde effects reported bySackeim et al. (1993, 2000) shortly after either unilateral orbitemporal brief-pulse ECT could be detected 2 months later,regardless of dosage condition.

Thus, the literature provides no evidence that any form ofECT is ca pable of producing long-term or persistentanterograde amnesia.

Nonmemory Cognitive Effects

P.192

Short-TermNeuropsychological issues of lateralized hemisphericspecialization of function that were raised by theintroduction of unilateral ECT stimulated most of the studiesof ECT-induced nonmemory cognitive impairment reviewed

in this section. Earlier studies have been reviewed elsewhere(e.g., Campbell, 1960; Fink, 1979; Price, 1982b) and aredifficult to assess because they were not blind, oftencontained mixed diagnostic groups, and frequently testedpatients after much longer courses of ECT (e.g., 20 or more)than would be given today. In general, however,impairments in various perceptual and psychomotorperformance tasks were observed immediately after sine-wave, bitemporal ECT, recovering to pretreatment levels orbetter by about 2 weeks after the last seizure.

Subsequent sine-wave studies comparing right unilateral andbitemporal placements have generally reported nonmemorycognitive functions to improve or remain unchanged afterECT (McAndrew, Berkey, and Matthews, 1967; Small et al.,1972 , 1973; Annettt, Hudson, and Turner, 1974; Weeks,Freeman, and Kendell, 1980; Taylor and Abrams, 1985;Williams et al., 1990). When left unilateral ECT is includedfor study, impaired verbal per formance is typically observed(Annett, Hudson, and Turner, 1974; Kronfol et al., 1978).

In general, the ability to learn new material increasesremarkably throughout a course of sine-wave bitemporal ECT(Thorpe, 1959), as does the ability to reproduce a complexvisual figure (Rossi et al., 1990). Scores on a brief, mostlynonmemory, cognitive screening test improved substantiallyshortly after ultra-brief right unilateral ECT, compared withsubstantial worsening after the other 3 methods: ultra-briefpulse bitemporal ECT, and 1.5 ms pulsewidth unilateral andbitemporal ECT (Sackeim et al., 2001b).

Long-Term Follow-Up StudiesWeeks, Freeman, and Kendell (1980) found no evidence fornonmemory cognitive inpairment 1 week after a course ofsine-wave right unilateral or bitemporal ECT; indeed,patients improved significantly on certain tasks. Comparedwith the performance of non-ECT treated depressives and

P.193

with matched normal controls, ECT patients also showed noimpairment at 4 or 7 months posttreatment.

A subsequent long-term follow-up study after sine-waveunilateral or bitemporal ECT, using an extensive test batteryof nonmemory cognitive functions, confirmed and extendedthese results (Abrams and Taylor, 1985). Patients weretested before ECT and 1 day, 30 days, 6 months, 1 year,and 2 years after the treatment course. Compared with age-matched normal controls, ECT patients exhibited significantlymore global cognitive impairment at baseline and 1 day after6 ECTs, regardless of electrode placement, but wereindistinguishable from controls at all follow-up intervals.Within the ECT group, global cognitive scores werenonsignificantly improved from baseline levels the day afterthe treatment course, with further improvements observedduring the follow-up period: 1 to 2 years later globalnonmemory cognition was substantially better than beforeECT.

Calev, Nigal, and Shapira (1991) included 3 nonmemorycognitive measures in their brief-pulse, bitemporal ECTstudy (complex figure copying,

paired-associates learning, and the nonmemory items of theMini-Mental State examination), none of which changedsignificantly after ECT or during the 6-month follow-upperiod.

Intelligence QuotientIntelligence quotient (IQ) tests are mixed neuropsychologicaltest batteries that contain only modest numbers of memory-related items. The Wechsler Memory Scale, although notspecifically an IQ test, correlates very highly with theWechsler-Bellevue and Wechsler Adult Intelligence Scales(WAIS), so it should come as no surprise that the olderliterature shows that bitemporal ECT does not impair (andmay improve) performance on these and similar instruments(Huston and Strother, 1948; Fisher, 1949; Stieper, Williams,and Duncan, 1951; Foulds, 1952; Summerskill, Seeman, andMeals, 1952). Squire et al. (1975) found no difference in theverbal IQ and arith metic subtests of the WAIS immediatelyafter a course of sine-wave, bitem poral ECT.

P.194

Effects of Subconvulsive StimulationBecause subconvulsive stimulation involves passing anelectric current through the brain, it is reasonable to predictthat it will have adverse memory consequences. Cronholmand Ottosson (1961b) examined this possibility in a study ofthe effects on retrograde amnesia of three different forms ofbitemporal ECT: (1) conventional suprathreshold seizureinduction; (2) conventional suprathreshold seizure inductionfollowed by 1 minute of weak subconvulsive stimulation; and(3) conventional suprathreshold seizure induction followed by1 minute of strong subconvulsive stimulation. The amount offorgetting (immediate recall minus delayed recall) wasevaluated before and 1 hour after ECT using a variety ofmemory tasks. Compared with simple seizure induction, theaddition of weak subconvulsive stimulation increased theforgetting score by 24%, and the addition of strongsubconvulsive stimulation increased the forgetting score by71%, thus dem onstrating a graded effect of prolongedsubconvulsive stimulation in aggra vating the retrogradeamnesia consequent to a single ECT.

In view of the American Psychiatric Association Task Force(2001) recommendation concerning stimulus titration viamultiple subconvulsive stimulations, and the report (Farahand McCall, 1993) that about 40% of US practitionersroutinely employ this method, it is important to know howthis procedure affects memory and orientation. Theexpectation based on Cronholm and Ottosson's (1961b)article is that multiple subconvulsive stim uli administered inthe same treatment session, at least via bitemporal electrode placement, should have a dysmnesic effect.

It is therefore disappointing that the only modern article tostudy the cognitive effects of subconvulsive stimulation didnot examine the stimulus titration procedure per se, butlimited its investigations to single subconvulsive stimuli thatchanced to occur later on in treatment (Prudic et al., 1994).In this study, which compared treatment sessions in whichgeneralized seizures were preceded by a single subconvulsivestimulation with those in which only single, convulsivestimulations were administered, no adverse cognitiveconsequences of single unilateral or bitemporal

subconvulsive stimulations were found.

Comparison with AntidepressantDrugsThe adverse memory effects of tricyclic antidepressants iswell-estabished (McElroy Keck, and Friedman, 1995; Settle,1998). In an open clinical trial, Calev et al. (1989)compared the anterograde and retrograde amnestic effectsof 3 weeks of imipramine, 200 mg/day, with those of acourse of 7 brief-pulse, bitemporal ECTs administered at1.5Ã— threshold. Both treatment methods produced similarand significant deficits in performance on the verbalanterograde memory task of paired-associates retention, butonly ECT impaired retrograde memory for autobiographicaland public events. Neither treatment induced deficits inimmediate memory span (digit recall) or visual retention(complex figure reproduction).

Subjective Memory DysfunctionThere is no doubt that the subjective experience ofdisordered memory after ECT has been the driving forcebehind the anti-ECT movement; had no one ever experiencedmemory disturbance after ECT, there would be no anti-ECTmovement today.

Because claims of subjective memory disturbance are justthatâ!”sub-jectiveâ!”they are not amenable to scientifictesting, and must be taken on faith alone. All we can do asscientists is to catalog the complaints, examine theirpattern, andâ!”so far as possibleâ!”determine whether theyshare some underlying commonality.

We are fortunate in this regard that there exists asubstantial body of literature on the subjective experience ofECT-induced memory dysfunction, the most recent studies ofwhich employ subjective memory questionnaires obtainedfrom those who have received ECT. Most of these studieshave also included objective assessments of memoryfunction and depressed state, thereby providing theopportunity for the investigators to examine the relationshipamong these 3 variables. The most strikingâ!”and clinically,by far the most importantâ!”finding of every one of thesestudies is that the vast majority of patients rate their

P.195

memory as either unchanged, or far more commonly,distinctly improved, shortly following a course of ECT(Shellenberger et al., 1982;

Pettinati and Rosenberg, 1984; Weiner et al., 1986b; Matteset al., 1990; Calev, Nigal, and Shapira, 1991; McCall et al.,1995; Coleman et al., 1996; McCall et al., 2000; Sackeim etal., 2000). Most investigators also find that a subjectivesense of improved memory function correlates significantlywith improvement in the depressed state, but not withobjective assessments of memory function or any aspect ofthe stimulus.

The studies of subjective memory after ECT readily dividethemselves into an earlier group that examined the effectsof sine-wave, bitemporal ECT, usually at much higherstimulus doses than used today, and a more recent groupthat studied the effects of brief-pulse unilateral andbitemporal ECT, typically at dosages only a fraction of thoseused in the early studies.

The Early Studies (1963-1983)In one of the Cronholm and Ottosson (1963a) studies ofimmediate and delayed recall and retention after partialsine-wave bilateral ECT, patients were asked 1 week after acourse of treatment whether they felt their memory wasworse, unchanged, or better. The experience of relief fromdepression was associated both with a subjective sense ofimproved memory, and with improved performance onlearning (but not retention) tasks, and verbal (but notvisuospatial) tasks. These results suggest that intactlearning and verbal performance abilities may be central tothe subjective experience of an intact memory, but thatability to retain information is not.

Although, as noted earlier, Squire and Chace (1975) couldnot detect objective evidence of impaired learning,retention, or remote memory 6 to 9 months after sine-waveright unilateral or bitemporal ECT, 63% of the bitemporallytreated sample complained of impaired memory since theirtreatment, compared with 30% of the right unilateral group(and 17% of the no-ECT controls). The authors interpretedthe patients' subjective experiences as a psychologicalphenomenon related to the patients' expectation, based on

P.196

the intial dysmnesic effects of ECT, that they might sufferlong-term memory difficulties.

Subsequently, Squire, Wetzel, and Slater (1979) developed asubjective memory questionnnaire that required subjects torate themselves on a continuum from â!œworse than everbeforeâ! ! to â!œbetter than ever beforeâ! ! on 18 itemsreflecting recognition, recall, retention, comprehension,concentration, and alertness. The phrasing of theinstructions makes it clear that the reference time intervalmust precede the onset of illness. These authors foundpersistent complaints of memory impairment in somepatients as long as 6 months after an average course of 11sine-wave bitemporal ECTs. These complaints closelyresembled those occurring 1 week after treatment but weresharply different qualitatively from those observed beforeECT, suggesting to the authors that the patient's experienceof altered memory 6 months after bitemporal ECT could notbe attributed solely to illness vari ables (e.g., depressedmood, low self-esteem).

Freeman, Weeks, and Kendell (1980) examined anonsystematically obtained sample of patients whocomplained of persistent cognitive dysfunction a mean of 10years after having had a course of ECT, most with sine-wavebitemporal technique. These patients generally performed aswell as no-ECT controls on most subtests of a battery of 19memory and nonmemory cognitive tasks but weresignificantly impaired on several tasks, a few even scoring inthe organic impairment range. The authors concluded that asmall subgroup of patients receiving ECT might sufferpermanent memory impairment.

In a 3-year follow-up study of the sample of Squire, Wetzel,and Slater (1979), Squire and Slater (1983) found thatpatients who had received sine-wave bitemporal ECTgenerally felt that they had difficulty remembering eventsfrom about 6 months before ECT to 2 months afterward. Infact, 50% of the patients who had received sine-wavebitemporal ECT felt that their memory had never returned tonormal. Again, the pattern of memory complaints resembledmuch more closely that reported 1 week after ECT, whenpatients were still amnestic, than the pattern beforetreatment, when they were depressed. The significance of

P.197

the patients' subjective estimate that ECT induced 6 months'retrograde amnesia is doubtful because they reported nearlythe same phenomenon (5 months' retrograde amnesia) whenexamined before ECT 3 years earlier. Right unilateral sine-wave ECT was generally not associated with the subjectiveexperience of memory impairment.

It is worth noting at this point that, although sine-wavebitemporal ECT is unsurpassed in its ability to relievedepression, a number of patients who have received thisform of treatment nevertheless report a subjective sense ofmemory impairment.

The Recent Studies (1984-1996)The study of Weiner et al. (1986b), which Squirecoauthored, also included a subjective memoryquestionnaire. Overall, patients rated their memory asimproved shortly following ECT, which correlated withimprovement in their depression scores, but not with any ofthe objective memory test results, treatment electrodeplacement, or stimulus wave-form.

In the negative study of Mattes et al. (1990) on the effectsof vasopressin on the memory disturbance of bitemporalbrief pulse ECT, the authors observed improvement inpatients' self-ratings of memory function.

The brief-pulse, bitemporal ECT study of Calev, Nigal, andShapira (1991c) included a subjective memory questionnaire.A significant subjective feeling of impaired memory wasreported 1 month post-ECT, but not immediately post-ECT or6 months later. An item analysis of the questionnairerevealed that no individual item was specifically related to asubjective perception of memory impairment, nor was such aperception associated with objective performance on any ofthe memory tasks administered. In contrast to Weiner et al.(1986b), no significant association was detected in thisstudy

between improvement in depression and the subjectiveimpression of im proved memory.

The patients of Sackeim et al. (1993) reported markedsubjective memory improvement that was related to

improvement in depressive symptoms. Even more striking,because of the much higher stimulus doses used are theobservations of McCall et al. (1995, 2000). In the firststudy, patients reported significantly fewer problems insubjective memory after, rather than before, high-dose (403mC) brief-pulse, unilateral ECT; in the second study, usingthe same stimulus dose, patients reported their memory tobe un changed.

Coleman et al. (1996) obtained subjective memoryquestionnaires from patients receiving unilateral orbitemporal ECT administered at either 1Ã— or 2.5Ã—threshold (from the sample of Sackeim et al., 1993).Patients receiving all 4 forms of treatment reported markedimprovent in memory function shortly after the course ofECT and 2 months later (at which time there was also nodifference between the patient's memory self-ratings andthose of normal controls). After accounting for the varianceattributable to clinical improvement (which correlated with asubjective sense of improved memory), the direction of thefinding remained unchanged, but higher dosages andbitemporal placement were both associated with less markedsubjective improvement.

The study of Sackeim et al. (2000) tabulates a remarkableeffect of dosage level and electrode placement on subjectivememory ratings during the week following ECT. All groupsshowed improved self-ratings relative to baseline, but thereally striking result is the stepwise and very substantialprogression in the degree of self-rated memory improvementas the cognitive impact of the dosage-electrode placementcombinations moves from lowest to highest. Thus, subjectivememory scores improved 25%, 33%, 53%, and 71% in thelow-dose unilateral, moderate-dose unilateral, high-doseunilateral, and high-dose bilateral groups, respectively. Inmy view, no more counterintuitive finding exists in theannals of ECT research. Of course, as in Coleman et al.'s(1996) report, it is likely that improvement in depressedstate accounts for a portion of this variance; however, itseems unlikely to be entirely determinant because, althoughthe high-dose unilateral and bitemporal groups, forexample, show identical degress of clinical response, high-dose bitemporal patients rated their memory as one-thirdmore im proved than did high-dose unilateral patients.

P.198

Despite the fact that objective memory test scores are oftenreduced during the immediate post-ECT period, patients'subjective assessments that their memory works better thanever before may also be accurate. Most of the memorytesting performed in ECT studies is designed to assessmemory contentâ!”the retention and retrieval of specificmemoriesâ!”whereas patients are more likely responding toimprovement in memory functionâ!”including especially theability to learn new materialâ!”which relies heavily on cognitive variables known to improve with ECT: concentration,attention, and motivation.

In view of the multiform phenomenon of ECS-inducedneurogenesis reported in rodents, which is associated withreversal of atrophy of stress-vulnerable neurons, protectionfrom further damage, and increased neuronal synapticstrength, survival, and growth (Duman and Vaidya, 1998),similar neuronal effects might occur in depressed patientswho receive ECT, many of whom already exhibit MRI signsand cognitive symptoms of hippocampal atrophy (Sheline etal., 1999). Were this, in fact, the case, improved memoryfunction after ECT would not be unexpected. It is evenconceivable that ECT might exert beneficial effects onmemory function in certain neurodegenerative conditions(e.g., Alzheimer's, and Parkinson's diseases). The reliablybeneficial effects of ECT in Parkinson's disease, described indetail in Chapter 2, have always been attributed to either anincreased availability of dopamine at postsynaptic receptorsites, or an increased sensitivity of those receptors. MightECT-induced neogenesis of dopamine-containing cells in thesubstantia nigra constitute the real basis of the observedclinical improvement?

Prediction of Post-ECT CognitiveImpairmentScant attention has been paid in the literature to predictorsof ECT-induced cognitive impairment other than age,treatment electrode placement, seizure duration, andstimulus dose. In the first study of whether preexistingcognitive impairment constitutes a risk factor, Sobin et al.(1995) examined a sample of 71 depressed patientsrandomly assigned to high-or low-dose right unilateral or

P.199

bitemporal ECT, and found that pretreatment globalcognitive status assessed via Mini -Mental State Examination(Folstein, Folstein, and McHugh, 1975), and postictalorientation recovery time, were both significantly associatedwith retrograde amnesia for autobiographical memories 1week and 2 months after their treatment course, regardlessof treatment group assignment. These findings suggest thatthe Mini -Mental State Examination might be used as ascreening test to guide the selection of ECT treatmentparameters demonstrated to affect cognition (treatmentelectrode placement, stimulus characteristics, and frequencyof administration), and that postictal disorientation shouldbe routinely assessed for the same pur pose.

Management of Memory ImpairmentNo specific treatment is available to reduce the extent orduration of the memory impairment of ECT, which occurs inits most pronounced form after bitemporal ECT, especiallywhen administered at a higher dosage levels. The hopegenerated by early case reports (Weingartner et al., 1981;Partap, Jos, and Dye, 1983) that vasopressin mightattenuate ECT-induced memory impairment provedephemeral in a controlled trial (Mattes et al., 1990).

The physician is thus left with several practical measures totake in the event that emergent confusion and memory lossbecome problematic during a course of ECT. If sine-wave orconventional bitemporal ECT are being used, the obviousstep is to switch to a brief-pulse stimulus and unilateral orbifrontal electrode placement. If a thrice-weekly treatmentschedule is in force, reducing the frequency to twice a weekwill help substantially. The physiological principles discussedin Chapter 6 governing neuronal discharge and recoveryshould also be applied here, because increasing theefficiency of the stimulus parameter package should reducecognitive side effects as well as augment the therapeuticimpact. Lowering the pulse width to the 0.25 msec to 0.5msec range, and increasing stimulus duration to 68 secondswithout altering the dosage, should reduce memory loss andconfusion by reducing the charge rate. If bitemporal ECT isbeing used, the dosage can be reduced to the lowest levelthat is compatible with inducing a well-developed EEG

P.200

seizure pattern. However, I recommend against lowering thedosage with unilateral ECT for reasons already specified inChapters 6 and 7. Fortunately, dosage effects of unilateralECT on memory and cognition are reported to beundetectable 2 weeks after concluding a treatment course(Sackeim et al., 1993).

Does ECT Cause Persistent orPermanent Memory Impairment?Aside from the presentation by Weiner et al. (1986b), whichsuffers from the problems already discussed, the articlesreviewed above include every published study that soughtevidence for long-term or persistent memory andnonmemory cognitive effects of brief-pulse ECT. None ofthese studies demonstrate that any form of brief-pulse ECT,regardless of dosage, is capable of inducing memory deficitseven as early as 2 weeks following a course of treatment,much less months later. The fact that significant differencescan be found 2 months posttreatment between the amnesticeffects of brief-pulse bitemporal and brief-pulse unilateralECT (Lisanby et al., 2000; Sackeim et al., 2000) in no waydemonstrates that persistent adverse effects relative tobaseline performance were caused by either method alone;no such information in fact exists. (For example, bothmethods could cause memory improvement that is simplygreater with unilateral placement, a phenomenon that, asnoted above, does in fact occur for some measures.)

Moreover, in a unique follow-up study of patients who hadreceived unusually large numbers of ECTs, Devanand et al.(1991) compared 8 patients who had each received morethan 100 lifetime treatments with sine-wave, bitemporal ECT(mean of 12 courses and 160 ECTs per patient) with aclosely matched control group of depressives who had neverreceived ECT. The 2 groups did not differ on any of themany measures examined of objective or subjective memoryfunctioning.

How then to account for the assertions in a recent editorialthat, following ECT, â!œvirtually all patients experiencesome degree of persistent and, likely, permanent retrogradeamnesia.â! ! (Sackeim, 2000)? The editorial containsadditional statements to the effect that â!œin many patients

the recovery from retrograde amnesia will be incomplete,and there is evidence that ECT can result in persistent orpermanent memory loss,â! ! â!œsome patients mayexperience persistent amnesia extending several years priorto ECT,â! ! and, most strikingly, â!œincreasing evidence hasaccumulated that some degree of persistent memory loss[with ECT] is commonâ! ! (Sackeim, 2000). Notably, none ofthese statements is supported by literature citation.

The last of these comments suggests that the editorial is notreferring primarily to sine-wave therapy because the phraseâ!œincreasing evidence has accumulatedâ! ! can hardly referto a method that was essentially abandoned more than 20years ago. Assuming the claims apply at least equally tobrief-pulse therapy, what supporting evidence is provided toback them up? Unfortunately, noneâ!”the reader is requiredto accept the statements on faith alone.

It is notable that this editorial introduced a special issue ofJournal of ECT (â!œCognition and ECTâ! !) in which therealso appears a memoir by a recipient of ECT, entitled â!œElectroconvulsive Therapy and Memory Loss: A PersonalJourneyâ! ! (Donahue, 2000). The author of the memoir is noovert enemy of ECTâ!”indeed, she states both that ECT mayhave saved her life, and that had she the decision to makeover again, her choice would still be ECT. Moreover, she ishighly educated, writes cogently and well, has a thoroughknowledge of the relevant literature, and engages in nopolemics. Nevertheless, her conviction that she suffered â!œdevastating and permanent memory loss with ECTâ! ! isjust that: a personal conviction, and one that is, like manyother personal convictions, unsupported by any objectiveevidence.

Because the memoir was solicited for the special cognitiveissue of the journal it did not undergo the usual peer-reviewprocess; therefore the validity of the assertions it containsis both unknown and unknowable. The sincerity of its authoris not in question; the difficulty lies elsewhere, in thedisjunction between objective science and subjectiveexperience. No amount of memory or cognitive testing canever prove or disprove the truth of subjective experience,which is, by definition, inaccessible to the scientific method.(Subjective experience can, of course, be described,classified, catalogued, and the results of that cataloguing

P.201

analyzed, as in the many useful studies of subjectivememory described earlier in this chapter.)

In the absence of any published articles cited in support ofthe editorial's assertion that virtually all patients receivingECT likely experience a degree of permanent retrogradeamnesia, one might reasonably inquire whether the inclusionof a personal memoir in the special cognitive issue of

Journal of ECT was intended to substitute for the missingfacts. This, of course, it can not do.

Moreover, although as documented above, the overwhelmingmajority of patients who have received brief-pulse ECT atany dose consider their memory function to be verysubstantially improved, and although a recent articlestrikingly reports that patients who receive the most intenseform of

titrated brief-pulse ECT (2.5 Ã— threshold bitemporal ECT)rate their memory 71% better after ECT than ever before intheir life (Sackeim et al., 2000), these self-assessments arediscounted as â!œunlikelyâ! ! and â!œof doubtful validityâ! !(Sackeim, 2000). At the same time, however, the editorialunquestioningly accepts as accurate the self-assessments ofother patients who â!œreport that a large segment of theirlife is lost (after having received ECT).â! ! Neither the sourceof these self-assessments, nor an estimate of their validity,is presented in the editorial, leaving the reader in the darkas to how they should be evaluated.

One problem that has bedeviled all objective follow-upstudies of the memory effects of ECT is the fact thatcomparisons have so far only been made with the patient'scognitive status at pre-ECT baselineâ!”which is to say, whilethey were depressedâ!”or with normal or depressed no-ECTcontrol groups. Because whatever cerebral process causesmelancholia also impairs memory performance (whichimpairment, in its severest form, presents as the dementiasyndrome of depression), memory is invariably impairedimmediately prior to ECT, and often for weeks or monthsearlier. Comparing post-ECT memory performance withimmediate pre-ECT performance is therefore problematic:the comparison variable of greatest interest, pre-illnesscognitive performance, is unavailable for comparison (hencethe instructions of the subjective memory questionnaire that

require the patient to compare his present level with thebest he has ever before experienced).

Test results obtained from normal individuals can control forthis lack to a certain degree, but melancholic patients maynever have been entirely normalâ!”their brain function mayhave been different from the outset, just as the occurrenceof Alzheimer's dementia can be predicted from the writtencompositions of individuals six decades prior to the onset ofcognitive symptoms (Snowdon, Greiner, and Markesbery,2000). Certainly, to be convincing, post-ECT follow-upstudies should at least demonstrate improvement incognitive functioning relative to pre-ECT baseline. Asreviewed earlier in this chapter, 2 studies report suchimprovement: one for mostly non-memory cognitivefunctions examined 1-2 years following sine-wave unilateraland bitemporal ECT (Abrams and Taylor, 1985) and one forboth anterograde and retrograde memory examined 6months following brief-pulse bitemporal ECT (Calev, Nigal,and Shapira, 1991c).

Because absence of proof is not proof of absence, sciencecannot prove that ECT of whatever variety is incapable ofcausing long-term or persistent memory loss. Moreover,there is no a priori reason to believe that, under certaincircumstances, such an effect cannot occur. It is the task ofscience to design and carry out further studies of thequestion, using ever more sensitive and specific measures,with the aim of definingâ!”if indeed they existâ!”thosecircumstances under which long-term or persistentdysmnesia might be induced by ECT. Until and unless thosestudies are performed, and their results found to beconfirmatory, all claims that ECT can induce per manentmemory loss must receive the Scotch verdict: â!œunproved .â! !




> Table of Contents > Chapter 11 - Neurobiological Correlates and

Mechanisms

Chapter 11

Neurobiological Correlates andMechanisms

In this chapter, the word â!œneurobiologicalâ! ! subsumes allof the following terms unless more particularly specified:neurochemical, neurohumoral, neu rotransmitter,neuroendocrine, neurohormonal, neuropeptide, neurogenetic,neurophysiological, and neurometabolic.

Data from Animal StudiesThe following 2 quotations, a decade apart, serve to put thetopic of the usefulness of animal studies of ECS inperspective:

â!" which of the many establishedpharmacological actions of ECS [in animals]are important to the behavioral changes indepression? This is a difficult question toaddress in patients, so for the most part,one has to make inferences from animalstudies of nondepressed rodents. (Nutt,Gleiter, and Glue, 1989)

For more than three decades the effects ofelectroconvulsive shock (ECS) on synaptic(primarily monoaminergic) transmission insmall animal models have been a majorfocus. This approach is intrinsicallyreductionistic, and the relevance ofneurochemical changes in the brains ofnormal rodents to the clinical effects of ECTin depressed or otherwise ill patients hasbeen questioned. (Lerer, 1999)

The literature on neurobiological consequences of induced

P.203

seizures contains a dearth of human ECT studies comparedwith electroconvulsive shock (ECS) trials in animals, yetnormal animal data have not helped to explain how ECTworks in mentally ill patients. Moreover, although animalbrains are unarguably better suited than their humancounterparts to prodding, chopping, grinding, and liquefying,one searches in vain for a single supporting structure of ourpresent, painfully limited, neurochemical understanding ofhow ECT works that requires the underpinning of animaldata. Discrepant results abound, and the few consistentlyreplicable observations in animals simply do not generalizeto man. Thus, electrically induced seizures have oppositeeffects in man and in rodents on beta-adrenergic receptors(Mann et al., 1990), serotonin-2 receptors (Newman andLerer, 1988),

and plasma insulin (Thiagarajan, Gleiter, and Nutt, 1988);ECS increases the sensitivity of postsynaptic serotonin-1Areceptors in animals, but ECT does not do this in man(Shapira et al., 2000); naloxone lowers baseline andpostictal prolactin levels in rats but not in humans (Swartz,199la); and if seizure termination is mediated by opioidsystems in rats, it almost certainly is not in man (Sperlinget al., 1989; Rasmussen and Pandurangi, 1999).

The literature on the subject contains paper after paperreporting the effects of ECS on total or regional rodent brainconcentrations of a dizzying variety of substances for nobetter reason than that the techniques exist for measuringthem. Moreover, when the multiple statistical testsinvariably employed chance to reveal a significant change inone or another of these substances, the authors hasten tosuggest the potential relevance of their â!œfindingâ! ! to themode of action of ECT in psychiatric patients.

The problem presented by such studies is that the extremevariability in results within and between studiesâ!”dueprimarily to technical differences in experimental methodand designâ!”renders them so difficult to interpret.

Disparities in these receptor data mostlikely reflect the neurobiologicalcomplexities of ECS-induced receptoralterations coupled with interpretivedifficulties stemming from themethodological variables involved in thedetection of these changes â!" Thus,experimental outcomes may to some degree

P.204

be dictated by the extent of membranepurification, perhaps by influencing thecomposition of receptor sites available forassay. In addition, the relative receptorselectivities of the different radioligands,and their susceptibilities to the influence ofthe binding assay ionic composition areimportant factors determining experimentaloutcome. (Tortella et al., 1989)

and,

Different techniques may yield differentresults in regard to the same receptors,such as the effect of ECS on 5 HT-1Areceptor-mediated responses.Electrophysiological studies suggest thatthese receptors are rendered supersensitivein the hippocampus while biochemicalapproaches indicate down-regulation.(Lerer, 1999)

A few examples illustrate these points: One research groupfinds that transauricular ECS increases gamma-aminobutyricacid levels in rat brain in all areas except the hippocampusand nucleus accumbens, whereas trans-corneal ECSincreased it in hippocampus, frontal cortex, hypothalamus,and olfactory bulb, but reduces it in the nucleus accumbens(Ferraro, Golden, and Hare, 1990). The authors concludethat selective activation of neuronal pathways may beachieved by altering the placement of ECS stimulatingelectrodes.

Another study finds that ECS increases cyclohexyladenosinebinding sites in rat cerebral cortex for 14 days and reducesforskolin binding in hippocampus and striatum for 2 days(Gleiter et al., 1989). After substantial discussion on howdifferences in rat strain or assay method might account

for their failure to reproduce other investigators' results, the

authors conclude that different seizure-inducing agents andantidepressant treatments seem to have different effects onthe mouse adenosine neuromodulatory system.

Yet another study finds that ECS does not significantlychange rat brain regional concentrations of substance P,neurotensin, or vasoactive intestinal polypeptide in thehypothalamus, pituitary, or left or right frontal or occipital

cortices, but does significantly elevate neuropetide Y inhippocampus and occipital cortex. Multiple additional trends,tendencies, and marginal changes are offered, all fluctuatingabout p = 0.05 (Stenfors, Theodorsson, and Mathe, 1989).The authors conclude that regional increments inneuropeptide Y might constitute, in part, the therapeuticmode of action of ECT.

To make matters worse, it now appears that most of thepre-1990 lit erature on rodent brain neurotransmitters mustbe discounted because the studies relied on, of all things,data from dead animals!

There is considerable conflicting earlierliterature on the changes inneurotransmitter turnover produced by ECS(the entire animal neurotransmitterliterature through 1989 is cited). Thesestudies relied on postmortem neurochemicalstudies of neurotransmitter concentrationsand turnover to infer the effect of ECS ontransmitter release. (Nutt, Gleiter, andGlue, 1989)

Two major reviews of the animal ECS literature haveappeared during the last decade (Fochtmann, 1994a;Newman and Lerer, 1998), but among the nearly 1000studies tabulated there is not a single one that bears anyproven relevance to the action of ECT in man.

However, as can be seen, experiments onECS in animals have not always yieldedconsistent results. In the absence ofdefinite knowledge as to which brain areaor areas are involved in depression, andwhich particular neurotransmitter pathwayor pathways are involved in thepathogenesis of the disorder, it is hard todetermine which of the proposedmechanisms applies most closely to theclinical effects of ECT. (Newman and Lerer,1998)

Actually, other than the 65-year-old observation in a Romeslaughterhouse confirming that an electric current passedthrough the head is not fatal, there is not a single piece ofthe admittedly meager accumulation of hard data that we

P.205

have concerning ECT in man that was derived from animalresearch. This being the case, the remainder of this chapteris devoted solely to human studies of ECT.

Human DataHuman data are plentiful enough for our purposes, if onlythey can survive the close scrutiny prescribed by Kety et al.(1967) before incorporating them into our understanding ofthe ECT process:

â!" the difficulty lies not in demonstratingsuch changes, but in differentiating betweenthose which are more fundamental andthose which are clearly secondary, and alsoin attempting to discern which of thechanges may be related to the importantantidepressive â!" effects and which arequite irrelevant to those. (Kety et al.,1967)

Unfortunately, the disorders for which ECT is prescribed arequite possibly the result of pathological states that derivefrom, or cause, changes in the particular neurochemicalstudiedâ!”either before, during, or after ECT â!”thusincreasing the chance that a given result will be spurious:

An important assumption, which is veryhard to test, is that the observedalterations in pharmacological processesprecede rather than are a consequence ofthe behavioral changes. (Nutt, Gleiter, andGlue, 1989)

The problem is compounded by the frequently small samplesizes studied and the multiple statisticalâ!”especiallycorrelationalâ!”tests employed (usually without appropriatecorrection for multiple testing), and by the inclusion of â!œtrends,â! ! â!œtendencies,â! ! and â!œnear-significantâ! !results that, within a year or two of publication, are widelycited as fact by other investigators desirous of establishingthe probity of their own trends, tendencies, and near-significant results.

Tenuous theoretical constructs from other studies withequally inadequate methods are thus elevated to the level ofâ!œsupporting data,â! ! thereby creating the general illusion

P.206

of a rather substantial and compelling body of knowledge. Toinsist that studies meet standard methodological andstatistical (including power analysis) requirements andinclude an untreated or sham-treated control group to beaccepted as valid would eliminate virtually every referenceon the subject. This melancholy excercise has, in fact, beentwice performed for the literature on ECT compared withantidepressant drugs (Abrams, 1982b; Rifkin, 1988), eachtime with devastating results.

Measurement of various brain neurotranmitters and theirmyriad derivatives in blood and urine is at once the easiestand least productive approach to elucidatingneurotransmitter mechanisms of action of ECT because ofthe confounding effects of diagnostic heterogeneity, previousand concomitant drug adminstration, effects of peripheralneurotransmitter production and metabolism, anduncontrolled motor activity level. Even spinal fluid studiesare bedeviled by active transport of metabolites into thecerebrospinal fluid and the effect of motor activity (includingespecially body position at the time of the tap) on cephalad-caudad concentration gradients.

Still further removed from cerebral events are the studies oflymphocyte or platelet neurotransmitter receptors. Nocompelling data on possible mechanisms of action of ECThave been forthcoming from any of these lines ofinvestigation (e.g., Linnoila et al., 1984; Lerer, 1987;Buckholtz et al., 1988; Rausch, Rich, and Risch, 1988;Cooper et al., 1988; Devanand et al., 1989; Lykouras et al.,1990; Stain-Malmgren, Tham, and Aber-Wistedt, 1998),

which have mostly been abandoned in favor of indirectneuroendocrine re lease and challenge studies ofmonoaminergic function.

However, although ECT-induced pituitary hormone releaseâ!”alone, and in response to drugs that have known effectson central monoaminergic systemsâ!”can reflect ECT-induced changes in receptor functioning, care must be takento choose hormones that are not altered in depressiveillness because changes in hormonal responses that are dueto relief from depression may be interpreted incorrectly asreflecting changes in receptor sensitivity (Swartz, 1991a).Growth hormone, thyrotropin, oxytocin, vasopressin,corticotropin, luteinizing hormone, follicle-stimulatinghormone, endorphins, and prolactin responses to ECT haveall been studied, with remarkably, and often disappointingly,variable results.

P.207

The design and interpretation of many human ECT studies isalso unfortunately based on a presumed analogy with theaction of psychotropicâ!” especially antidepressantâ!”drugs:because ECT and antidepressants both treat depression, forexample, they must have the same mechanism of action.The fact that, after 40 years of intensive investigation, nomechanism of action has yet been defined for the efficacy ofany antidepressant drug in man, should give pause toinvestigators attempting to base their ECT studies onpresumed or postulated effects of antidepressants. The logicof such an approach is faulty because ECT is the morepowerful agentâ!”psychotropic drug studies should from theoutset have been based on leads developed from ECTresearch, rather than vice versa.

A recent study of the neuropeptide response to ECTconducted at a leading research center will illustrate someof the foregoing points. Because past studies ofpsychotomimetics, antipsychotics, and lithium had revealedeffects on neuropeptide systems, the decision was made tostudy neuropeptide changes as a potential mechanism ofaction of ECT in man. CSF was examined before and afterECT for levels of various neuropeptide transmitters, with theresult that immunoreactivity for several of them increased:neuropeptide Y, somatostatin, corticotrophin-releasingfactor, and endothelin. Moreover, the patients who got ECTimproved. On the basis of these findings, it washypothesized that neuropeptides were involved in themechanism of action of ECT, and that distinct combinationsof neuropeptide and monoamine changes in selectedneuronal populations were essential to the effect of ECT onspecific disease symptoms, possibly even regardless ofdiagnosis.

The problem with this interpretation for a treatment such asECT, that releases so many substances into the CSF, is thatwhen most or all the patients studied improve withtreatment, the necessary conditions are lacking fordemonstrating causality. A range of treatment outcomes isneeded, so that the extent of neurotransmitter release canbe related to the degree of improvement: a studydemonstrating that ECT responders exhibited increased CSFlevels of neuropeptide Y, but that nonresponders did not,would provide a strong basis for further examination of therelationship. (Such a result would not in itself demonstratecausality, of course, any more than the observed strongcorrelation during neuroleptic therapy between the severity

of extrapyramidal symptoms and the degree of clinicalimprovement demonstrates that the former causes thelatter.)

The enormous difficulties for human studies posed by all ofthe above considerations are most clearly shown by thefailure of investigators to uncover the neurochemical basisfor the anti-Parkinson action of ECT. The clinical efficacy ofECT in relieving the motor symptoms of both Parkinson'sDisease and neuroleptic-induced parkinsonism has beenabundantly demonstrated over the past 30 years (seeChapter 2). Moreover, the pathophysiology of Parkinson'sdisease is extremely well-documentedâ!”precisely localizedto dopaminergic cells of the substantia nigraâ!”and effectivetreatment of Parkinson's disease with both L-DOPAreplacement and dopamine-receptor stimulants has beenthoroughly worked out. In short, Parkinson's disease is anECT-investigator's dream.

Yet a recent review of the effects of ECT on the humandopaminergic system (Mann, 1998) consists of 2 shortparagraphs containing only the news that (1) most CSFstudies find no change in HVA (the primary dopaminemetabolite) after a course of ECT; (2) neuroendocrinestudies are equally divided as to whether or not ECTenhances dopamine receptor sensitivity; and (3) the self-evident statement that

â!" the observation that Parkinson'sdisease responds to ECT suggests that ECTincreases dopaminergic transmission in thebasal ganglia.

ProlactinBy far the most consistent neurochemical result of ECT-induced seizures is the 10-to 50-fold increase in immediatepostictal serum prolactin levels, which peak at 10 to 20minutes poststimulus and return to baseline within 2 hours(Ryan et al., 1970; Ohman et al., 1976; Klimes et al., 1978;O'Dea et al., 1978; Meco et al., 1978; Arato et al., 1980;Skrabanek et al., 1981; Balldin, 1982; Whalley et al., 1982,1987; Linnoila et al., 1984; Swartz and Abrams, 1984;Haskett, Zis, and Albala, 1985; Swartz, 1985; Papakostas etal., 1985, 1990; Scott et al., 1986; Tauboll et al., 1987;Johansson and von Knorring, 1987; Turner, Ur, Grossman,1987; Weizman et al., 1987; Markianos, Papakostas, andStefanis, 1987; Cooper et al., 1989; Sperling et al., 1989;

P.208

Devanand et al., 1989; Zis et al., 1991, 1993 , 1996; McCallet al., 1996; Lisanby et al., 1997; Sundblom et al., 1999;Chaudhry et al., 2000). More than 30 research reports showthis effect, and no contradictory data exist.

Bilateral ECT releases more prolactin into the bloodstreamthan unilateral ECT (Swartz and Abrams, 1984; Papakostaset al., 1984; Swartz, 1985; Clark, Alexopoulos, and Kaplan,1995; Lisanby et al., 1998), as do higher, compared withlower, electrical doses (Abrams and Swartz, 1985b; Robin,Binnie, and Copas, 1985; Zis et al., 1993, 1996; Lisanby etal., 1998), and multiple, compared with single, electricalstimuli (Abrams and Swartz, 1985b).

Thus, variations in technique that increase the therapeuticimpact of ECT also increase prolactin release.

Electroconvulsive therapy-induced prolactin release is notblocked by naloxone (Haskett, Zis, and Albala, 1985;Papakostas et al., 1985; Turner, Ur, and Grossman, 1987;Weizman et al., 1987; Sperling et al., 1989), so it is notdependent on opioid receptor mechanisms; it is notenhanced by phentolamine (Klimes et al., 1978), so it is notan alpha-adrenergic phenomenon; it is not blocked byketanserin or ritanserin (Zis et al., 1989a; Papakostas et al.,1990), so it is not mediated by serotonin-2 receptors; and itis not blocked by ondansetron so it is not mediated byserotonin-3 receptors (Yatham et al., 1996).

The fact that methysergide, a nonselective serotoninreceptor blocker that also exhibits dopaminergic properties,attenuates ECT-induced prolactin release (Papakostos,Markianos, and Stefanis, 1988; Zis et al., 1989b), suggestedto these authors that such release is either serotonin-1-mediated or secondary to acute antidopaminergicmechanisms, as proposed by Deakin et al. (1983). The latterhypothesis received support from Cooper et al. (1989), whoasserted that their finding of a sharp increase in ECT-induced prolactin and luteinizing hormone levels withoutconcomitant release of follicle-stimulating hormone favored â!œa powerful acute antidopaminergic action for ECTâ! !(although they found significant increases in cerebrospinalfluid homovanillic acid levels induced by the first ECT).

The contribution of serotonergic mechanisms to ECT-inducedprolactin release finds support in the study of Shapira et al.(1992), who reported that ECT enhanced fenfluramine-induced prolactin release, suggesting an increase in centralserotonergic responsivity (that was not, unfortunately,

P.209

correlated with the treatment response to ECT). Whalley etal. (1987), likewise conclude that ECT-induced increases inplasma prolactin and adrenocorti cotropic hormone, withlittle or no effect on luteinizing hormone or thyro tropin,probably reflect the activation of serotonergic neurones.

Because pituitary prolactin release is under tonicdopaminergic inhibitory control, the many studies showing ablunting of ECT-induced prolactin release across a course oftreatment (Meco et al., 1978; Whalley et al., 1982, 1987;Balldin, 1982; Deakin et al., 1983; Abrams and Swartz,1985a; Aperia, Thoren, and Wetterberg, 1985; Haskett, Zis,and Albala, 1985; Lisanby et al., 1998; Scott, Milner, andShering, 1989; Chaudhry et al., 2000) suggest that repeatedECTs produce a sustained increase in postsynaptic pituitarydopamine receptor sensitivity, an effect thatâ!”if it alsooccurs in the brain â!”neatly explains the phenomenon ofECT-emergent dyskinesias, the need for the reduction oflevodopa dose in some patients with Parkinson's disease whoreceive ECT (Balldin et al., 1980b; Douyon et al., 1989), andthe almost universally favorable effects of ECT on the motormanifestations of Parkinson's disease (Rasmussen andAbrams, 1991, 1992; Fall et al., 1995). The report of Balldinet al. (1982) that ECT systematically increased apomorphinesuppression of serum prolactin in depressed subjects isconsistent with this view, although contradictory andequivocal data exist (Coppen et al., 1980a; A. Christie et al.,1982; Lerer and Belmaker, 1982; Linnoila, Karoum, andPotter, 1983).

Lisanby et al. (1998) related blunting of the prolactin surgeacross a course of ECT to clinical response: treatmentrespondersâ!”and those who showed the greatestimprovement in depression scale scoresâ!”exhibited suchblunting; nonresponders did not. Because the prolactin surgewith ECT reflects reduction in the tonic hypothalamicinhibition of pituitary prolactin release (Zis et al., 1992),this particular result of Lisanby et al. (1998) is compatiblewith a diencephalic effect of ECT.

Electroconvulsive therapy-induced prolactin release is afunction of seizure duration in some studies (Skrabanek etal., 1981; Balldin, 1982; Johansson and von Knorring, 1987;Mitchell et al., 1990) but not in others (Aperia, Thoren, andWetterberg, 1985; Swartz and Abrams, 1984; Swartz, 1985;Papakostas et al., 1986a,b , 1990; McCall et al., 1996;Lisanby et al., 1998). Balldin (1982) proposed that ECT-induced prolactin release was a biochemical marker of

P.210

seizure activity because it correlated significantly with motorseizure but not ECT stimulus duration; this was, of course,prior to the research findings cited above that ECT-inducedprolactin release was also a function of stimulus dose.

The finding of Linnoila et al. (1984) that ECT-inducedprolactin elevations did not correlate with changes in plasmalevels of the dopamine metabolite, homovanillic acid,suggests either that the phenomenon is independent ofpituitary dopamine metabolism, orâ!”more likelyâ!”that thecontribution of tuberoinfundibular dopamine metabolism toplasma homo-vanillic acid concentration is trivial. Papakostaset al. (1986a) found a close correlation between theprolactin responses to ECT and to thyrotropin-releasinghormone, suggesting a possible common underlyingprolactin-releasing mechanism.

Despite the reliable observations catalogued above thatmore clinically potent ECT stimulus methods release moreprolactin, a consistently positive correlation between ECT-induced prolactin release and clinical improvement has failedto materialize: some have found it (Scott et al., 1986;Whalley et al., 1987), others have not (Deakin et al., 1983;Abrams and Swartz, 1985a; Clark, Alexopoulos, and Kaplan,1995; Lisanby et al., 1998).

Swartz (1991b) argues that because ECT-induced prolactinrelease probably primarily reflects pituitary chemistry, itsstudy is unlikely to reveal much about neurochemicalmechanisms of ECT. However, the amount of prolactinreleased during ECT is sensitive to stimulus dosing andtreatment electrode placement, and may also reflect thephysiological impact of the ECT technique used. The resultsof the study of Zis et al. (1992) support this view. Theseauthors examined the effects of ECT on prolactin releasewith and without prior administration of metoclopramide toblock dopamine receptors. Following metoclopramide (whichinduced a substantial prolactin surge), no further increase inprolactin levels was seen with ECT, and the authorsconcluded, in contrast to Swartz's (199la) view, that ECT-induced prolactin release was not merely an extracerebralphenomenon but resulted

from decreased hypothalamic dopaminergic inhibition of thepituitary lac totroph.

Moreover, there is evidence from depth electrode studies inpatients with partial epilepsy for a specific relation betweenintracerebral stimulation, seizures, and prolactin release. In

1 study (Sperling and Wilson, 1986), electrical stimulationswere administered to amygdala, hippocampus, and frontalcortex: Only stimulations producing a high-frequencyregional limbic afterdischarge resulted in serum prolactinelevation. In a second study (Sperling et al., 1986), serumprolactin levels always rose after complex partial sei -zuresâ!”all of which involved limbic dischargesâ!”and only roseafter simple partial seizures if they induced limbicdischarges. The authors concluded that prolactin releaseduring partial seizures resulted from spread of dischargesfrom mesial temporal structures to ventromedialhypothalamus. Just as occurs after ECT, postictal serumprolactin levels after complex partial seizures peaked at 15minutes and returned to baseline in about 1 hour, withoutsignificant correlation with seizure duration.

Prolactin release is thus a reliable marker of the ECT-induced seizure and sensitive to both dose and electrodeplacement effects, and the anatomical brain regions involvedin the seizure. Lisanby et al. (1998) are the firstinvestigators so far to have systematically examined therelation between prolactin release and the clinical responseto ECT, varying both dosage and electrode placement, andthereby obtaining a sufficiently wide range of clinicalresponses to provide a valid study. Their failure to find adirect relation between the prolactin surge and the clinicalefficacy of ECT may be attributable to the inefficientstimulus parameters they used, which, although not specifiedin their article, were likely to have been essentially limitedto a pulsewidth of 1.5 ms and a stimulus train duration of 1second (Boylan et al., 2000). It is therefore quite possiblethat studies from other research sites where morephysiologically efficient stimuli are employed (e.g., 0.5 mspulsewidth, 6-8 sec stimulus train) may provide a better testof the sought-after correlation between ECT-inducedprolactin release and clinical anti depressant response.

Thyrotropin, Thyrotropin-ReleasingHormoneAcute release of thyrotropin (thyroid -stimulating hormone)from the posterior pituitary during ECT is reported by someinvestigators (Aperia, Thoren, Wetterberg, 1985; Tauboll etal., 1987; Dykes et al., 1987; Scott, Milner, and Shering,1989; Papakostas et al., 1990) but not others (Thorell andAdielsson, 1973: O'Dea et al., 1979; Whalley et al., 1987;Cooper et al., 1989). Cooper et al. (1989) neverthelessreported a significant correlation between the averagechanges in thyrotropin and prolactin across 4 ECTs.

P.211

changes in thyrotropin and prolactin across 4 ECTs.

The amount of thyrotropin released falls across a course ofECT (Aperia, Thoren, and Wetterberg, 1985; Scott, Whelley,and Legros, 1989), although the thyrotropin response tothyrotropin-releasing hormone is variously

reported diminishing (Decina et al., 1987), increasing(Kirkegaard et al., 1977; Nerozzi et al., 1987), or remainingunchanged (Coppen et al., 1980b) over the treatmentcourse. Cooper et al. (1989) found less thyrotropin releasedafter the last than after the first ECT of a series, but thereduction was not significant. As for prolactin, theattenuation in thyrotropin release across a treatment coursehas been attributed to an ECT-induced increase inpostsynaptic dopamine receptor function (Aperia, Thoren,and Wetterberg, 1985). Although Scott, Milner, and Shering(1989) found no relation between ECT-induced thyrotropinrelease and antidepressant efficacy as measured by thechange in Hamilton depression scale score, they did obtain aperfect correlation between summed total EEG spike activityand change in thyrotropin release across a course oftreatment. This confirmed an earlier finding (Dykes et al.,1987) that significant thyrotropin release does not occurduring seizures lasting less than about 30 seconds, andprovides a potential biological rationale for the arbitraryclinical dictum to repeat ECT-induced seizures that fail tolast this long. However, because thyrotropin release wascorrelated with seizure duration and not with clinicalefficacy, Scott, Milner, and Shering (1989) concluded thatECT-induced thyrotropin release is probably anepiphenomenon of seizure activity and unlikely to provideinsights into the antidepressant mode of action of ECT.

The effect of a course of ECT on the thyrotropin response toexogenous thyrotropin-releasing hormone has also beenstudied, with the usual conflicting results. Kirkegaard andCarroll (1980) reported an increase in the thyrotropinresponse to thyrotropin-releasing hormone following ECT,the extent of which was inversely correlated with the risk ofrelapse. However, not only were Decina et al. (1987) unableto find a single patient who met Kirkegaard and Carroll's(1980) criterion for predicting relapse among 23 patientsstudied, but they observed the same change in thyrotropinresponse in responders as in nonrespenders: No patient witha pretreatment blunted thyrotropin response showednormalization after ECT, andâ!”most striking of allâ!”themean thyrotropin response became significantly moreblunted after ECT, despite a generally favorable response to

P.212

treatment. Thyrotropin-releasing hormone is not reported tobe released by ECT (Whalley et al., 1982).

The most recent studies on the subject have been entirelynegative. In a sample of 42 unipolar depressives, Lykouraset al. (1993) found no difference between responders andnonresponders for the thyrotropin response to TSH, andHofmann et al. (1994) reported similarly disappointingresults in a sample of 20 patients. There have been nostudies since then. Based on all available data, therefore,further investigation of this topic is unwarranted.

Corticotropin-Releasing Factor,Corticotropin, and CortisolIt was already known by the late 1960s that elevatedplasma cortisol levels accompanied the stress of depressiveillness; that ECT temporarily induced

acute postictal increases in plasma levels of corticotropinand cortisol; that the elevated baseline cortisol levels fellwith improvement in the depressed state; and that thecortisol response to ECT was unrelated to this improvement(Gibbons and McHugh, 1962; Hodges et al., 1964; Bersonand Yalow, 1968). Since then, our understanding of thissubject has been modestly increased by the finding that theplasma cortisol response to ECT is dose-dependent (Zis etal., 1996). Any enthusiasm for investigating the cortisol -releasing effects of ECT will surely be dampened by thefinding that cardioversionâ!”which does not involve passingany current through the brainâ!”has essentially the sameeffects on plama cortisol as ECT (Flor kowski et al., 1996).

Although ECT does not release corticotropin -releasing factorinto the blood (Widerlov et al., 1989), it neverthelessacutely elevates plasma levels of corticotropin (Berson andYalow, 1968; Whalley et al., 1987; Widerlov et al., 1989)and cortisol (Gibbons and McHugh, 1962; Hodges et al.,1964; Elithorn et al., 1968; Arato et al., 1980; Deakin etal., 1983; Haskett, Zis, and Albala, 1985; Whalley et al.,1987 Weizman et al., 1987; Mathe et al., 1987; Widerlov etal., 1989; Mitchell, Smythe, and Torda, 1990) and increasesthe cortisol response to intravenous infusions of the alpha-2receptor agonist, methylamphetamine (Slade and Checkley,1980). Turner, Ur, and Grossman (1987) provide the onlycontradictory data.

Basal and ECT-induced cortisol levels fall across a course ofECT (Whalley et al., 1982; Christie et al., 1982; Linkowski

P.213

et al., 1987; Widerlov et al., 1989), as do the cortisol (butnot the corticotropin) response to exogenous corticotropin -releasing factor (Dored et al., 1990) and the cerebrospinalfluid levels of corticotropin -releasing factor itself (Nemeroffet al., 1991).

Dexamethasone Suppression TestNumerous investigators have studied the effects of ECT onthe response of the elevated plasma cortisol levels indepressed patients to exogenously administereddexamethasone, examining the phenomenon as eithercontinuous (postdexamethasone plasma cortisol levels) ordichotomous (suppressor compared with nonsuppressorstatus). Four studies (Albala et al., 1981; Papakostas et al.,1981; Varma et al., 1988; Palmer et al., 1990) found that acourse of ECT significantly reduced postdexamethasoneplasma cortisol levels; one found that it did not (Devanandet al., 1987). Similar variability has been reported for theeffects of ECT on suppressor status (Coryell andZimmerman, 1983; Lipman et al., 1986a,b; Fink, Gujavarty,and Greenberg, 1987; Devanand et al., 1987; Katona et al.,1987). Swartz and Saheba (1990) found that doubling thestandard dexamethasone dose to 2 mg sharply reducedinterpatient variability in postdexamethasone plasma cortisollevels after but not before a course of ECT; in fact, all oftheir patients tested with the 2-mg dose showed post-ECTsuppression.

Interindividual variability has been a major problem withclinical use of the dexamethasone suppression test as apredictor or correlate of ECT response (Swartz, 1997). Anumber of studies (but none since 1990) have attempted touse the dexamethasone suppression test to predict relapseafter successful treatment with ECT. Although several ofthese studies report that suppressor status following an ECTcourse is associated with continued remission at follow-up,the studies are plagued with serious methodological flaws:ECT was usually not the onlyâ!”or even the predominantâ!”mode of treatment; treatment response was oftenassessed with full knowledge of suppressor status, andtreatments received during the follow-up interval wereneither controlled nor systematically applied (Bourgon andKellner, 2000). In summary, the dexamethsone suppressiontest appears to have no demonstrably valid use forassessing any aspect of the clinical response to ECT.

Because melancholia is associated with a state-dependent,

P.214

stress-related, hypercortisolemia that attenuates with clinicalimprovement regardless of treatment method, changes inbasal and postdexamethasone levels of cortisol â!”as well asin its antecedentsâ!”have also been uniformly uninformativewith regard to mechanisms of action of ECT. Thus, thereliable observation that a course of ECT reduces elevatedbaseline plasma cortisol levels in depressed patients is mostlikely due to a nonspecific reduction in stress-inducedcorticotropin release associated with relief of depression.

Growth HormoneGrowth hormone levels are generally reported unchanged orreduced immediately after ECT (Yalow et al., 1969; Ryan etal., 1970; O'Dea et al., 1979; Arato et al., 1980; Whalley etal., 1982, 1987; Deakin et al., 1983; Linnoila et al., 1984;Robin, Binnie, and Copas, 1985; Haskett, Zis, and Albala,1985; Scott et al., 1986; Linkowski et al., 1987; Turner, Ur,and Grossman, 1987; Weizman et al., 1987), althoughcontradictory data exist (Skrabanek et al., 1981; Aperia,1986).

Oxytocin, Vasopressin, andAssociated NeurophysinsOxytocin, vasopressin, and associated neurophysins arereleased by ECT-induced seizures (Whalley et al., 1982,1987; Scott et al., 1986; Smith et al., 1990; Scott, Whalley,and Legros, 1989; Scott et al., 1991), but the amountreleased neither correlates with seizure duration (Scott,Whalley, and Legros, 1989) nor changes over a course ofECT (Scott et al., 1991). The amount of oxytocin-associatedneurophysin released to the first ECT was significantlycorrelated with ECT-induced improvement in depression(Scott, Whalley, and Legros, 1989; Scott et al., 1991), butthe original reports of an overall correlation between ECT-induced neurophysin release and clinical improvement (Scottet al., 1986; Whalley et al., 1987) were not confirmed

in a subsequent study from the same laboratory (Scott etal., 1991). Subsequent studies from independentlaboratories (Smith et al., 1994; Devanand et al., 1998)have not confirmed any of these clinical correlations, andthere is no present viable hypothesis of the mechanism ofaction of ECT that invokes oxytocin, vasopressin, orassociated neurophysins.

Endorphins

P.215

Electroconvulsive therapy reliably increases baseline(Alexopoulos et al., 1983; Misiaszek et al., 1984; Weizmanet al., 1987) and immediate postictal plasma endorphinlevels (Emrich et al., 1979; Alexopoulos et al., 1983;Misiaszek et al., 1984; Inturrisi et al., 1982; Weizman et al.,1987; Ghadirian, Gianoulakis, and Nair, 1988; Griffiths etal., 1989; Dored et al., 1990; Chau dhry et al., 2000), 9observations in search of a clinical correlation.

The possibility that endogenous opioid mechanisms mediatethe termination of ECT-induced seizures and therefore thereduction in seizure threshold and duration across atreatment course (Tortella et al., 1989; Nakajima et al.,1989) is confuted by the inability of naloxone at any dose toalter the length of ECT-induced seizures (Sperling et al.,1989; Rasmussen and Pandurangi, 1999; Prudic et al.,1999). To the extent that the anticon vulsant hypothesis ofECT depends on such a mechanism, it is weakened by theseresults.

Neuroendocrine Challenge TestsDrugs that release hormones from the anterior pituitary viastimulation of specific monoamine systems have been usedas challenge tests in patients receiving ECT. Two suchcompounds are apomorphine and clonidine.

ApomorphineHormonal responses to the dopamine agonist apomorphinehave been used indirectly to probe dopamine receptorchanges induced by ECT, with inconsistent results. Modigh etal. (1984) found that, although all of 17 patients treatedimproved with ECT, some exhibited increases and somedecreases in apomorphine-induced growth hormonesuppression. Costain et al. (1982) found a significantincrease in apomorphine-induced growth hormone releaseafter a course of ECT, which the authors noted wasconsistent with an ECT-induced increase in dopaminereceptor sensitivity. Although Christie et al. (1982) claimedfailure to confirm such an effect, their published figure isvirtually identical to that of Costain et al. (1982) and showsthat, after a course of ECT, growth hormone levels weresubstantially increased at baseline and at each of 6postapomorphine sampling intervals, although with

statistical significance found only 15 minutespostapomorphine. Balldin et al. (1982) found that ECTenhanced apomorphine-induced suppression of prolactin

levels in parkinsonian patients, although this effect did notcorrelate with improvement in motor symptoms.

It is now more than a decade since anything has appeared inthe liter ature on the subject.

ClonidineSlade and Checkley (1980) failed to find any change in thegrowth hormone response to intravenous infusions of thealpha-2 receptor agonist clonidine in depressed patientsbefore and after a course of ECT or to infusions of thealpha-2 receptor agonist, methylamphetamine; however,they did observe a significant increase in the cortisolresponse to this drug. In addition, Balldin et al. (1992)found that clonidine-stimulated growth hormone secretionwas significantly blunted by ECT, suggesting down-regulationof hypothalamic alpha-2 receptors. Unaccountably, althoughthe authors had collected depression rating scale data, theydid not examine the relationship between changes in growthhormone secretion and clinical outcome. Coote et al. (1998)did not confirm that ECT blunted the growth-hormoneresponse to clonidine. They found that although patientswho received ECT had a significantly more blunted growth-hormone response to a clonidine challenge at baseline thancontrols, this was no longer true after a course of ECT;moreover, the modest change induced by ECT did notcorrelate with im provement.

Mechanisms of ActionFor a neurobiological event or process to be consideredcentral to the mechanism of action of the antidepressanteffect of ECT, the occurrence, intensity, or duration of thatprocess should correlate with the degree of clinicalimprovement induced by ECT, measured either as acontinuous (e.g., depression scale change) or discrete (e.g.,responder/nonresponder status) variable. For such an eventor process to be considered the primary causative agent ofthe efficacy of ECT in depression, it must be demonstratedto be both necessary and sufficient for the antidepressanteffect of ECT.

As of this writing, although the occurrence of a generalizedseizure is widely considered central (e.g., necessary) to themechanism of action of ECT, there is abundant evidencefrom low-dose unilateral ECT studies that such a seizure isnot alone sufficient to account for the antidepressant effectof ECT (e.g., Sackeim et al., 1987a).

P.216

Moreover, considerable evidence has recently been adducedfor an antidepressant action of brain electrical stimulation inthe absence of a generalized seizure, when that stimulationis induced via magnetic fields generated

close to the skull (see Chapter 13). Because the brainelectrical potentials induced by magnetic fields cannot bedifferentiated from those induced by electrical currentsapplied directly to the scalp, the central role of the seizurenow appears somewhat less strongly established, albeit onlywith regard to the categories of depressed patients thathave been shown to repond to nonconvulsive magneticstimulation. (As of this writing, although there may beminimal overlap in the clinical indications for nonconvulsivemagnetic stimulation and for ECT, it is generally concededthat the two treatment methods have substantially differentindications.)

In any event, it is the occurrence of the ECT-inducedseizure, and not its duration, that is reported to be centralto the antidepressant efficacy of ECT (Ottosson, 1960), justas the occurrence, but not the duration, of high-frequencylimbic discharges is prerequisite for prolactin release duringpar tial seizures (Sperling et al., 1986).

In like manner, it may be the occurrence, and not theduration, extent, or intensity of other neurobiologicalprocesses that is central to the efficacy of ECT, in whichcase a significant correlation with clinical improvement maybe detected only for the occurrence of that process, and notfor its duration, extent, or intensity.

The two main neurobiological hypotheses of the mechanismof action of the antidepressant effect of ECT are theanticonvulsant hypothesis (Sackeim, 1999b), and the seizuregeneralization hypothesis (Krystal and Weiner, 1999); thediencephalic stimulation hypothesis (Abrams and Taylor,1976a; Abrams, 1997) has evolved into a subset of thelatter.

The Anticonvulsant HypothesisAs noted in Chapter 6, the original basis for theanticonvulsant hypothesis of ECT is the reliable observationthat the seizure threshold typically increases across a courseof ECT (Sackeim et al., 1983a; 1987b). A key element wasadded when it was further reported that patients whoseseizure thresholds do not increase substantially across acourse of treatment fail to enjoy a good clinical outcome

P.217

(Sackeim et al., 1987c, 1993; Sackeim, 1994a; Sackeim,1999b).

Independent confirmation of these latter results has notmaterialized. Neither the absolute value of the threshold,nor the magnitude of its increase over a course of ECT havebeen found by subsequent investigators to predict clinicaloutcome; on the contrary, most studies show the oppositerelationship: the greater the seizure threshold increase, theworse the clinical response (Krystal et al., 1998; Delva etal., 2000; Scott and Boddy, 2000).

Coffey et al. (1995b) employed a flexible stimulus titrationschedule to estimate seizure threshold change across acourse of 6 unilateral or bitemporal ECTs, and found thatwith moderately suprathreshold ECT almost half of theirdepressed patients showed no threshold increase. Moreover,in the patients that exhibited such an increase, there was noassociation between

the rise in threshold and either therapeutic response statusor speed of response. In a subset of this patient sample,Krystal et al. (1998) later demonstrated that a decrease instimulus intensity relative to seizure threshold over a courseof ECTâ!”and therefore an increase in seizure thresholdâ!”was actually associated with a reduced antidepressantresponse. Likewise, Shapira et al. (1996) used a stimulustitration schedule to determine seizure threshold changeacross a course of ECTs and found no relation between anobserved 42% threshold increase and improvement intreatment -resistant depressives in response to moderatelysuprathreshold bitemporal ECT. In their reanalysis of datafrom an earlier study of 3 different electrode placements,Delva et al. (2000) reported threshold increases overbaseline by the end of the treatment course ranging from57% to 142%, but those increases did not predict clinicalresponseâ!”on the contrary, the group with the smallestthreshold increase exhibited the best clinical outcome, andvice versa. Similar results were reported by Scott and Boddy(2000) for titrated 2Ã— threshold bitemporal ECT: meanseizure thresholds increased a modest 23% across a courseof 6 ECTs, there was no association between the extent ofthe increase and clinical response, and the thresholdincrease was actually sub stantially larger in thenonresponders than the responders.

Finally, although seizure duration by no means always variesinversely with seizure threshold, a fall of seizure durationacross a course of ECT is typically associated with a

P.218

corresponding increase in seizure threshold: â!œthe seizurethreshold of patients is related to their seizure duration.Higher thresholds are associated with shorter seizures.â! !(Sackeim et al., 1987c). Kales et al. (1997) found nocorrelation between the 37% fall in seizure duration over acourse of ECT (which they designated an â!œanticonvulsantâ! ! effect) and the amount of clinical improvement indepression scale scores, a result subsequently confirmed byLisanby et al. (1998).

Gamma-Aminobutyric Acid (GABA)Because depressive illness has been associated with a GABAdeficit (Petty, 1995; Shiah and Yatham, 1998), andGABAergic functions are often enhanced by anticonvulsants,it was natural to question whether GABA plays any role inECT-induced seizures. A critical test was provided by a studyin which 7 depressed patients who served as their owncontrols provided plasma GABA levels during each of 4ECTS: 2 genuine bilateral brief-pulse ECTs and 2 sham ECTs(Devanand et al., 1995). There was no difference betweenthe effect of genuine and sham ECT on plasma GABA levels,and thus no support for the anticonvulsant hypothesis.

Data reported in the same article from a larger samplestudied at a different study site, without a sham ECT control,did indeed show an effect of ECT on plasma GABA levels, butin the opposite direction from that predicted by theobservation of reduced plasma GABA levels in depression(Petty et al., 1992): plasma GABA levels were reduced forabout an hour

after each ECT, and were lower after, rather than before,the treatment course. A somewhat strained interpretation ofthe results succeeded in bringing them into line with thetheory, but not without doing some damage to thehypothesis that ECT enhances GABAergic functioning (Mann,1998).

Naloxone and Seizure DurationOne theory of the neurochemical basis of the seizure-threshold increase with ECT invokes an opiod mechanism,based partly on studies demonstrating that the potentnarcotic antagonist and opiate receptor blocker, naloxone,reduces the anticonvulsant activity of opioids in rats(Tortella et al., 1989). If such a mechanism obtained inman, then naloxone should lengthen ECT-induced seizures;however, it does not (Sperling et al., 1989; Rasmussen and

P.219

Pandurangi, 1999).

ConclusionAlthough the anticonvulsant hypothesis is not falsified bythe negative results of the several studies reviewed above(only the isolation and identification of an endogenousanticonvulsant specifically associated with ECT-inducedseizures in man, followed by the demonstration thatblockade of this anticonvulsant did not also block clinicalimprovement, could do that), its viability is seriouslyimpaired.

The Seizure GeneralizationHypothesisThis hypothesis is different from all others concerning themechanism of action of ECT because (at least until veryrecently) it does not invoke neurochemical mechanisms, butrather neurophysiological ones. Although no one doubts thatneurochemical mechanisms underly all neurophysiology, thehypothesis does not require the elucidation of suchmechanisms for its vi -abilityâ!”it is one step removed fromthemâ!”and to that extent, it is less specific. This lack ofspecificity, however, is at least partly responsible for thefact that the hypothesis has remained strongly in therunning while vir tually all others have fallen by thewayside, falsified by a single, critical experiment.

Stated in a nutshell, the seizure generalization hypothesisproposes that the more (and more efficiently) that ECT-induced seizure activity spreads throughout the brain, thebetter is the clinical antidepressant response.

The hypothetical construct of seizure generalizationintegrates diverse findings reported in the complexrelationships among the technical aspects of ECTadministration, observed physiologic (includingelectrophysiologic) responses, and therapeutic impact(Abrams, 199la). As noted previously, seizure generalizationis proposed to be manifested in ictal EEG amplitude, thedegree of postictal EEG suppression, the extent of ictal andpostictal

EEG coherence, the peak and duration of the heart-rateresponse, the dura tion of the induced seizure, the amountof prolactin released, and the inter correlation amongdifferent estimates of seizure duration.

The view that EEG seizure activity in primary generalized

epilepsy (the model most closely resembling the ECT-induced seizure) is driven by a centrencephalicâ!”presumably thalamicâ!”pacemaker is an old one that hasrecently been revived in several animal models(Niedermeyer, 1996) and is again invoked by someinvestigators of clinical epilepsy (e.g., de Curtis andAvanzini, 1994). Subcortical driving of generalized seizurespread via secondary bilateral synchronyâ!”withoutspecifying a thalamic origin of the driverâ!”is also described(Yoshinaga et al., 1996). However, Niedermeyer (1996)asserts that, in the case of spontaneus epileptic seizures, â!œEEG evidence indicates a superior frontal origin ofbilateral-synchronous spikes and spike-waves; depth EEGrecordings in patients have failed to demonstrate primarythalamic spike generation.â! ! Such a view is more consistentwith the spread of seizure activity from primary foci inducedin the frontal lobes as is proposed by the intracerebralcharge density model described below.

The existence of a subcortical pacemaker for this bilateralsynchrony is supported by the increased ictal EEG coherencethat is reported during, compared with following, the seizure(Krystal, Weiner, and Coffey, 1995). The substantialintercorrelations observed during ECT-induced seizuresamong the durations of the ECT-induced motor activity, EEGspike activity, and tachycardia response (Swartz and Larson,1986), also support this view: because these variablesderive from anatomically separate brain areas, theirconcordance must result from their being driven from acommon source, presumably subcortical.

The interictal EEG may also reflect seizure generalizationwith ECT. The same treatment factors that are associatedwith increased ictal amplitude and coherence, and postictalsuppression (e.g., bilateral relative to unilateral placement,higher versus lower stimulus intensity), are also associatedwith increased abundance of EEG slowing (particularly in thedelta frequency band) on non-ECT days and for a varyinginterval after the treatment course (Weiner et al., 1986a;Sackeim et al., 1996). As noted in Chapter 4, Sackeim et al.(1996) found that increased antidepressant activity of ECTwas associated with increased interictal slowing in anterior,prefrontal regions, a finding that dovetailed nicely with theirearlier report that ECT-induced cerebral blood flowreductions in the same anatomical areas were positivelyassociated with ECT reponse (Nobler et al., 1994). Thedecreased regional glucose metabolism observed in thefrontal lobes after bitemporal ECT, although not examinedfor its relation to outcome, is also consistent with these

P.220

results (Nobler et al., 2001).

Role of the Diencephalon

Although the EEG measures electricalactivity generated by cortical neurons, itcan also reflect subcortical activity. Thesubcortical influences are

seen indirectly, with diencephalic structuresmodulating the degree of synchronizationamong cortical pyramidal cells. (Krystal andWeiner, 1999)

The diencephalic hypothesis of ECT (Abrams and Taylor,1976b; Abrams, 1997) now finds itself nested comfortablywithin the seizure generalization hypothesis as a specialfocus of the latter. The diencephalic hypothesis posits thatfor the fully developed therapeutic effect of ECT to becomemanifest, the ECT-induced seizure must be sufficientlygeneralized to involve those diencephalic centers implicatedin the regulation and modulation of appeti tive behaviors,diurnal rhythms, hormone release, and physiological homeostasis.

The proposition that convulsive therapies reduce thevegetative changes in depression through an effect ondiencephalic structures is an old one (Carney and Sheffield,1973; King and Listen, 1990). Pollitt (1965) has detailed theway in which the diencephalon directly controls a variety ofbiological functions that are frequently altered in depressionthrough its hypothalamic releasing factors and pituitary andautonomic connections, pointing out that melancholia isfrequently associated with disturbances in appetite,gastrointestinal functions, sleep, temperature regulation,secretory patterns, diurnal rhythms, menstrual cycle, libido,and cardiovascular func tion, all more or less regulated bydiencephalic centers and virtually always restored towardnormal by a course of ECT.

Dr. Michael Taylor and I tested the diencephalic hypothesisin a series of studies that systematically manipulatedelectrode placement for unilateral, bilateral, and bifrontalECT in an attempt to relate presumed diencephalicstimulation to the efficacy of ECT in depression (Abrams andTaylor, 1973, 1974b, 1976b).

In the third, and to me, most compelling of these studies

P.221

(Abrams and Taylor, 1976b), we randomly assignedmelancholic patients to treatment with either bitemporal ECTor with left and right unilateral ECT administeredsimultaneously via 2 separate ECT devices. The percentimprovement in blindly obtained Hamilton depression scalescores after 6 ECTs was 75% for bitemporal ECT, comparedwith 35% for simultaneous left and right unilateral ECT (p <

0.01); significantly more patients in the left -right uni lateralECT group went on to receive additional ECTs as prescribedby a psychiatrist blind to treatment assignment.

Because dose and stimulus parameters were identical forboth methods, and all patients received simultaneousbilateral stimulation over both temporal areas, the markedtherapeutic advantage for the bitemporal ECT group couldonly have resulted from differences in the intracerebraldistribution of either the stimulating current or the inducedseizure, presumably through the greater diencephalicstimulation of bitemporal ECT predicted by the model ofWeaver, Williams, and Rush (1976).

Further evidence that bitemporal ECT stimulates diencephaliccenters more than right unilateral ECT comes from ourdemonstration of a greater duration of ECT-induced heart-rate elevation with the former method (Lane et al., 1989).

There is little doubt that ECT-induced heart-rate elevation isdiencephalically-driven via the brainstem, rather than causedby circulating catecholamines, because the full hemodynamicresponse to ECT occurs in the complete absence of theadrenal glands (Listen and Salk, 1990). Moreover, oursubsequent finding (Swartz, 1994b) of a differential effect ofleft and right unilateral ECT on the persistence of ECT-induced heart-rate ele vations can only be explained by alateralized diencephalic effect on the brainstem cardio-acceleratory center.

The most readily tested prediction of the diencephalichypothesis of ECT is that ECT-induced seizures that are notassociated with a well-developed tachycardia response willhave a reduced clinical antidepressant response. Thisproposition was tested by Swartz (2000) in 24 patients whoserved as their own controls. Each patient was givenstandard (half -age) or high-dose (4Ã—-5Ã— age) stimuli onseparate days, administered via a variant of bifrontalelectrode placement (Swartz, 1994). Seizures induced withhigh-dose stimulation yielded higher peak heart rates withinpatients, and those patients whose heart rates weremaintained closer to their peaks required fewer ECTs as

P.222

determined by a psychiatrist blind to heart rate response;that is, they had a better or faster clinical response. Thus,the peak heart rate with ECT is both sensitive to stimulusdose, and directly associated with clinical antidepressantefficacy.

Although an apparently powerful combination of ictal andinterictal events can thus be invoked in support of theseizure generalization hypothesis of ECT, its confirmationwill require the demonstration that one or more of theseictal or interictal variables not only correlates with thedegree of ECT-induced clinical improvement in depression,but is prerequisite for such improvement to occur.

Prefrontal ModelOn the basis of investigations into the differential efficaciesof titrated unilateral and bitemporal ECT given at varyingmultiples of the threshold dose, it has been proposed(Sackeim and Mukherjee, 1986; Sackeim, 1994a) that theefficacy of ECT is related to the intracerebral spatialdistribution of charge density and resultant initation ofseizure activity, with unilateral ECT producing an evendistribution of charge density across the anterior two thirdsof the stimulated hemisphere, and bitemporal ECTconcentrating its effects in the prefrontal regions. In thisview, increasing the stimulus intensity with unilateral ECTresults in its becoming more like bitemporal ECT in thespatial distribution of its charge density, thus enhancingparticipation of the prefrontal regions in seizure initiation,which is theorized to be central to the antidepressantefficacy of ECT.

The reported greater efficacy of bifrontal ECT than bothbitemporal and right unilateral ECT (Letemendia et al.,1993) is quite consistent with this model, as are theobservations of an association between clinical ECT response

and both greater interictal prefrontal slowing (Sackeim etal., 1996) and greater prefrontal cerebral blood flowreductions (Nobler et al., 1994). If Niedermeyer's (1996)description of the superior frontal origin of the bilateral-synchronous spikes and spike-waves of primary generalizedepilepsy is correct, this would also be consistent with themodel. The prefrontal model is the simplest and mostpragmatic of those considered above, and thus cannotaccount for some of the clinical response-associated featuresexplained by the more complex models (e.g., therelationship of EEG postictal suppression or the tachycardia

response to clinical outcome with ECT).

But there is absolutely no reason why the models cannot belinkedâ!” in fact, it makes great sense to combine thembecause they represent different phases of the ECT-inducedseizure: the prefrontal model specifies the criticalanatomical sites for seizure initiation, and the seizuregeneralization model specifies the EEG characteristics ofsubsequent seizure spread and the anatomical requirementfor diencephalic participation. This unified neurophsyiologicaltheory of ECTâ!”which I call the anatomico-ictal theoryâ!”is more heuristic than any of its individual components andis more likely to be correct, just as multifactorial geneticmodels of mental illness are more likely to be correct thansingle-gene models.

To summarize, the anatomico-ictal theory of ECT actionspecifies that those ECT-induced seizures will have thegreatest clinical antidepressant impact that are initiated inthe prefrontal regions of the brain and then spreadmaximally throughout the cortex and subcortex, involvingdiencephalic centers in particular. The theory includes theproposition, for example, that seizures initiated prefrontallygeneralize more effectively throughout the brainâ!”andtherefore more effectively involve diencephalic participationâ!” than seizures initiated at other cortical sites; additionalpropositions of the same type are readily generated by themodel.

Other modern theories of the mode of action of ECT havenot advanced much beyond the general understanding of theproblem that was already widely proposed by the early1960s:

â!" many workers consider that one type ofdepressive illness, called variouslyendogeneous or physiological depression,reflects a failure of hypothalamic centresconcerned with instinctive drives andadrenal stress mechanisms, and it has beenfrequently suggested that E.C.T. acts bystimulating these mid-brain centres.(Hodges et al., 1964)

This notion that convulsive therapies worked to reverse thevegetative changes in depression by providing â!œmassageof diencephalic centersâ! ! was already clearly expressed by1939 (King and Listen, 1990), the year of the first English-language publication on ECT. Observations, such as the one

P.223

offered by Weizman et al. (1987), that

It seems that ECT affects variousneurotransmitters and neuromodulators andthat the interactions among these systemsresult in the neuroendocrine changesobserved following an ECT course.

are no more helpful in explaining the mechanisms involvedthan the statement that â!œclimate results from a complexinteraction of cosmological, meteorological, and geologicalforcesâ! ! is in explaining the weather. It is further notablethat a comprehensive review of the data supporting aneurobiologic mechanism for the therapeutic effects of ECTin man (Kapur and Mann, 1993) failed to uncover a singlecandidate for the magic bullet, concluding that â!œDespiteconsiderable investigation, the mechanism of antidepressant(sic) effect of ECT is still a mystery.â! !

It should be clear from this brief review that no coherentgeneral neu rochemical theory of the action of ECT yetexists. King and Listen (1990) have, in my view, correctlysummarized the situation:

Taken together, however, the myriadbiochemical sequelae of electroconvulsivestimulation, the gaps in our knowledge oftheir functional correlates, and thedifficulties in extrapolating from animal toman or from normal to pathophysiologicalstates makes it tenuous to entrust thetherapeutic efficacy of ECT to theobservable change in specific transmitter,receptor, or system.

The broadly drawn hypotheses that ECT exerts its beneficialeffects in depression through hypothalamic mechanisms(Carney and Sheffield, 1973; Abrams and Taylor, 1973,1974b, 1976b; Fink and Ottosson, 1980) or byneurotransmitter release, otherwise unspecified (King andListen, 1990), are not specific enough to be heuristic, orâ!”in the latter case, even testableâ!” and the more specifichypotheses (Abrams, 1986b, 1990; Staton et al., 1988; Finkand Nemeroff, 1989; Fink, 1990; Sackeim, 1994b; 1999b)are either too narrow, premature, unconfirmed, or just plainwrong.

Modern ECT researchers, regardless of their species ofpredilection, have hardly more insight into the relationshipbetween brain biological events and treatment response inECT than they did at the time of the first edition of thisbookâ!”which is to say, very little, indeed. Moreover,modern neurohumoral theories of the action of ECTâ!”evenas formulated by sophisticated investigators with impeccablecredentialsâ!”have not surpassed in conceptual elegance the18th-century claim that things burned because theycontained phlogiston; ECT awaits its Lavoisier.




> Table of Contents > Chapter 12 - Patients' Attitudes, Medicolegal

Considerations, and Informed Consent

Chapter 12

Patients' Attitudes, MedicolegalConsiderations, and InformedConsent

Patients' Attitudes TowardElectroconvulsive TherapyGomez (1975) was the first to examine the incidence andseverity of subjectively experienced side effects of ECT. Shetried to understand why, as she believed, â!œmany patientsand their relatives view the prospect of ECT with horror.â! !She interviewed 96 patients, most of them depressed, 24hours after each treatment, for a total of 500 consecutivetreatments, of which 420 were given with bilateral electrodeplacement. The self-reported incidence of side effects wasremarkably low: Muscle pain topped the list at 8.2%, withsubjective memory impairment and headache occurring onlyabout 3% of the time. Surprisingly, subjective complaintswere no less frequent in patients receiving unilateralcompared with bilateral ECT. The aspect of ECT mostdisliked by patients who were interviewed after they hadreceived 3 or 4 ECTs related primarily to fearâ!”mostfrequently, a fear of permanent impairment of memory orintellectual abilities, followed about equally by fear ofentering the ECT room, fear of death or serious damage,and the apprehension experienced while simply waiting toreceive treatment. A quar ter of the sample expressed nospecial fear or dislike of ECT.

Hillard and Folger (1977) administered a questionnaire topatients receiving ECT on 2 different state mental hospitalwards and found that attitudes toward this treatment were

P.225

significantly more favorable on the ward in which ECT wasmore frequently given, despite the fact that neither hos pitalmade any effort to brief patients beforehand on the safetyor therapeutic aspects of ECT.

Freeman and Kendell (1980) conducted a systematic study ofthe experiences and attitudes toward ECT among 166patients who had received treatment up to 6 years earlier.Almost half of the patients either had no particular feelingsbefore receiving ECT or were reassured and pleased thattreatment was about to begin. The vast majority had nofears about the procedure itself, or of sustaining braindamage as a result. More than 80% of the patients said thathaving ECT was no worse than a visit to the dentist and50% actually preferred ECT (65% said they would be willingto have

it again if necessary). Unique to this study (see Chapter 10)was the fact that 30% felt their memory had never returnedto normal after ECT. Only 1.2%, however, felt that thetreatment had worked by making them forget theirproblems; 90% of the patients said that either the consentprocedures had been adequate or were satisfied to defer tothe doctor's recommendation.

Hughes, Barraclough, and Reeve (1981) used the samequestionnaire to interview 72 patients who received ECT forsevere mental illness: 83% said they had improved and 81%were willing to have ECT again. Slightly more than halfthought a visit to the the dentist was more distressing.Almost half reported memory impairment after treatment,and 18% said it was still present when interviewed. About10% thought the consent procedure should be changed, andmost thought the decision about ECT should be left to thedoctors.

Results similar to those obtained in the preceding 2 studieswere re ported by Kerr et al. (1982).

A prospective study of patients' attitudes regarding ECT wasconducted by Baxter et al. (1986) in Berkeley, California.More than 50% of patients who were about to have ECT forthe first time were concerned about experiencing memoryloss, having a seizure, or sustaining brain damage, andabout 25% were worried about the anesthesia, the

P.226

possibility of pain, and the use of electricity. After a courseof ECT, most of these patients felt that they had receivedadequate information to decide about having the treatment,that they were helped by it, and that their decision to haveit had been a good one. In contrast, patients who hadreceived ECT in the past (some of them presumably withoutmodern anesthesia techniques) were more fright ened aboutthe treatment and pessimistic about its outcome, althoughall consented to a new course of ECT.

Pettinati et al. (1994) surveyed patient attitudes toward ECTin a sample of 56 depressed patients before and after ECT,compared with 22 depressed patients never treated withECT. Virtually all (98%) of the ECT-treated group statedthey would be willing to have ECT if they became depressedagain, compared with 70% in the no-ECT group. In astriking confirmation of the earlier studies cited above, amajority (62%) of the ECT-treated group rated theexperience less upsetting than having a tooth pulled,compared with 14% of the no-ECT group. The differencebetween the 2 groups in favorable attitudes toward ECT wasmaintained 6 months later.

Walter, Koster, and Rey (1999a) conducted a 53-itemtelephone survey of all available patients in New SouthWales, Australia, who had received ECT as adolescents (age< 19 years) during 1990-1998. Of the 26 respondents, halffelt ECT had been helpful, three quarters believed theeffects of their illness had been worse than either ECT ordrug therapy, and most felt that the side effects of ECT anddrug therapy had been similar. The vast majority consideredECT to be a legitimate treatment and were willing both tohave it again if needed, and to recommend it to others.

These same investigators (Walter, Koster, and Rey, 1999b)conducted a similar interview with 28 parents of theadolescent patients who had received

ECT and found that 61% felt the treatment had beenbeneficial. Just as for the adolescent ECT recipientsinterviewed, the vast majority of their parents also said theywould support a medical recommendation for their child tohave ECT again if it were needed, and would advise otherparents to do the same.

Goodman et al. (2000) administered a 44-item survey oftreatment attitudes and satisfaction to 24 patients duringand shortly after a course of ECT, and to 24 psychiatricpatients who had never received ECT. Attitudes toward ECTwere significantly more favorable among those who had received ECT than among controls, and 91% of the ECTpatients endorsed the statement: â!œI am glad that Ireceived ECT.â! !

In view of the fact that most of the several hundred patientsinterviewed in the studies described above and in a largerreview by Freeman and Cheshire (1988) had rather positiveviews about ECT and did not find the treatment especiallyfrightening, upsetting, painful, or unpleasant, how is itpossible to account for the apparently widespread negativepublic image of ECT (Frankel, 1982; National Institutes ofHealth, 1985) and the history of leg islative restrictions onits use (Winslade et al., 1984)? I believe there are severalcontributory factors.

Past Abuse of ElectroconvulsiveTherapyFink (1983) points out that professional concerns about ECTare hardly new, citing the critical 1947 report of the Groupfor the Advancement of Psychiatry (which began with theintroductory statement â!œIn view of the reportedpromiscuous and indiscriminate use of electroshock therapy,your Committee on Therapy decided to devote its firstmeeting to an evaluation of the role of this type of therapyin psychiatryâ! !). In 1972, Dr. Milton Greenblatt, thenCommissioner of Mental Health for the State ofMassachusetts, organized a Task Force to Study andRecommend Standards for the Administration of Electro-Convulsive Therapy (Frankel, 1973). This Task Force wascreated in response to local allegations about the excessiveuse of ECT, particularly for outpatient maintenance therapy,by one or two zealous practitioners, as later exemplified inan article by Regestein, Murawski, and Engle (1975) entitledâ!œA case of prolonged, reversible dementia associated withabuse of electroconvulsive therapy.â! ! These authorsdescribed in detail the appalling (but apparently ultimatelyreversible) cognitive consequences in a 57-year-old womanwho had received a â!œmaintenanceâ! ! ECT every Saturday

P.227

morning for more than 2 years (a procedure I believe wouldbe more accurately described as â!œremunerative ECTâ! !).They also briefly described the devastating effects of asimilar treatment schedule on an executive, concluding withthe statement that â!œthe ready insurance payments forany number of ECT further encourage errors in judge mentconcerning the efficacy of such treatment.â! !

The Massachusets Task Force on ECT responded by limitingthe number of ECTs to be given in a single treatmentcourse, and further constrained

the use of outpatient and maintenance ECT. Indeed,concerns over the â!œpotential for misuse and abuse of ECTand the desires to ensure the protection of patients' rightsâ! ! were central to a United States government-sponsoredconference to assess the place of ECT in medical practice(National Institutes of Health, 1985), described in detail inChapter 1.

Inadequate Consent ProceduresFor many years, psychiatry lagged behind other medicalspecialties in developing and promulgating the doctrine ofinformed consent. This is likely to have resulted in largemeasure from the nature of the patients treated: theirfrequent impairments of perception, judgment, andreasoning often required others, including their physicians,to make decisions for them. Several articles written on thesubject during the decade ending in 1985 (Salzman, 1977;Culver, Ferrell, and Green, 1980; Gilbert, 1981; Frankel,1982; Senter et al., 1984; Winslade et al., 1984) made itabundantly clear that a sea change was in orderâ!”therewas, in fact, little choice in the matter as state after stateintroduced mental health legislation with the same 3themes: (1) voluntary patients may not receive involuntarytreatment; (2) competent patients, even if involuntary, maynot receive involuntary treatment; and (3) involuntary,incompetent patients may only receive involuntary treatmentunder court order (or, in some states, in the presence of adocumented real and immediate threat to life or limb).

Misrepresentation ofElectroconvulsive Therapy in the

P.228

Electroconvulsive Therapy in theMediaAlthough the depiction of ECT in the 1940s movie The Snake

Pit was accurate for that time, the portrayal of electricitystiffening MacMurphy's body in One Flew Over the Cuckoo's

Nest was not, because the nature of ECT had beencompletely transformed by modern anesthesia techniquesdecades before that film was made. Nevertheless, bothfilms, and several others that followed them, createdindelible images of the apparent brutality of ECT and thecallousness of those who administer it, inculcating negativefeelings towards ECT (and even refusal of ECT) among itspotential recipients (Walter, 1998).

Whether because public opinion is less easily swayed bysuch representations than many have feared or because suchportrayals are occasionally effectively countered by morebalanced media presentations, public surveys among thosewho have not received ECT and are not about to, have generally revealed an unexpectedly overall positive attitudetoward the treatment (Kalayam and Steinhart, 1981; Baxteret al., 1986).

A particularly favorable view of ECT was presented by thepopular and widely respected talk show host Dick Cavett intelevision interviews and

news articles in the early 1990s. It is hard to overestimatethe positive effects on public opinion of the followingstatement, which he made concerning his personalexperience of receiving ECT for a severe depressive illness:

In my case, ECT was miraculous. My wifewas dubious, but when she came into myroom afterward, I sat up and said, â!˜Lookwho's back among the living.' It was like amagic wand. (News and Notes, 1992).

Increased Awareness of Patients'RightsFrankel (1982) pointed out that the introduction of civilrights concerns into the mental health controversy over themedicolegal and ethical aspects of ECT has transformed â

!œwhat might have been a narrow medical debate into apolitical challenge involving litigation and legislation, whilejurists and legislators largely unfamiliar with ECT have beendrawn into the debate.â! !

Although there is understandably great interest and concernover the issue of procedural constraints on civil commitment(e.g., Stone, 1977; American Psychiatric Association, 1978;Frankel, 1982), this is a problem that does not directlyimpinge on the administration of ECT, which comes into playonly after a patient's legal status has been determined. Atthis point, there are 2 major ethical and legal concerns: Thecompetency of a patient to consent to treatment and thecomponents of valid informed consent.

Competency IssuesThe general issue of legal competency is enormouslycomplex and far beyond the scope of this volume. Roth,Meisel, and Lidz (1977), from the Law and PsychiatryProgram at Western Psychiatric Institute, have pointed outthe dearth of legal guidance on the question of competencyto consent to medical treatments. They note that a personmay at the same time be considered competent for somelegal purposes and incompetent for others; a person is notjudged incompetent merely because of the presence ofmental illness; the consent to treatment of an incompetentperson does not validly authorize a physician to performsuch treatment; and a physician who does not takereasonable steps to obtain legal authorization for treating anincom petent patient who refuses treatment may be heldliable. These authors agree that, although there is no singlevalid test for competency,

It has been our experience that competencyis presumed so long as the patientmodulates his or her behavior, talks in acomprehensible way, remembers what he orshe is told, dresses and acts so as toappear to be in meaningful communicationwith the environment, and has not beendeclared legally incompetent.

The ability of a patient to understand the risks, benefits,

P.229

and alternatives to treatment (including no treatment atall), however, is increasingly becoming

the commonly applied standard of competence for consentingto ECT. Interestingly, although the critical element of this â!œtestâ! ! of competence is the patient's ability tocomprehend the elements that the law presumes to be apart of treatment decision making, such decision makingneed not be rational to be legally acceptable. This isexemplified by the patient of Roth, Meisel, and Lida (1977)who fully understood the nature of the ECT that was beingoffered to her, but accepted it because she hoped it wouldkill her.

The right to refuse treatment is, of course, implicit in theconcept of competency. Even committed psychotic patientswho are deemed by the law to require continued involuntarypsychiatric hospitalization and treatment may refuse ECT, asshown by N.Y. City Health and Hospitals Corp. vs. Stein.Paula Stein had been committed to Bellevue Hospital bycourt order to receive whatever course of treatment thepsychiatric staff deemed advisable, including ECT, regardlessof her consent and without the necessity of prior judicialapproval. Because of a recently introduced state mentalhygiene law, however, the hospital nevertheless attemptedto obtain the pa-tient's consent; when this was notforthcoming, the hospital petitioned the court to authorizetreatment. In a landmark decision, the court denied thepetition, concluding that although

she is sufficiently mentally ill to requirefurther retention â!" that determinationdoes not imply that she lacks the mentalcapacity to knowingly consent or withholdher consent to electroshock therapy â!"Itdoes not matter whether this court wouldagree with her judgement; it is enough thatshe is capable of making a decision,however unfortunate that deci sion mayprove to be.

The Doctrine of Informed Consent

P.230

The criterion of the ability of the patient to understand thenature of the treatment being offered is, of course, fullyconsistent with the doctrine of informed consent (Meisel,Roth, and Lidz, 1977). One of the 2 1960 landmark cases ofthe informed consent doctrine, Mitchell v. Robinson, involveda psychiatric patient who had sustained spinal fractures as aresult of ECT. Even though the patient's consent had beenobtained and the physician was not deemed negligent in theperformance of the procedure, the court nevertheless ruledthe consent invalid (and the physician liable) because hehad not adequately informed the patient beforehand of thehazards of treatment.

Meisel, Roth, and Lidz (1977) pointed out that in order forconsent to treatment to be valid, the patient must be ableto act voluntarily; must be provided with a particular amountof information concerning the treatment (specifically, itsrisks, benefits, and alternatives); must have the capacity tounderstand the information provided (that is, the patientmust be competent); and must actually make a decisionregarding the treatment (although consent

may sometimes be implied by the patient's passiveacceptance of the treatment).

It was apparent from the testimony from the floor at theNational Institute of Mental Health Consensus DevelopmentConference that the anti-ECT sentiment derived almostentirely from two kinds of patients: Those who had receivedinvoluntary ECT and those who had received ECT voluntarilybut without sufficient warning about the possibility ofpersistent memory impairment. Psychiatrists who wish tocontinue having ECT available to them as a therapeuticmodality should therefore do 3 things: (1) discontinue theuse of sine-wave devices in favor of those administering abrief-pulse, square-wave stimulus; (2) refrain from givinginvoluntary ECT except under court order; and (3) carefullyinform potential ECT candidates and their families about thepossibility, however infrequent, of permanent memory lossafter ECT, especially with sine-wave bilateral electrodeplacement. Moreover, all patients should be started onsubstantially suprathreshold unilateral ECT unless theirclinical condition is so severe as to warrant bilateral ECT atthe outset (Abrams and Fink, 1984; American Psychiatric

P.231

Association, 1990). Following these simple guidelines willhelp defuse the ECT controversy and ensure the continuedavailability of ECT for the many patients who are likely tobenefit from no other form of treatment.

When obtaining consent for ECT, the physician shouldpresent the advantages and disadvantages of the proposedtreatment in sufficient detail to permit the patient a trulyinformed choice, without either exaggerating the potentialbenefits (e.g., promising favorable results) or undulyalarming the patient with a litany of every conceivable risk,no matter how remote. A good relationship with both patientand family is essential in this process, which is ultimatelybased on personal trust. Of particular help are videotapesthat have recently become available that portray therationale and procedures of ECT in a straightforward andunbiased fashion (Baxter and Listen, 1986; Frankel, 1986;Ries, 1987). A clearly written, concise consent form is alsorequired, one that can be readily understood by patients whohave only a limited ability to focus their attention. Theexample in the Appendix mod ified from Abrams and Swartz(1991) can be adapted to suit the needs of a particularphysician or institution.

Biomedical Ethics andElectroconvulsive TherapyOttosson (1992, 1995) has reviewed the principles ofbiomedical ethics as they apply to ECT. Autonomy,beneficence, nonmaleficence, and justice are the basicconcepts of biomedical ethics. With regard to ECT, theprinciple of respect for individual autonomy is satisfied byadherence to the doctrine of informed consent as presentedabove; the principle of beneficence is satisfied by theresults of the studies of ECT versus sham ECT, and ECTversus drug therapy (Chapter 2); the principle ofnonmaleficence is satisfied by the results of the studies ofECT and brain damage (Chapter 4), ECT in the

high-risk patient (Chapter 5), and the memory and cognitiveeffects of ECT (Chapter 10). The principle of justice is moredifficult to apply with regard to ECT as it concerns, amongother things, effects on societyâ!”for example, whether aprocedure judged as ethically right for one person has

negative consequences for others (Reiter-Theil, 1992).Setting this somewhat murky principle aside, then, it iscrystal -clear that ECT as a procedure is ethical whenadministered with informed consentâ!”but is compulsory ECTever ethical?

This raises the issue of paternalism and relates to questionsof competence, guardianship, and judicial orderâ!”that is,whether society is justified in treating patients against theirwill for their own good or the good of others. This questionhas historically been answered in the affirmative for othermedical procedures (e.g., the forced medical isolation oflepers, the quarantine for smallpox) and other psychiatrictherapies (e.g., involuntary commitment), and the abovediscussion of the principles of beneficence andnonmaleficence shows that there is no objective reason toview compulsory ECT differently. Indeed, the vast majorityof requests made by psychiatrists to administer involuntaryECT are approved by the courts (Srinivasaraghavan andAbrams, 1996), emphatically demonstrating that societyviews such a procedure as appropriate.

In my view, a far more important question is whether it isethical to withhold ECT from a patient who wants it andmight benefit from it, as daily occurs in the hundreds ofstate and other hospitals in the United States that do notoffer this form of therapy. If such hospitals do not have aneffective referral procedure for sending the patientelsewhere to receive ECT, then they are in the clearlyunethical position of refusing to administer a therapy ofdemonstrated efficacy to a patient who requests it andwhose illness might benefit from it, and it alone.

Malpractice LitigationSlawson and Guggenheim (1984) reviewed the totalmalpractice experience of the American PsychiatricAssociation (APA) professional insurance program from 1974to 1978 and found that ECT accounted for only 15 out of71,788 claims (0.02%) and only 0.0051% of the loss($45,000 out of $876 million), an experience that led toreduction, and then elimination, of the program's ECTsurcharge. In a more focused review of the Association'ssubsequent ECT malpractice experience to 1983, Slawson(1985) found only 16 cases that related to ECT: 15 were

P.232

settled out of courtâ!”8 in favor of the psychiatrist, 7 infavor of the patientâ!”and the only case that went to trialwas won by the psychiatrist. Moreover, of the 7 casessettled in favor of the patient, 4 involved amounts thataveraged only $1000 (2 of which were payment of dentalbills).

The 3 cases that required larger settlements, averaging$110,000 per settlement, included 2 claims of wrongfuldeath (one in which a psychiatrist

administered the anesthesia and failed to resuscitate thepatient when he developed laryngospasm), and one entirelyunrelated to the administration of ECT: failure to diagnosecancer. (The patient improved with ECT but was found ayear later to have carcinoma of the tail of the pancreas.)

This unusually favorable ECT malpractice experience wasconfirmed in subsequent reviews (Slawson, 1989, 1991) ofthe APA's liability program during the years 1984 through1990, after it had switched from commercial coverage tomanaging its own claims. During that interval there werejust 6 malpractice claims involving ECT, only 2 of whichwere lawsuits, and the only payment to date had been forthe dental bill of 2 patients who claimed they had sustaineddental damage during ECT. No claims involving ECT went totrial.

Dr. Slawson's reviews affirm the generally small liabilityincurred by physicians who administer ECT, although onlyabout 60% of American psychiatrists are insured through theAPA. I am acquainted as an expert witness with 7 additionalECT-related cases during and since the period most recentlyreviewed by Slawson. Of the 3 cases tried, the jury foundfor the psychiatrist in two, and the third was settled infavor of the plaintiff after the trial ended in a hung jury. The2 trial cases won by the psychiatrist are instructive(Abrams, 1989a) and demonstrate that juries are moreimpressed by facts than emotional claims of plaintiffs andtheir expert witnesses.

In the first case, a 32-year-old housewife claimed that acourse of 6 bilateral, brief-pulse ECT received 2 years earlierfor an episode of major depression had caused her to sufferbrain damage with permanent memory loss. In support of

P.233

her claim, the plaintiff presented the expert opinions of aclinical psychologist, who testified that his testingdemonstrated recently acquired brain damage, and aneurologist, who testified that an MRI performed a year afterECT showed cortical atrophy and ECT-induced bitemporalbrain hemorrhages. The defense impeached the reliabilityand validity of the neuropsychological testing employed tomake the diagnosis of recently acquired brain damage andcited published data failing to show any deleterious effectsof ECT on MRIs obtained before and after ECT. Trialtestimony took 5 days, and the jury deliberated 2 hoursbefore finding for the defendant psychiatrist.

In the second case, a 78-year-old retired architect claimedthat a course ECT he had received for the treatment ofmajor depression had caused him to suffer brain damageand permanent memory loss and had also induced a severeand totally disabling progression of preexisting Parkinson'sdisease. The plaintiff's expert, Dr. John Friedberg, held thatECT had been contraindicated because of the preexistingdiagnosis of Parkinson's disease, and that the plaintiff hadsuffered a stroke during ECT, leaving him with a permanentorganic mental syndrome. The defense cited the extensiveliterature supporting the use of ECT to treat depressedpatients with Parkinson's disease, and presented the resultsof its own testing of the plaintiff, which demonstratednormal intellectual functioning. Trial testimony took 5 days;

the jury deliberated less than an hour before finding for thedefendant psychiatrist.

My experience with these and several additional cases hasled me to identify the following key elements as common tomany suits:

1. The plaintiff claims ECT-induced brain damage withresulting permanent memory loss and social andeconomic disability, although objectiveneuropsychological and memory testing reveals normalto superior functioning.

2. The doctor is charged with

a. improper diagnosis

b. failing to document a valid indication for ECT

c. ignoring concurrent medical conditions.

d . rushing to treatment

e. failing to obtain informed consent

f. failing to advise the patient of the possibility oflong-term memory loss

g . improperly administering or documentingtreatments

3. In the past, expert witnesses for the plaintiff oftenincluded Drs. Peter Breggin (psychiatrist) or JohnFriedberg (neurologist), both of whom authored booksand articles that are highly critical of ECT (Friedberg,1976, 1977; Breggin, 1979, 1980). Dr. Breggin hasalso appeared on virulently anti-ECT talk shows intandem with Dennis Clarke, president of theScientology-funded Citizen's Commission on HumanRights.

Both Dr. Friedberg and Dr. Breggin were afforded theopportunity to present their views at ECT conferencesconvened by the APA or the National Institute of MentalHealth. Dr. Friedberg presented his view of the evidence forECT-induced neuropathology at the annual meeting of theAPA in 1976, during a symposium held in parallel with anongoing fact-finding mission of the APA Task Force on ECT.His paper (Friedberg, 1977) concluded with the statementthat â!œit is well to reaffirm the individual's right to pursuehappiness through brain damage if he or she so chooses,â! !but that doctors should ask themselves whether they shouldbe offering it. Dr. Fred Frankel, chairman of the session(and of the first APA Task Force on ECT) characterized Dr.Friedberg's data as â!œcarelessly culled from the literatureand frequently reported inaccurately,â! ! and his presentationas inaccurate, indis criminate, inconsistent, andcontradictory (Frankel, 1977).

Dr. Breggin has been equally guilty of misrepresenting hissources. For example, in his review of the study of Globus,van Harreveld, and Wiersma (1943) of whether ECS inducedneuropathological changes in animals, Dr. Breggin claimed

P.234

that:

Globus et al. found extreme, permanentpathological changes in every single one oftheir animals, including â!œghostâ! ! cellsand other signs of dead and dying cellsthroughout the brain. (Breggin, 1979)

In fact, Globus, van Harreveld, and Wiersma (1943) foundprecisely the opposite:

Under high magnification (fig. 2A) there areno pathologic alterations visible in thenerve cells. These cells are of the usualform and display a normal amount of tigroidsubstance. Occasionally there is a cell inwhich the tigroid substance is not veryprominent, but this is probably due to thethinness of the section of these cells in theslide. For similar reasons, an occasional cellshows a faint outline of the cell body, oftenassuming the character of a ghost cell. Butwhen the aforementioned features arecompared with those found in a sectionobtained from a corresponding area of thebrain of a normal dog (fig. 2B) theoccasional cellular pseudo-defects will befound to be as common in the control as inthe experimental animal. (Globus, vanHarreveld, and Wiersma, 1943).

Two reviewers of his book (Breggin, 1979) characterized itas follows:

â!" personal opinions are presented as ifthey were scholarly deliberations. In fact,most scientific research concerning ECTcontradicts Breg-gin's hypotheses about theefficacy and adverse effects of thetreatment. Breggin repeatedly ignoresconsiderable recent literature on unilateral

P.235

ECT, as well as low-energy stimulation,both of which have dramatically reducedthe side-effect of memory loss. â!"Dr.Breggin's arguments fail because he usessupporting data uncritically andinaccurately. (Mandel, 1980)

and,

â!" both quantitatively and qualitatively,factual inaccuracies and notable absencesensure an exceedingly biased presentation.(Weiner, 1980c)

Several years ago, judicial decisions effectively neutralizedboth the value of Dr. Breggin's testimony to plaintiffs'lawyers and his undue influence on the public's view of thetherapeutic roles of ECT and psychopharmacology (Abrams,1995). Dr. Breggin's credentials and opinions as an expertmedical witness were scathingly rejected by the court in 2separate and unrelated lawsuits, one in which the plaintiffclaimed damages from ECT, and another in which theplaintiff claimed damages from a psycho pharmacologicagent, Halcion.

In the first case, involving ECT, 1 Judge Hilary Kaplan in theCircuit Court for Baltimore City issued a 17-page ruling tostrike Dr. Breggin's testimony from the record,characterizing his testimony as â!œa house of cards,â! ! â!œerroneous,â! ! â!œwithout factual basis,â! ! and â!œjustincredible,â! ! and concluded with the statement that

The court believes that this witness Breggindoes not have sufficient familiarity with thesubject matter in this case to express anopinion, and that opinion has no rationalbasis in fact.

In the second case,2 U.S. Magistrate Judge Crigler of theU.S. District Court, Western District of Virginia, rendered theopinion that:

Breggin has conducted no original research,has performed no personal reanalysis of thework of others, has prescribed Halcion onlyonce since 1983, has no academic trainingor regulatory experience and has neverparticipated in any FDA-related proceedingsaddressing what constitutes an adequatewarning. Simply put, the court believes thatDr. Breggin's opinions do not rise to thelevel of an opinion based on â!˜goodscience.' The motion to exclude histestimony as an expert witness should begranted.

The Risk of LitigationIn my view, the prudent physician who gives ECT can bestprotect himself against even the small risk of a malpracticesuit, and position himself to prevail in the unlikely eventthat one is filed, by adhering to the following few simpletenets of good medical practice.

1. Always obtain explicit, fully informed, written consentfor ECT. Give your patient an ECT informationpamphlet to read and note this in the chart. If moreinformation is requested, have the patient view avideotape on ECT and note this in the chart as well.Use the sample consent form in the Appendix, or anyother that includes a reasonably complete presentationof the risks and potential benefits of ECT, includingmention of bilateral compared with unilateral ECT. Allof the main risks of ECT â!”including the possibility ofpersistent memory dysfunctionâ!”should be written intothe form, rather than simply being discussed with thepatient. If your hospital uses only a general consentform that omits specific language covering the detailsof ECT, have the patient sign your own form as well.Have someone reliable physically present during yourexplanation of this material to sign the consent form asa witness. Write everything you tell the patient aboutECT in the chart, along with the patient's queries andyour responses. If during the treatment course the

P.236

patient expresses doubts or fears about receivingfurther ECT, do not proceed with treatment until andunless you have alleviated these apprehensions anddocumented your discussion and the patient'sacquiescence in the chart.

2 . Explain your decision to recommend ECT by recordingeach of its clinical indications in the chart. Do notsimply assert that the patient has â!œmelancholia,â! !but rather list each of its syndromal elements (e.g.,early waking, guilt, anorexia, etc.). Use theauthoritative sources of the 2001 APA ECT Task Forcerecommendations, the Journal of ECT , and this volumeto document all supporting evidence for primary orsecondary use of ECT. The primary indicationsâ!”justifying the use of ECT as the initial treatment ofchoiceâ!”include:

a. The need for a rapid, definitive treatmentresponse

b . The opinion that alternative treatments are riskierthan ECT

c. A history of poor drug response or good ECTresponse

d . Patient preference

The secondary indicationsâ!”justifying use of ECT afteranother treatmentâ!”include:

a. Treatment failure (e.g., drug therapy resistance)

b . Adverse effects of other treatments

c. Deterioration of patient while receiving othertreatments

3. If the indication for ECT is atypical or unclear, obtain asecond opinion from a consultant who is not associatedwith your practice.

4. Attend at least one educational course on ECT eachyear, own the 2001 APA ECT Task Force Report, readat least one up-to -date text that includes a fulldescription of the principles and practice of ECT, and

P.237

subscribe to and read the American Journal of

Psychiatry and the Journal of ECT .

5 . Document carefully the details of each ECT treatmentsession, including drugs and dosages given, stimulusparameters, treatment electrode placement,oxygenation, and seizure monitoring method andduration.

6. Avoid giving ECT to patients with pronounced histrionicpersonality traits, dysthymia, or hypochondriasis, andespecially those who frequently express anger andresentment over alleged misdiagnosis or mistreatmentby previous physicians, no matter how highly theypraise your skills in comparison. Such patients rarelyrespond favorably to ECT.

7. Read the nursing notes and document each instance ofclinical improvement as it occurs. Regularly examineand record the patient's cognitive status. Mostimportantly, write legibly while imagining each notebeing read aloud to the jury by a plaintiff's attorney.

Notes1Lightner versus Alessi, 22 February 1995

2The Estates of Lam v. Upjohn & Co., 20 March 1995

Appendix

Consent for ElectroconvulsiveTherapy

Information:ECT, previously known as shock therapy, is a method fortreating certain mental or emotional conditions bystimulating the brain electrically in order to produce acerebral seizure. The procedure is carried out by doctors andnurses while the patient is fully asleep under general anesthesia.

Description of the Procedure:While the patient is lying on a stretcher, a needle is placedin a vein and an anesthetic medication is injected. After thepatient is asleep, a muscle-relaxing medication is then giventhrough the same needle, and the patient is given pureoxygen through a mask. When the patient's muscles arerelaxed, an electrical stimulus is briefly applied to the scalpin order to stimulate the brain into a period of intense,rhythmical, electrical activity. This seizure lasts 1 or 2minutes and is accompanied by mild contractions of themuscles. When the seizure is over, the patient is taken to arecovery area and is observed by trained staff until heawakens, usually in about 20 minutes. ECT is usually givenevery other day for about 6 to 12 treatments, althoughsome patients may require more than 12 treatments toreach maximum improvement.

Risks of the Treatment:ECT is among the safest of medical treatments given undergeneral anesthesia. The risk of death or serious injury withECT is about 1 in 50,000 treatments, much smaller than thatreported for child birth. The extremely rare deaths that dooccur are usually due to cardiovas cular complications.

Side Effects and Complications:Patients may be confused just after they awaken from ECT;this confusion generally clears up within an hour or so.Memory for recent events may be disturbed, and dates,names of friends, public events, addresses, and telephonenumbers may be forgotten. In most patients, this memorydifficulty goes away within a few days or weeks, although avery few may continue to experience memory problems formonths or years afterwards. Certain treatment techniquesprevent or minimize the occurrence of such memoryproblems (for example, brief-pulse, right unilateral ECT),and your doctor will discuss these options with you. No long-term effects of ECT on intellectual ability (IQ) or memorycapacity have been found.

Results of Treatment:

P.238

Although many patients experience significant improvementafter a course of ECT, no specific treatment results can bepromised. As is true with all medical treatments, somepatients will recover quickly, some slowly, and a few mightnot recover at all. Even when recovery is complete, relapseis still possible. Medication therapy or maintenance ECT isoften prescribed after a successful course of ECT in order toprevent such relapses.

Availability of AlternativeTreatments:Medications and other therapies may be available to treatyour particular condition, and it is possible that some ofthem might work as well as, or better than, ECT. The advantages and disadvantages of alternate therapies will bediscussed with you by your doctor.

Right to Withdraw Consent:Even though a patient voluntarily signs an agreement toreceive ECT, he may withdraw his consent at any time, evenbefore the first treatment is given. Withdrawal of consentfor ECT does not in any way prejudice the patient'scontinued treatment with the best alternative methodsavailable.

Risks of Not Having ElectroconvulsiveTherapy as Recommended:It is possible that ECT may be more effective for yourcondition than any other available treatments, and that ifyou choose not to accept your doctor's recommendation tohave ECT, you might experience a longer or more severeperiod of illness and disability. Medication and othertherapies have their own risks and complications and maynot be safer than ECT.

I, ________________, have read the above description ofthe ECT treatment that has been recommended to me, andit has also been explained to me by, who has answered anyquestions I had. I agree to have the treatments andunderstand that Dr. ______________ will be in charge ofadministering the treatments.

Patient signature _________________________ Date__________________

Witness signature _________________________ Date__________________




> Table of Contents > Chapter 13 - Transcranial Magnetic

Stimulation Therapy (TMS)

Chapter 13

Transcranial MagneticStimulation Therapy (TMS)

There is no such thing as magnetic stimulation therapy. Themagnetic fields applied to the scalp in both nonconvulsiveand convulsive magnetic methods induce electrical current toflow in the brain, and it is that current that has therapeuticproperties; viewed in this way, there is no inherentdifference between electrical and magnetic brain stimulationmethods in psychiatry. Indeed, attempts to induce seizuresin monkeys via magnetic fields were only successful whenthe intracerebral peak voltage levels induced matched thoserecorded during the conventional application of electriccurrent for ECS (Lisanby et al., 2001a).

There are important practical differences between themethods, however, the most striking of which is that theskull, which deflects at least 90% of applied electricalcurrent, is entirely transparent to magnetic fields. Thus, anyâ!œblurringâ! ! effects that might be attributable to skullimpedance during electrical stimulation simply do not occurwith magnetic fields, allowing for far more accurate focusingon the desired site of cortical stimulation with magnetic, ascompared with electrical, methods. It is irrelevant that wemay have little idea at this writing where to focus theinduced currents, and no ability to induce them any deeperthan about 2 cm from the surface of the coil (that is, in thecortex only)â!”more rational and more efficient stimulationmethods will doubtless soon materialize, and one need onlyconsider the great advances in cancer therapy consequent tothe ability to precisely focus radiation to appreciate theenormous potential advantage of focused versus unfocusedelectrical stimulation.

Also irrelevant is the fact that nonconvulsive electricalstimulation methods proved mostly ineffective when studiedin the 1940s and 1950s. This is because the high impedanceof the skull made it difficultâ!”if not impossibleâ!”to delivercurrent sufficient for brain stimulation without also causinga seizure. Moreover, the conclusion from those early studiesthat nonconvulsive electrical stimulation of the brain lacksantidepressant properties has been confuted by thenumerous sham-controlled nonconvulsive magnetic studies inthe modern era (see below) demonstrating a clearlysignificant, albeit limited, antidepressant effect.

There is also little point in establishing an artificial conflictbetween magnetic induction and ECT, as between John Henryand the steam hammer.

P.240

P.241

First of all, the indications and treatment results at thispoint seem rather different: ECT for the more severe formsof major depression (variously designated psychotic,endogenous, suicidal, delusional, or melancholic), andmagnetic induction for less severe forms of major depression(and perhaps also including atypical, reactive, neurotic,anxious, or dysthymic depression). This may well change,but what difference could it possibly makeâ!”except toadvance the care of patientsâ!”whether one or anothermethod turns out to be more or less effective or efficient intreating a particular disorder?

Stimulus Generation and ParametersIn the broadest sense, TMS resembles giving ECT to a metalcoil. A powerful electrical impulse is delivered to a coilplaced parallel to the head, generating a magnetic field that,in turn, induces electrical current to flow in the brain. (Ofcourse, the stimulus delivered to the coil must be manythousands of volts greater than that used for ECT, in orderto generate the required magnetic field strength, which is inthe same range as for MRI.) The electrical energy for thispurpose is stored in one or more condensers until the timeof stimulation, when it is discharged in a fraction of asecond directly into the coil, generating a magnetic field of1-2 Tesla (the magnetic equivalent of joules). The electricalimpulse thereby induced in the brain typically lasts 0.1-0.3ms (similar in duration to an ultra-brief stimulus for ECT),but more closely resembles a spike than a square-wave.

Stimulus frequencies used for nonconvulsive stimulation inpsychiatry typically range from 1 Hz to 20 Hz; those used toinduce convulsions are in the neighborhood of 60 Hz.(Neurological testing typically employs stimulus frequencies<1 Hz.) For stimulus frequencies below 1 Hz, the term TMSis used; for 1 Hz and above, the word â!œrepetitiveâ! ! isadded (repetitive TMS, rTMS); frequencies > 1 Hz are also

designated â!œfast.â! ! Although most published studies haveused fast rTMS (in the unsupported belief of a direct

relationship between stimulus frequency and the intensity ofcurrent flow in the brain), expert interest has recentlyshifted towards the potential advantages of 1 Hzstimulation, for 2 reasons: (1) 1 Hz stimulation has provenas effective in depression as faster frequencies; and (2) 1Hz stimulation has neurophysiological effects in animalsreminiscent of long-term depression, and may thereforeexhibit potentially useful â!œquenching,â! ! or antikindling,properties (Weiss et al., 1997; Post et al., 1999; George,Lisanby, and Sackeim, 1999; Lisanby and Belmaker, 2000).

A 1 Hz repetition rate is slow enough to allow for rechargingof the condenser between stimuli, and for sufficient passiveheat dissipation to prevent excessive heating of the coil.Faster repetition rates often require multiple condensers (â!œboostersâ! !) that are alternately charged and dis charged,and water-cooled coils (this is especially true for coppercoils; ironcore coils are less subject to heat build-up).

Figure-8 coils produce more focal stimulation; circular(doughnut -shaped) coils produce more diffuse stimulation.

P.242

Although virtually all published studies of rTMS have used

figure-8 coils (based on an unsupported belief in thedesirability of focused stimulation), expert interest hasrecently shifted in the direction of the more diffusestimulation obtained with circular coils (Lisanby andBelmaker, 2000).

Anatomical Site of StimulationVirtually all rTMS studies have applied the stimulus over the

dorsolateral prefrontal (DLPF) area, about 5 cm anterior tothe site of the maximal evoked motor response, and almostalways on the left side of the head. Some apparently cogentrationales have been offered in support of this choice, butthe reality is that most investigators simply followed whatothers did before them. At present, the party line is thatslow right-sided rTMS and fast left -sided rTMS should have

equivalent antidepressant effects, based on theories ofhemispheric specialization of emotional expression anddifferential excitatory and inhibitory neurophysiologicaleffects of fast and slow rTMS, largely based on post-hoc

reasoning. Unfortunately for the theories, however, both leftand right 1 Hz rTMS have yielded significant antidepressant

effects (Epstein et al., 1999; Klein et al., 1999), while someof the worst results have been obtained with left -sided fast

rTMS (George, Wasserman, and Kim brell, 1997; Kimbrell,

Little, and Dunn, 1999; Loo et al., 1999).

Selection of DosageTo date, all studies of rTMS have determined dosage by first

stimulating over the dorsolateral scalp to determine thesmallest dose to produce a clearly visible motor response inthe contralateral hand, and then moving the coil anteriorly 5cm and administering treatment stimulation at 80%-110% ofthe motor threshold obtained. This procedure is analogousto selecting ECT dosage via stimulus titration (see Chapter6), and makes about as much sense. There is no knownrelation between the motor threshhold and the stimulus doserequired to produce a therapeutic effect. Indeed, althoughthe distance from the stimulus coil to the motor cortex is animportant determinant of the motor threshold, there is nocorrelation between the motor threshold and the distancefrom the coil to the prefrontal cortex (McConnell et al.,2001).

Clinical Indications and EfficacyAt the time of this writing, only nonconvulsive rTMS is

available for clinical use, and only outside of the UnitedStatesâ!”it has not been yet been

approved by FDA. Convulsive, TMS has been reported in 1pilot study (reviewed below), but there is as yet nopublished literature on the subject. Therefore this chapterconcentrates on a review of the nonconvulsive rTMS

literature.

Now that a sufficient number of systematically conductedtrials of rTMS has accumulated, including 9 random-

P.243

assignment, double-blind, sham-r TMS controlled studies,

there is no reason to review the earlier, exploratory studiesconducted prior to 1995; these have been covered in detailin several earlier reviews (e.g., Markwort, Cordes, andAldenhoff, 1997; George, Lisanby, and Sackeim, 1999).

Major depression is the only diagnostic group for whichunequivocal therapeutic effects of rTMS have been reported

when compared with sham, TMS. The present review andassessment of the efficacy of rTMS will there fore

concentrate on results obtained in major depression;applications in other diagnostic groups will only be referredto incidentally.

Systematically Conducted Open TrialsTriggs et al. (1999) treated 10 drug-free major depressiveswith an open trial of 2 weeks of 20 Hz rTMS administered

daily on weekdays at 80% of motor threshold via a figure-8coil positioned over the left frontal area. Each sessionconsisted of 50 2-second trains of 40 stimuli each, for atotal of 2000 stimuli per session. Hamilton depression scoreswere reduced 41% at the end of the 2-week course, and50% of the patients were classed as treatment respondersusing a cutoff of at least 50% reduction in depressionscores.

Grunhaus et al. (2000) conducted an open-assessment,random-assignment comparison of rTMS and brief-pulse ECT

in 40 major depressives (20 per group), all of whomreceived unlimited concomitant psychotropic drug therapy.10 Hz rTMS was admininstered to the left dorsolateral

prefrontal area via a figure-8 coil at 90% of motor thresholdfor 20 trains of either 2 seconds each (first 8 patients) or 6seconds each (final 12 patients), for a total of either 400 or1200 stimuli/day, 5 days/week on weekdays for a total of 4weeks (20 days of stimulation). Brief-pulse ECT wasadministered at 2.5Ã— threshold via right unilateralplacement for the first 6 treatments, after which a switch tobitemporal ECT could occur if > 30% improvement inHamilton depression score had not been obtained (this occurred in 8 patients). ECT patients received an average of9.6 treatments (range, 7-14) during somewhat less than 4weeks.

At the end of the treatment course, ECT was significantlymore effective than rTMS in reducing Hamilton depression

scores (61% versus 40%, respectively) and 80% of ECT-treated patients were designated responders by the criterionof at least 50% reduction in depression score, comparedwith 45% who received rTMS. This ECT advantage was even

more striking for the 19 patients who were psychoticallydepressed: 100% responded to ECT, compared to only 22%to, TMS. Presumably, an even larger overall ECT

advantage would have been obtained with the 6Ã— thresholddosing for uni lateral ECT that has now replaced the onlymoderately effective 2.5Ã— thresh old dosing used in thestudy (Sackeim et al., 2000).

The article's title: â!œRepetitive Transcranial Magnetic

P.244

Stimulation is as Effective as Electroconvulsive Therapy inthe Treatment of Nondelusional Major Depressive Disorderâ! !reflects the results obtained when the authors eliminatedfrom the analysis all patients (about half the sample) thatexhibited the delusional feature that contributed heavily tothe ECT advantage; when that was accomplished, the 2methods were indeed equally effective.

Studies Including a Sham-r TMS

Control GroupPascual-Leone et al. (1996) reported the first random-assigment, double-blind, sham-controlled study of rTMS in

unipolar major depression, in 17 medication-resistantpsychotic depressives who were treated as outpatientsduring a 5-month multiple cross-over trial. During the 5-month trial, each patient underwent rTMS only for the first 5

days of each month, randomly assigned to receive active,TMS at 3 different scalp positions (vertex, and left or rightdorsal prefrontal area), and sham rTMS at only the left or

dorsal prefrontal positions. Thus, during the 5-month trial,each patient received a total of 15 days of genuine rTMS.

Active rTMS was administered via a figure-8 shaped coil at

90% of motor threshold, in 5 daily sessions, each sessionconsisting of 20 trains of 10-second duration at 10 Hzfrequency, with a 1-minute pause between trains, for a totalof 2000 stimuli per day.

The only objective outcome measure obtained was the scoreon the Hamilton rating scale for depression, which wasreduced, on average, 45% below baseline after 5 months oftreatment. A similar result was obtained with the Beck self-rating scale for depression. The authors inexplicably focus,however, on the unsupported claim that, after the study wascompleted, 11/17 patients (65%) retrospectively providedtheir â!œsubjective recollectionâ! ! of having experienced â!œpronounced improvementâ! ! only after genuine left orright dorsal prefrontal rTMS. Nowhere in the article do the

authors de scribe how or under what circumstances thisinformation was obtained from the patient.

All patients received the calcium-channel blocker nimodipinethroughout the study, for an intended â!œmood-stabilizingâ! ! effect, and more than half continued to receiveantidepressants and antipsychotics as well. Difficult as it isto imagine how the authors were able to collect a sample of17 psychotic depressives without turning up a single bipolardepressive, it simply beggars the imagination to conceive ofmanaging such severely-ill patients as out patients,especially when they were to receive only 15 days of active,TMS out of more than 150 study days.

Even accepting the study results at face value, a 45%improvement in depression over 5 months is hardlyimpressive, especially since more than

half were also receiving antidepressants. (Compare theseresults with the 21% and 50% reductions in Hamiltondepression scale scores obtained in medication-free patientsby George, Wasserman, and Kimbrell [1997] and Klein et al.

P.245

[1999] after only 2 weeks of sham rTMS, as described

below.)

George, Wasserman, and Kimbrell (1997) randomly assigned12 major depressives (all but one were unipolar) to receiveeither 2 weeks of genuine r TMS or 2 weeks of sham rTMS,

after which treatment group assignment was crossed overand 2 additional weeks of treatment given. Active treatmentconsisted of 10 daily weekday sessions of 20 minutes of leftdorsolateral prefrontal stimulation administered via a figure-8 coil at 80% of motor threshold and 20 Hz frequency, to atotal of 800 stimuli per session. During genuine rTMS

Hamilton depression scores fell about 8%, compared with anincrease of about 1% with sham rTMS, a difference that was

statistically significant, but clinically trivial (a point thatapparently eluded the authors of the article).

Klein et al. (1999) treated 57 major depressives who wererandomly assigned to receive genuine or sham rTMS for a

total of 10 daily sessions during a 2-week double-blind trial.Active, TMS was administered at 110% of motor thresholdvia a circular coil placed over the right prefrontal area. Eachof the daily sessions consisted of 2 trains of 60 stimuli each,spaced 3 minutes apart, with stimulus parameters of 0.1 mspulse duration and 1 Hz frequency, for a total of 120 stimuliper day. Blindly rated Hamilton depression rating scalescores were obtained at baseline, after 5 treatment sessions,and after the final rTMS.

Active rTMS reduced Hamilton depression scores 36% after 1

week, and 50% after 2 weeks, significantly better than thesham-r TMS reductions of 18% and 25%, respectively. When

patients were classified as responders or nonresponders bythe criterion of at least 50% reduction in one or the otherdepression scale score, essentially the same figures resulted.As in Pascual-Leone et al. (1996), most patients receivedconcomitant antidepres sant drug therapy; however, nosignificant treatment (active versus sham) Ã— druginteraction was detected.

Loo et al. (1999) randomly assigned 18 medication-resistantmajor depressives to 2 weeks of treatment with either sham

rTMS, or active rTMS administered daily on weekdays via a

figure-8 coil placed over the left dorsolateral prefrontalarea. 10 Hz stimulation was given at 100% of motorthreshold for a total of 1500 stimuli per day. After 2 weeksof treatment there was no difference between the sham andgenuine rTMS groups for the improvement in Hamilton

depression scale scores. More than 70% of pa tientscontinued to receive antidepressants during the study.

The within-subjects crossover study of Kimbrell, Little, andDunn (1999) in 13 major depressives is difficult tosummarize because of limited correspondence between whatis claimed in the text and what is shown in the table,nonsystematic randomization, nonsystematic assignment tosham r TMS (in fact, only 3/13 patients received the sham

condition), and a lack of forthrightness in discussing therather limited clinical results obtained.

Nevertheless, it appears that most patients received either 2weeks of daily left prefrontal 1 Hz stimulation followed byanother 2 weeks of stimulation at 20 Hz, or the sametreatments in the reverse order, often by randomassignment. A figure-8 coil was used to deliver stimulationat 80% of motor threshold at the rate of 800 stimuli perday. Blindly obtained Hamilton depression scores showed aslight worsening after 2 weeks of 20 Hz r TMS, compared

with a modest (20%) improvement after 2 weeks of 1 Hzstimulation, a difference that was statistically significant.(Indeed, there was a substantial negative correlation withinsubjects between the effects on de pression of the 2different stimulus frequencies.)

Because the 3 patients who received sham rTMS received

only 20 Hz stimulation as an active treatment, no shamcontrol is available for the 1 Hz effects on depression.However, since the antidepressant effects of 20 Hzstimulation in those patients who also received sham rTMS,

as well as in those who did not, were no different from thoseof sham stimulation, it is reasonable to infer that the studyalso showed left frontal 1 Hz rTMS to be modestly better

than sham rTMS.

Padberg et al. (1999) conducted a double-blind, random-assignment comparison in 18 drug-resistant majordepressives among fast (10 Hz), slow (0.3 Hz), and sham

rTMS applied at 90% of motor threshold over the left

dorsolateral prefrontal area, at a rate of 250 stimuli per dayfor 5 successive days. After 5 days of treatment, 0.3 Hz,

rTMS reduced Hamilton depression scores by 19%, compared

with 6% for 10 Hz rTMS and -1% with sham r TMS, a

marginal, but statistically significant difference.

George et al. (2000) randomly assigned 30 medication-freemajor depressives to receive daily genuine (Â« = 20) orsham (n = 10) rTMS at 100% of motor threshold, further

dividing the genuine, TMS group equally to receive 5 Hz or20 Hz stimulation. Stimulation was administered via afigure-8 coil positioned over the left prefrontal area for 20min/day over a 2-week period, for a total of 1600 stimuliper day. Psychotic and actively suicidal patients wereexcluded from study.

Blindly obtained Hamilton depression scale scores fell 35%with genuine rTMS by the end of the 2-week trial period,

versus 21% with sham, TMS, a difference that was notsignificant. When patients were dichotomized into respondersand nonresponders by the criterion of > 50% improvementin depression score, significantly more responders werefound in the genuine rTMS group (9/20) than the sham rTMS

group (0/10). There were no significant differences inantidepressant efficacy between 5 HZ and 20 Hz stimulation.

Berman et al. (2000) randomly assigned 20 medication-free,treatment -resistant major depressives to receive eithersham, TMS or genuine 20 Hz r TMS administered daily on

weekdays for 2 weeks via a figure-8 coil over the leftdorsolateral prefrontal area at 80% of the motor threshold,for a total of 800 stimuli per day. Blindly obtained Hamilton

P.246

depression scores fell about 40% with active treatment,compared with no improvement after sham r TMS. Only 1

patient receiving active treatment exhibited robust (i.e.,85%)

improvement; the 3 next best improvements with genuine

rTMS were in the 40%-45% range.

Tormos et al. (in preparation) randomly assigned 45 drug-free, unipolar, nonpsychotic, recurrent major depressives toreceive either sham rTMS, or one of 3 forms of genuine

rTMS: left dorsolateral prefrontal stimulation at 1 Hz, left

dorsolateral prefrontal stimulation at 10 Hz, or rightdorsolateral prefrontal stimulation at 1 Hz. Stimulation wasgiven daily for 10 consec utive weekdays over 2 weeks, viaa figure-8 coil at 110% of motor threshold, for a total of1600 stimuli per day, delivered over 20-30 minutes.

Blindly obtained Hamilton depression scales scores fell 46%and 43% after the 2-week trial for the genuine 10 Hz left -sided and 1 Hz right-sided groups, respectively. Neithersham rTMS nor genuine 1 Hz left -sided rTMS exhibited any

treatment effects. Dichotomizing treatment response using a> 50% cut -off in Hamilton depression scale scores revealedsimilar results: the left -fast and right-slow groups bothyielded 50% response rates, and neither of the other 2groups provided any treatment responders. At a 4 weekpost-r TMS follow-up assessment, the 2 effective methods

continued to yield virtually identical levels of improvement.

The results of the above studies are summarized in Table13-1, arranged in order of decreasing percent improvementobtained in Hamilton depression score. The 9 sham-controlled studies that included more than one type of activestimulation are tabulated as if each active methodconstituted a separate comparison with sham rTMS, yielding

a total of 14 active versus sham comparisons.

Four points are immediately apparent from inspection of theTable 13-1:

1. Sham rTMS is clearly inferior to the genuine article: all

but one of the sham rTMS-controlled studies showed a

significant advantage for active treatment. Therefore,

rTMS has significant antidepressant properties.

2. No dramatic or even very substantial antidepressanteffects are reported with rTMS for the stimulation

methods shown; its efficacy as presently administeredis moderate at best: 50%, either by Hamiltondepression score reduction, or response rate accordingto a criterion of at least 50% Hamilton score reduction.

3. Both the percent improvement and treatment responserates obtained with r TMS are inferior to those reported

for bitemporal ECT and high-dose unilateral ECT (seeChapter 7, Table 7-1), and approximate the resultsoften obtained with antidepressant drug therapy.

4. There is no obvious or striking association between theantidepressant efficacy of rTMS and coil shape,

P.247

stimulation level, stimulus frequency, stimulation site,number of stimuli administered per session, or the presence of a sham-r TMS control group.

That being said, there is a tendency for both 1 Hz, andright-sided, stimulation, independently to be more frequentlyassociated with better treatment

outcomes. At the very least, it is easy to see that there isno specific advantage for either left -sided stimulation orfrequencies >1 Hz.

Table 13-1 Systematically conducted studies of theefficacy of rTMS n i major depression

Study Coil Thresh Hz DLPF Stimuli %Imp %Resp %Sham

Klein et al(1999)

round 110 1 R 120 50 50 25

Tormos etal. (inpreparation)

F-8 110 10 L 1600 46 50 0

Pascual-Leone et al.(1996)

F-8 90 10 L/R 2000 45 8


F-8 110 1 R 1600 43 50 0

Triggs et al.(1999)

F-8 80 20 L 2000 41 50 NA

Herman etal. (2000)

F-8 80 20 L 800 40 0

Grunhaus(2000)

F-8 90 10 L 400-1200

40 45 NA

George etal. (2000)

F-8 100 5 L 1600 35 45 21

George etal. (2000)

F-8 100 20 L 1600 35 45 21

Loo et al.(1999)

F-8 100 10 L 1500 25 23

P.248

Kimbrell,Little, andDunn(1999)

F-8 80 1 L 800 20 0

Padberg etal. (1999)

F-8 90 0.3 L 25 19 -1

George,Wasserman,andKimbrell(1997)

F-8 80 20 L 800 8 -1

Padberg etal. (1999)

F-8 90 19 L 250 6 -1

Kimbrell,Little, andDunn(1999)

F-8 80 20 L 800 0 0


F-8 110 1 L 1600 0 0 0

Coil, coil type (round or flgure-8 [F-8]); Thresh, % of motor threshold at whichstimulus delivered; Hz, stimulus frequency; DLPF, the side (L or R) of application ofdorsolateral prefrontal stimulation; Stimuli, number of stimuli per treatment session;%Imp, percent reduction in baseline Hamilton depression score; %Resp, percentlabeled â!œrespondersâ! !; %Sham, percent improvement with sham, TMS; NA, notapplicable.

An examination of only the 9 sham-controlled studies forthe differences between active and sham rTMS in Hamilton

depression score reduction, treating each within-studycomparison of more than one type of active treatment as aseparate study (yielding a total of 14 active versus shamcomparisons), shows the mean improvement across studiesfor active rTMS to be almost 4 times as large as for sham

rTMS (29% compared with 8%, respectively), a figure that

compares favorably with that for placebo-controlledantidepressant drug trials, which typically yield only a 2:1ad vantage for active treatment.

The only absolutely negative sham-r TMS-controlled study is

that of Loo et al. (1999), who reported a 25% Hamiltonscore reduction with 2 weeks of active rTMS compared with a

23% reduction with sham. In sham, TMS the stimulus coil isplaced at a 45-90 degree angle to the head in order not toinduce electrical potentials in the brain, while at the sametime stimulating the scalp to maintain the â!œblind.â! ! Looet al. (1999) placed the edge of

P.249

a 2-wing-figure-8 coil at a 45 degree angle to the head. Twoarticles have shown that this form of sham rTMS is actually

active, inducing an intracerebral electrical field estimated tobe from less than 50% (Loo et al., 2000) to as much as76% (Lisanby et al., 200Ib) of that induced by intendedstimulation. Thus, the unintended use of an active shamcondition by Loo et al. (1999), even at an intensitysubstantially less than 50% of the intended active condition,raises the possibility that theirs was not a negative studyafter all. (From a practical point of view, however, even ifzero effect had been recorded for the sham condition, 25%efficacy in major depression is no better than the resultsobtained with placebo in countless antidepressant drugtrials.)

Another study with a substantial (21%) response to sham

rTMS, resulting in only a moderate therapeutic effect of

active relative to sham rTMS, also used a 2-wing-figure-8

coil held at a 45 degree angle to the head (George et al.,2000), suggesting that their results might also have beensubstantially better using a truly inactive sham condition.

Clinical Trials of rTMS in Other

DisordersThere have been a handful of clinical trials of rTMS in other

diagnostic groups, of which the only controlled study wasconducted in 16 patients with mania (Grisaru et al., 1998).The control in this instance was the assignment (randomly inabout one third of subjects) of patients to receive r TMS to

either the left or right hemisphere; no sham-r TMS condition

was included. Assessment was blind in about two thirds ofsubjects, and almost all received concomitant uncontrolledpsychotropic drug therapy. Stimulation at 20 Hz wasadministered daily on weekdays for 2 weeks at 80% ofmotor threshold, 20 2-second trains per day (800 stimuli perday).

At 7 and 14 days, mania scale scores fell 15% and 20%,respectively, in the left -r TMS group, compared with

reductions of 35% and 71% for the right-r TMS group,

differences that were significant at both assessmentintervals. An overall effect of rTMS in reducing baseline

mania scores did not survive post-hoc testing. Consideringthe lack of a sham-r TMS control, the virtually unlimited co-

administration of psychotropic drugs, and the nonrandom,nonblind, components of this study, it provides little usefulinformation concerning either the overall efficacy of rTMS in

mania, or the relative efficacy of left versus righthemisphere stimulation.

Safety and Side-Effects of rTMS

SafetyConsidering that during MRI, patients are routinely andsafely exposed to continuous (static) magnetic fields of >1.5 Tesla for up to an hour, the tiny

P.250

fraction (e.g., 0.1%) of that stimulation level that isdelivered during time-varying rTMS is certainly irrelevant to

health considerations.

CognitionSeveral studies have examined a variety of memory andother cognitive functions with simple cognitive screeningtasks as well as more complex neuropsychological testsbefore and at varying intervals after a course of r TMS (e.g.,

Avery, Claypoole, and Robinson, 1999; Triggs et al., 1999;Padberg et al., 1999; George et al., 2000; Grunhaus et al.,2000; Loo et al., 1999). No rTMS-induced cognitive

dysfunction has been detected; indeed, just as for ECT,several studies report a trend towards improved cognitivefunctioning after rTMS, the most comprehensive of which

revealed no cor relation between cognitive improvement andrelief of depression (Loo et al., 1999).

HearingEach discharge of electrical energy into the coil isaccompanied by a loud clicking noise that can causetemporary hearing loss or transiently increased auditorythreshold in man (Loo et al., 1999). For this reason,patients and staff should wear protective earplugs during theadministration of both active and sham rTMS.

Scalp and Facial PainPain or discomfort in the scalp and facial area under oradjacent to the site of stimulation is due to repeated musclecontraction and is occasionally severe enough to requirereducing stimulus intensity or to cause patients to withdrawfrom treatment (Epstein et al., 1999; Klein et al., 1999).

SeizuresSlow TMS (e.g., 0.3 Hz) as used for neurological assessmentis long known to have the potential for inducing seizures inpatients with preexisting neurological lesions (e.g., stroke,multiple sclerosis) or epilepsy), but only rTMS has induced

seizures in normals, the first reported instance of whichoccurred after 10 seconds of 10 Hz stimulation at 100% ofthe motor threshold (Pascual-Leone et al., 1992).

To date, epileptic seizures (grand mal or complex partial)have been accidentally induced in a total of at least 8patients undergoing rTMS. Because most were normal

volunteers, the causal link is most likely to the

nature of the stimulus rather than patient susceptibility.Many of these seizures have occurred at higher stimulusfrequencies, but Chen et al. (1997) reported that 15 minutesof 0.9 Hz stimulation induced â!œspread of excitation, whichmay be a warning sign for seizures â!" indicating the needfor adequate monitoring even with stimulations at lowfrequencies.â! ! Although EEG monitoring during theadministration of slow, or even fast, rTMS is not yet routine,

this report provides a clear rationale for it to become so.

P.251

The risk of unintended seizures at higher stimulusfrequencies has led to comprehensive limits on stimulustrain duration according to stimulus intensity and frequencyemployed (Wasserman, 1998). For example, the safe limitfor 10 Hz stimulation at various proportions of the motorthreshold is set at: 5 seconds of stimulation at 100%; 4.2seconds at 120%, and 2.9 sec at 130%. Similar figures areprovided for a complete range of stimulus frequencies, withthe further suggestion provided to add a 25% safety marginto reduce the risk of seizures. (The relative safety of low-frequency stimulation is underscored by the willingness ofthe US Food and Drug Administration to allow clinicalresearch to be conducted with devices up to 1 Hz maximumfrequency without requiring an investigational deviceexemption.)

Neurometabolic EffectsAs expected for stimulation at or near an intensity capableof inducing a motor response, the immediate cortical effectsof rapid rTMS are excitatory. Fox et al. (1997) used FDG-

PET to identify a focal blood flow increase of 12%-20% overthe left primary motor cortex during rTMS administered over

that site, and Paus et al. (1997) used 15O water-PET todemonstrate increased rCBF as a function of the number of

subthreshold 10 Hz stimulus trains over the left frontal eyefield. Siebner et al. (2000, 2001) used FDGPET to show that1800 stimuli of 2 Hz at 140% of motor threshold increasedglucose uptake in both directly stimulated and anatomico-functionally related areas; a similar effect occured withsubthreshold stimulationâ!”1800 stimuli of 5 Hzadministered below the motor threshold (Siebner et al.,2000 , 2001). A significant increase in oxyhemoglobin anddecrease in cytochrome oxidase as measured by near-infrared spectroscopy immediately following r TMS likewise

provides evidence for metabolic activation and increasedcerebral blood flow consistent with these findings (Olivieroet al., 1999).

A series of functional MRI, (MRI) studies obtained continuousblood oxygenation level-dependent images during interleaved

rTMS to demonstrate that:

1. The degree of rTMS-induced cortical activation varies

with the intensity of the stimulus (Bohning et al.,1999).

2. The 2%-3% increase in cortical activation associatedwith repetitive 1 Hz stimulation at 110% of motorthreshold is the same as that for an equivalentvoluntary motor act (Bohning et al., 2000a).

3. Even the l%-2% increased cortical activation withsingle rTMS pulses is detectable, both in ipsilateral and

contralateral cortices (Bohning et al., 2000b).

The only contradictory report concerning the excitatory

short-term effects of rTMS is the 15O water-PET replication

P.252

attempt of Paus et al. (1998), who unexpectedly found thatCBF decreased proportionally to the number of subthresholdtrains administered to the left primary sensorimotor cortex.

Longer-term effects of rTMS may be frequency-dependent.

Speer et al. (2000) examined patients at baseline and 72hours after a series of 10 daily sessions (1600 stimuli persession) with either 1 Hz or 20 Hz stimulation in a within-subjects, unbalanced crossover design, and found a tiny(1%) but significant increase over baseline in global CBF

(by15O water PET) for the fast condition only; an evensmaller (0.04%) effect of slow rTMS was in the opposite

direction. The authors' characterization of the 1% increasewith 20 Hz stimulation as â!œmore robustâ! ! than the â!œmore modestâ! ! 0.04% decrement with 1 Hz stimulationsurely overstates the case. Similar effects were obtained forregional CBF changes. None of the CBF changes wasassociated with improvement in depressive symptoms. Theauthors mention, but do not report any results for, a sham

rTMS condition to which patients were randomly assigned:

the same â!œactiveâ! ! sham method as discussed above forLoo et al. (1999) and George et al. (2000).

Neurophysiological EffectsChen et al. (1997) used changes in the motor evokedpotential to study the excitability of the motor cortex innormal humans receiving rTMS. Very slow (0.1 Hz)

stimulation for 1 hour (360 stimuli) had no effect on corticalexcitability, but 0.9 Hz stimulation for 15 minutes (810stimuli) significantly decreased the amplitude of the motorevoked potential by about 20%, an effect that lasted for 15minutes. The authors note that such a reduction in corticalexcitability is consistent with the phenomenon of long-termdepression, an inhibitory process related to the antikindlingphenomenon of quenching, and which has been produced inrodents with low-frequency electrical stimulation (Weiss etal., 1997; Post et al., 1999).

Summary and ConclusionsThe antidepressant properties of nonconvulsive rTMS have

now been amply demonstrated in several carefullyconducted, sham-controlled trials, almost entirely intreatment -resistant patients who have failed standardcourses of therapy with a variety of antidepressant drugs. Todate, the clinical results obtained with rTMS have

approximated those obtained in similar patient samples withantidepressant drugs (see Chapter 2) and with moderate-dose

(e.g., 2.5Ã— threshold) brief-pulse unilateral ECT (Sackeimet al., 1993, 2000 , 200la); therefore, rTMS may turn out to

be a reasonable therapeutic alter native to these 2treatment methods.

As presently configured, however, because nonconvulsive

rTMS almost certainly lacks the therapeutic potency of

bitemporal, bifrontal, and high-dose unilateral ECT (andindeed, has not yet been compared with any of those

methods under controlled, random-assignment, double-blindconditions in medication-free patients), it cannot beadministered to psychotic, delusional, suicidal, ormelancholic depressivesâ!”or other severely-or dangerously-ill patientsâ!”with any reasonable hope of success.

This caveat seems likely to change, however, as incrementalimprovements accumulate in the technique of administrationof nonconvulsive rTMS, just as they have over the years for

ECT. It is quite possibleâ!”even likely â!”that presentmethods of administering nonconvulsive rTMS bear as close a

relationship to the methods that will be employed 25 yearshence as un-monitored sine-wave bitemporal ECT bears tomodern ECT.

Magnetoconvulsive TherapySeparate and different considerations apply to magneto-convulsive therapy, or MCT. Because MCT induces ageneralized seizure via electrical stimulation of the brain, itfar more closely resembles ECT than does nonconvulsive

rTMS. Thus, there is no a priori reason to believe that with

appropriate technical development, MCT cannot achievetherapeutic outcomes identical to those presently obtainedwith the most potent forms of ECT. Indeed, it is reasonableto hope that the ability to focus the brain electricalstimulation of MCT by manipulating coil shape and positionwill confer a distinct advantage upon MCT compared withconventional ECT, ultimately providing response rates notyet achieved with the older method.

However, equipment cost is likely to play a decisive rolewith regard to the speed with which MCT is developed,tested, and, if proven effective, becomes generally available.Although nonconvulsive rTMS devices have comparable

manufacturing costs to ECT devices, and are widelycommercially-available, the far greater energy levelsrequired to deliver high-frequency (40-60 Hz) stimuli at theintensity needed for reliable seizure induction with MCT areextremely expensive to generate: 2 devices working intandem, with a total of 8 â!œboosterâ! ! modules wererequired to administer the first MCT in man, and 8 moremodules were added in a subsequent study (Lisanby et al.,2001c; 2001d). Even so, levels of intra cerebral currentdensity achieved during MCT were substantially below thatdeveloped with ECT.

Considering that years of safety and efficacy studies lieahead for MCT, that ECT devices already provide a safe,effective, and acceptably priced convulsive therapy method,and that medical equipment firms typically lack the kind offinancial research and development support enjoyed by thelarge pharmaceutical companies, it seems unlikely that MCTwill become avail able for general clinical use within thenext 20 years, if ever.




> Back of Book > References

References

Abiuso P, Dunkelman R, Proper M. (1978)Electroconvulsive therapy in patients with pacemakers.JAMA 240:2459-60.

Abramczuk JA, Rose NM. (1979) Pre-anaestheticassessment and the prevention of post-ECT morbidity. Br

J Psychiatry 134:582-7.

Abrams R. (1967) Daily administration of unilateral ECT.Am J Psychiatry 124: 384-6.

Abrams R. (1975) Drugs in combination with ECT. In: MGreenblatt, ed. Drugs in Combination with Other

Therapies . New York: Grune and Stratton, pp. 157-64.

Abrams R. (1982a) Clinical prediction of ECT response indepressed patients. Psy chopharmacol Bull 18:73-5.

Abrams R. (1982b) ECT and tricyclic antidepressants inthe treatment of endogenous depression.Psychopharmacol Bull 75:48-50.

Abrams R. (1986b) A hypothesis to explain divergentfindings among studies comparing the efficacy ofunilateral and bilateral ECT in depression. Convuls Ther

2:253-7.

Abrams R. (1987) ECT in schizophrenia. Convuls Ther

3:169-70.

Abrams R. (1988a) Interview with Lothar Kalinowsky.Convuls Ther 4:243-9.

Abrams R. (1988b) Electroconvulsive Therapy (1st ed.).New York, Oxford Univer sity Press.

Abrams R. (1989a) Malpractice litigation and ECT[letter]. Convuls Ther 5:365-7.

Abrams R. (1989b) ECT for Parkinson's disease[editorial]. Am J Psychiatry 146: 1391-3.

Abrams R. (1990) Termination of ECT-induced prolongedseizures [letter]. Convuls Ther 6:54-8.

Abrams R. (1991a) Seizure generalization and theefficacy of unilateral Electroconvulsive Therapy. Convuls

Ther 7:213-7.

Abrams R. (1991b) The FDA proposal to reclassify ECTdevices [editorial]. Convuls Ther 7:1-4.

Abrams R. (1992) Electroconvulsive Therapy . New York:Oxford University Press.

Abrams R. (1994a) Stimulus parameters and efficacy ofECT [commentary]. Convuls Ther 70:124-8.

Abrams R. (1994b) In honor of Max Fink, MD, foundingeditor [editorial]. Convuls Ther 70:89-90.

Abrams R. (1995) ECT malpractice issues. Convuls Ther

77:286-7.

Abrams R. (1996) ECT stimulus parameters asdeterminants of seizure quality. Psychiatric Annals

P.254

26:701-4.

Abrams R. (1997) Electroconvulsive Therapy (3d ed.).New York, Oxford University Press.

Abrams R. (2000) Electroconvulsive therapy requireshigher dosage levels: Food and Drug Administrationaction is required. Arch Gen Psychiatry 57:445-6.

Abrams R. (in press) Stimulus titration and ECT dosing. JECT. 253

Abrams R, Essman WB, eds. (1982) Electroconvulsive

Therapy: Biological Foun dations and Clinical

Applications . New York: Spectrum Publications.

Abrams R, Fink M. (1972) Clinical experiences withmultiple electroconvulsive treatments. Compr Psychiatry

73:115-22.

Abrams R, Fink M. (1984) The present status ofunilateral ECT: some recommendations [editorial]. JAffect Disord 7:245-7.

Abrams R, Fink M, Dornbush R, Feldstein S, Volavka J,Roubicek J. (1972) Unilateral and bilateral ECT: effectson depression, memory, and the electroen cephalogram.Arch Gen Psychiatry 27:88-91.

Abrams R, Swartz CM. (1985a) ECT and prolactinrelease: relation to treatment response in melancholia.Convuls Ther 7:38-42.

Abrams R, Swartz CM. (1985b) ECT and prolactinrelease: effects of stimulus parameters. Convuls Ther

7:115-9.

Abrams R, Swartz CM. (1985c) ECT Instruction Manual.Lake Bluff, IL: Somatics Inc.

Abrams R, Swartz CM. (1989) Electroconvulsive therapyapparatus and method for automatic monitoring ofpatient seizures [patent]. U.S. patent 4,878,498.

Abrams R, Swartz CM. (1991) Consent forelectroconvulsive therapy. In: R Abrams, CM Swartz,eds. The Technique of ECT: Documentation and Forms

Lake Bluff IL: Somatics Inc. p. 23.

Abrams R, Swartz CM, Vedak C. (1989) Antidepressanteffects of right versus left unilateral ECT and thelateralization theory of ECT action. Am J Psychiatry

746:1190-2.

Abrams R, Swartz CM, Vedak C. (1991) Antidepressanteffects of high-dose right unilateral electroconvulsivetherapy. Arch Gen Psychiatry 48:746-8.

Abrams R, Taylor MA. (1973) Anterior bifrontal ECT: Aclinical trial. Br J Psychi atry 722:587-90.

Abrams R, Taylor MA. (1974a) Unipolar and bipolardepressive illness: phenomenology and response toelectroconvulsive therapy. Arch Gen Psychiatry 30:320-1.

Abrams R, Taylor MA. (1974b) Electroconvulsive therapyand the diencephalon: A preliminary report. Compr

Psychiatry 75:233-6

Abrams R, Taylor MA. (1976a) Catalonia: A prospectivestudy. Arch Gen Psychiatry 33:579-81.

Abrams R, Taylor MA. (1976b) Diencephalic stimulation

P.255

and the effects of ECT in endogenous depression. Br J

Psychiatry 729:482-5.

Abrams R, Taylor MA. (1976c) Mania and schizo-affectivedisorder, manic type: A comparison. Am J Psychiatry

733:1445-7.

Abrams R, Taylor MA. (1979) Differential EEG patterns inaffective disorder and schizophrenia. Arch Gen


Abrams R, Taylor MA. (1981) Importance ofschizophrenic symptoms in the diagnosis of mania. Am J


Abrams R, Taylor MA. (1985) A prospective follow-upstudy of cognitive functions after ECT. Convuls Ther 7:4-9.

Abrams R, Taylor MA, Faber R, Ts'o T, Williams R, AlmyG. (1983) Bilateral versus unilateral ECT: efficacy inmelancholia. Am J Psychiatry 740:463-5.

Abrams R, Taylor MA, Volavka J. (1987) ECT-inducedEEG asymmetry and ther apeutic response inmelancholia: relation to treatment electrode placementAm J Psychiatry 744:327-9.

Abrams R, Vedak C. (1991) Prediction of ECT response inmelancholia. Convuls Ther 7:81-4.

Abrams R, Volavka J. (1982) Electroencephalographiceffects of convulsive therapy. In: R Abrams, WB Essman,eds. Electroconvulsive Therapy: Biological Foundations

and Clinical Applications . New York: SpectrumPublications, pp. 157-67.

Abrams R, Volavka J, Dornbush R, Roubicek J, Fink M.(1970) Lateralized EEG changes after unilateral andbilateral electroconvulsive therapy. Dis Nerv Syst 37:28-33.

Abrams R, Volavka J, Fink M. (1973) EEG seizurepatterns during multiple unilateral and bilateral ECT.Compr Psychiatry 14:25-8.

Abrams R, Volavka J, Schrift M. (1992) Brief pulse ECTin melancholia: EEG and clinical effects. J Nerv Merit Dis

180:55-1.

Accornero F. (1988) An eyewitness account of thediscovery of electroshock. Con vuls Ther 4:40-9.

Acevedo AG, Smith JK. (1988) Adverse reaction to use ofcaffeine in ECT. Am J Psychiatry 745:529-30.

Ackermann RF, Engel J, Jr, Baxter L. (1986) Positronemission tomography and autoradiographic studies ofglucose utilization following electroconvulsive seizures inhumans and rats. Ann N Y Acad Sci 462:263-9.

Addersley DJ, Hamilton M. (1953) Use of succinylcholinein ECT. Br Med J I: 195-7.

Addonizio G, Susman V. (1986) Neuroleptic malignantsyndrome and use of anes thetic agents. Am J Psychiatry

143:121-8.

Adityanjee, Jayaswal SK, Chan TM, Subramaniam M.(1990) Temporary remission of tardive dystonia followingelectroconvulsive therapy. Br J Psychiatry 156: 433-5.

Ahlskog JE, Kelly PJ, van Heerden JA, Stoddard SL, TyceGM, Windebank AJ, Bailey PA, Bell GN, Blexrud MD,

P.256

Charmichael SW. (1990) Adrenal medullarytransplantation into the brain for treatment ofParkinson's disease: clinical outcome and neurochemicalstudies. Mayo Clin Proc 65:3052-8.

Albala AA, Greden JF, Tarika J, Carroll BJ. (1981)Changes in serial dexamethasone suppression testsamong unipolar depressives receiving electroconvulsivetreatment. Biol Psychiatry 76:551-60.

Alexander RC, Salomon M, lonescu-Pioggia M, Cole J.(1988) Convulsive therapy in the treatment of mania:McLean Hospital 1973-1986. Convuls Ther 4: 1152-5.

Alexopoulos GS, Frances RJ. (1980) ECT and cardiacpatients with pacemakers. Am J Psychiatry 137:1111-2.

Alexopoulos GS, Inturrisi CE, Lipman R, Frances R,Haycox J, Dougherty JR, Jr, Rossier J. (1983) Plasmaimmunoreactive (3-endorphin levels in depression: effectof electroconvulsive therapy. Arch Gen Psychiatry

40:181-3.

Alexopoulos GS, Shamoian CJ, Lucas J, Weiser N, BergerH. (1984) Medical problems of geriatric psychiatricpatients and younger controls during electroconvulsivetherapy. J Am Geriatr Sac 52:651-4.

Alexopoulos GS, Young RC, Shamoian CA. (1984)Unilateral electroconvulsive therapy: An open clinicalcomparison of two electrode placements. Biol Psychiatry

79:783-7.

Alger I. (1991) History of modern psychiatry, andmodern application of electro convulsive treatment.Hospital and Community Psychiatry 42:355-6.

Allen RE, Pitts FN, Jr. (1978) ECT for depressed patientswith lupus erythematosus. Am J Psychiatry 755:367-8.

Allman P, Hawton K. (1987) ECT for post-strokedepression: beta blockade to modify rise in bloodpressure. Convuls Ther 3:218-21.

Almansi R, Impastato DJ. (1940) Electrically inducedconvulsions in the treatment of mental diseases. A' YState J Medicine 40:1315-6.

Alpers BJ and Hughes J. (1942) Changes in the brainafter electrically induced convulsions in cats. Arch Neural


American Psychiatric Association. (1978)Electroconvulsive Therapy: Task Force Report No 14.Washington, DC: American Psychiatric Association.

American Psychiatric Association. (1990) The Practice of

Electroconvulsive Ther apy: Recommendations for

Treatment, Training, and Privileging. Washington, DC:American Psychiatric Association.

American Psychiatric Association. (2001) The Practice of

Electroconvulsive Ther apy: Recommendations for

Treatment, Training, and Privileging . (2d. Ed.)Washington DC: American Psychiatric Association.

Ames D, Burrows G, Davies B, Maguire K, Norman T.(1984) A study of the dexamethasone suppression test inhospitalized depressed patients. Br J Psychiatry

744:311-3.

Ananth J, Samra D, Kolivakis T. (1979) Amelioration ofdrug-induced Parkinsonism by ECT. Am J Psychiatry

736:1094.

Andersen K, Balldin J, Gottfries CG, Granerus AK, ModighK, Svennerholm L, Wallin A. (1987) A double-blindevaluation of electroconvulsive therapy in Parkinson'sdisease with â!œon-offâ! ! phenomena. Acta Neural Scand

76:191-9.

Andrade C, Gangadhar BN, Channabasavana SM. (1990)Further characterization of mania as a side effect ofECT. Convuls Ther 6:318-9.

Andrade C, Gangadhar BN, Subbakrishna DK,Channabasavanna SM, Pradhan N. (1988a) A double-blind comparison of sinusoidal wave and brief-pulseelectroconvulsive therapy in endogenous depression.Convuls Ther 4:297-305.

Andrade C, Gangadhar BN, Swaminanth G,Channabasavanna SM. (1988b) Predicting the outcome ofendogenous depression following electroconvulsivetherapy. Convuls Ther 4:169-74.

Andrade C, Gangadhar BN, Swaminanth G,Channabasavanna SM. (1988c) Mania as a side effect ofelectroconvulsive therapy. Convuls Ther 4: 81-3.

Annett M, Hudson PT, Turner A. (1974) Effects of rightand left unilateral ECT on naming and visualdiscrimination analysed in relation to handedness. Br J


Anton AH, Uy DS, Redderson CL. (1977) Autonomicblockade and the cardiovas cular and catecholamineresponse to electroshock. Anesth Analg 56:46-54.

Aperia B. (1986) Hormone pattern and post-treatmentattitudes in patients with major depressive disordergiven electroconvulsive therapy. Acta Psychiatry Scand

73:271-4.

P.257

Aperia B, Thoren M, Wetterberg L. (1985) Prolactin andthyrotropin in serum during electroconvulsive therapy inpatients with major depressive illness. Acta Psychiatry

Scand 72:30-28.

Arana GW, Baldessarini RJ, Ornsteen M. (1985) Thedexamethasone suppression test for diagnosis andprognosis in psychiatry. Commentary and review. Arch

Gen Psychiatry 42:1193-204.

Arato M, Erdos A, Kurcz M, Vermes I, Fekete M. (1980)Studies on the prolactin response induced byelectroconvulsive therapy in schizophrenics. Acta

Psychiatry Scand 67:239-44.

Asnis GM, Fink M, Saferstein S. (1978) ECT inmetropolitan New York hospitals: A survey of practice,1975-1976. Am J Psychiatry 735:479-82.

Asnis GM, Leopold MA. (1978) A single-blind study ofECT in patients with tardive dyskinesia. Am J Psychiatry

755:1235-7.

Avery D, Lubrano A. (1979) Depression treated withimipramine and ECT: The DeCarolis study reconsidered.Am J Psychiatry 136:559-62.

Avery DH, Claypoole K, Robinson L. (1999) Repetitivetranscranial magnetic stimulation in the treatment ofmedication-resistant depression: preliminary data. JNMD

787:114-117.

Avramov MN, Husain MM, White PF. (1995) Thecomparative effects of methohexital, propofol, andetomidate for electroconvulsive therapy. Anesth Analg

81:596-602.

Babington RG, Wedeking PW. (1975) Blockade of tardiveseizures in rats by electroconvulsive shock. Brain Res

#8:141-4.

Bader GM, Silk KR, Dequardo JR et al. (1995)Electroconvulsive therapy and cerebral aneurysm.Convuls Ther 77:139-43.

Bagadia VN, Abhyankar RR, Doshi J, Pradhan PV, ShahLP. (1983) A double blind controlled study of ECT vschlorpromazine in schizophrenia. J Assoc Phy sicians

India 37:637-40.

Bailine SH, Rifkin A, Kayne E et al. (2000) Comparisonof bifrontal and bitemporal ECT for major depression.Am J Psychiatry 757:121-123.

Bajc M, Medved V, Basic M, Topuzovik N, Babic D,Ivancevic D. (1989) Acute effect of electroconvulsivetherapy on brain perfusion assessed byTc99mhexamethylpropyleneamineoxim and single photonemission computed tomography. Acta Psychiatr Scand

80:421-6.

Balldin J, Berggren U, Lindstedt G, Modigh K. (1992)Neuroendocrine evidence for decreased function ofalpha2-adrenergic receptor after electroconvulsivetherapy. Psychiatry Res 41:257-65.

Balldin J, Eden S, Granerus AK, Modigh K, Svanborg A,Walinder J, Wallin, M. (1980b) Electroconvulsive therapyin Parkinson's syndrome with â!œon-offâ! ! phenomenon.J Neural Transm 47:11-21.

Balldin J, Granerus AK, Lindstedt G, Modigh K, WalinderJ. (1981) Predictors for improvement afterelectroconvulsive therapy in parkinsonian patients withon-off symptoms. J Neural Transm 52:199-211.

P.258

Balldin J, Granerus AK, Lindstedt G, Modigh K, WalinderJ. (1982) Neuroendocrine evidence for increasedresponsiveness of dopamine receptors in humansfollowing electroconvulsive therapy. Psychopharmacology

76:371-6.

Balldin, J. (1982) Factors influencing prolactin releaseinduced by electroconvulsive therapy. Acta Psychiatr

Scand 65:365-9.

Barbour GL, Blumenkrantz MJ. (1978) Videotape aidsinformed consent decision. JAMA 240:2741-2.

Barbui C, Hotopf M. (2001) Amitriptyline v. the rest: stillthe leading antidepressant after 40 years of randomizedcontrolled trials. Br J Psychiatry 178:129-144.

Barker JC, Baker AA. (1959) Deaths associated withelectroplexy. J Ment Sci 105: 339-48.

Barton JL. (1977) ECT in depression: The evidence ofcontrolled studies. Biol Psychiatry 72:687-95.

Barton JL, Mehta S, Snaith RP. (1973) The prophylacticvalue of extra ECT in depressive illness. Acta Psychiatr

Scand 49:386-92.

Bates WJ, Smeltzer DJ. (1982) Electroconvulsivetreatment of psychotic self-injurious behavior in apatient with severe mental retardation. Am J Psychi atry

759:1355-6.

Baxter LR, Jr, Liston EH. (1986) Informed ECT for

patients and families with Dr. Max Fink [videotapereview]. Convuls Ther 2:301-3.

Baxter LR, Roy-Byrne P, Liston EH, Fairbanks L. (1986)Informing patients about electroconvulsive therapy:effects of a videotape presentation. Convuls Ther 2:252-9.

Bayles S, Busse EW, Ebaugh FG. (1950) Square waves(BST) versus sine waves in electroconvulsive therapy.Am J Psychiatry 707:34-41.

Beale MD, Kellner CH, Pritchett JT, Bernstein HJ, BurnsCM, Knapp R. (1994) Stimulus dose-titration in ECT: A2-year clinical experience. Convuls Ther 70:171-6.

Behar R. (1991) The thriving cult of greed and power[Cover Story]. Time May 6, pp. 50-57.

Bellett S, Kershbaum A, Furst W. (1941) Theelectrocardiogram during electric shock treatment ofmental disorders. Am J of Men Sci 207:167-76.

Benbow SM. (1987) The use of electroconvulsive therapyin old age psychiatry. Int J Ceriat Psychiatry 2:25-30.

Benbow SM. (1987) ECT for depression in dementia. Br J


Bennett AE. (1983) Electroshock and Berkeley. Biol


Berg S, Gabriel AR, Impastato DJ. (1959) Comparativeevaluation of the safety of chlorpromazine and reserpineused in conjunction with ECT. J Neuropsychi atry 7:104-7.

Bergman PS, Gabriel AR, Impastato DJ, Wortis SB.(1952) EEG changes following ECT with the Reiterapparatus. Confinia Neurologica 72:347-51.

P.259

Bergsholm P, Gran L, Bleie H. (1984) Seizure duration inunilateral electroconvulsive therapy. The effect ofhypocapnia induced by hyperventilation and the effect ofventilation with oxygen. Acta Psychiatr Scand 69:121-8.

Bergsholm P, Larsen JL, Rosendahl K, Holsten F. (1989)Electroconvulsive therapy and cerebral computedtomography. A prospective study. Acta Psychiatr Scand

80:566-12.

Berman RM, Narasimhan M, Sanacora G et al. (2000) Arandomized clinical trial of repetitive transcranialmagnetic stimulation in the treatment of majordepression. Biol Psychiatry 47:332-337

Berne RM, Levy MN. (1981) Cardiovascular Physiology .St. Louis: C.V. Mosby, pp. 145-78.

Berrouschot J, Rolle K, Kuhn HJ et al. (1997) Serumneuron-specific enolase levels do not increase afterelectroconvulsive therapy. J Neural Sci 750:173-6.

Berson SA, Yalow RS. (1968) Radioimmunoassay of ACTHin plasma. J Clin Invest 47:2725-51.

Bertagnoli MW, Borchardt CM. (1990) A review of ECT forchildren and adolescents. J Am Acad Child Adolesc

Psychiatry 2:302-7.

Bidder TG, Strain JJ. (1970) Modifications ofelectroconvulsive therapy. Compr Psychiatry 77:507-17.

Bini L. (1938) Experimental researches on epilepticattacks induced by the electric current. Am J Psychiatry

(Suppl) 94:172-4.

Birkett DP. (1988) ECT in parkinsonism with affectivedisorder [letter]. Br J Psychiatry 752:712.

Blachly P, Cowing D. (1966) Multiple monitoredelectroconvulsive treatment. Compr Psychiatry 7:100-9.

Blachly PH, Semler HJ. (1967) Electroconvulsive therapyof three patients with aortic valve prostheses. Am J


Black DW, Wilcox JA, Stewart M. (1985) The use of ECTin children: case report. J din Psychiatry 46:98-9.

Black DW, Winokur G, Hulbert J, Nasrallah A. (1988)Predictors of immediate response in the treatment ofmania: The importance of comorbidity. Biol Psychiatry

24:191-8.

Black DW, Winokur G, Nasrallah H. (1987) Treatment ofmania: A naturalistic study of electroconvulsive therapyversus lithium in 438 patients. J Clin Psychiatry 48:132-9.

Blaurock MF, Lorimer FM, Segal MM, Gibbs FA. (1950)Focal electroencephalographic changes in unilateralelectric convulsive therapy. Arch Neural Psychiatry

64:220-6.

Bodley PO, Fenwick PBC. (1966) The effect ofelectroconvulsive therapy on patients with essentialhypertension. Br J Psychiatry 7/2:1241-9.

Boey WK, Lai FO. (1990) Comparison of propofol andthiopentone as anaesthetic agents for electroconvulsivetherapy. Anaesthesia 45:623-8.

Bohning DE, Shastri A, McConnell KA et al. (1999) Acombined TMS/fMRI study of intensity -dependent TMS

over motor cortex. Biol Psychiatry 45:385-94.

Bohning DE, Shastri A, McGavin L et al. (2000a) Motorcortex brain activity induced by 1-Hz transcranialmagnetic stimulation is similar in location and level tothat for volitional movement. Invest Radiol 35:676-683

Bohning DE, Shastri A, Wassermann EM et al. (2000b)BOLD-f MRI response to single-pulse transcranialmagnetic stimulation (TMS). J Magn Reson Imaging

77:569-574

Bolwig T. (1987) Training in convulsive therapy inDenmark. Convuls Ther 3: 156-7.

Bolwig T. (1988) Blood-brain barrier studies with specialreference to epileptic seizures. Acta Psychiatr Scand

Suppl 345:15-20.

Bolwig TG. (1984) The influence of electrically-inducedseizures on deep brain structures. In: B Lerer, RDWeiner, RH Belmaker, eds. ECT: Basic Mecha nisms.London: John Libbey, pp. 132-8.

Bolwig TG, Hertz MM, Holm-Jensen J. (1977) Blood-brainbarrier during electroshock seizures in the rat. Eur J Clin

Invest 7:95-100.

Bonne O, Krausz Y, Shapira B et al. (1996) Increasedcerebral blood flow in depressed patients responding toelectroconvulsive therapy. J Nucl Med 37:1075-1080

Bouckoms AJ, Welch CA, Drop L, Dao T, Kolton K. (1989)Atropine in electroconvulsive therapy. Convuls Ther

5:48-55.

Bourgon LN, Kellner CH. (2000) Relapse of depression

P.260

after ECT: A review. J ECT 16:19-31

Bourne H. (1954) Convulsion dependence. Lancet

2:1193-6.

Bourne H. (1956) Convulsion dependence and rationalconvulsion therapy. Journal of the Indian Medical

Profession 3:1-6.

Boyd DA, Brown DW. (1948) Electroconvulsive therapy inmental disorders associated with childbearing. Journal of

Missouri Medicine 45:573-9.

Boylan LS, Haskett RF, Mulsant BH et al. (2000)Determinants of seizure threshold in ECT:benzodiazepine use, anesthetic dosage, and otherfactors. J ECT 16:3-18.

Braasch ER, Demaso DR. (1980) Effect ofelectroconvulsive therapy on serum isoenzymes. Am J


Bracha S, Hess JP. (1956) Death occurring duringcombined reserpine-electroshock treatment. Am J


Brandon S. (1981) The history of shock treatment. In:RL Palmer, ed. Electrocon vulsive Therapy: An Appraisal .Oxford: Oxford University Press, pp. 310.

Brandon S, Cowley P, McDonald C, Neville P, Palmer R,Wellstood-Easton S. (1985) Leicester ECT trial: results inschizophrenia. Br J Psychiatry 746:177-83.

Brandon S, Cowley P, McDonald C, Neville P, Palmer R,Wellstood-Eason S. (1984) Electroconvulsive therapy:results in depressive illness from the Leicestershire trial.

Br Med J 288:22-25.

Breggin PR. (1979) Electroshock: Its Brain -Disabling

Effects. New York: Springer.

Breggin PR. (1980) Electroconvulsive therapy fordepression [letter]. N Engl J Med 303:1305-6.

Bridenbaugh RH, Drake FR, O'Regan TJ. (1972) Multiplemonitored electrocon vulsive treatment of schizophrenia.Compr Psychiatry 73:9-17.

Brill NQ, Crumpton E, Eiduson S, Grayson HM, HellmanLI, Richards RA. (1959) Relative effectiveness of variouscomponents of electroconvulsive therapy. Arch Neural


Brodersen P, Paulson OB, Bolwig TG, Rogon ZE,Rafaelsen OJ, Lassen NA. (1973) Cerebral hyperemia inelectrically induced epileptic seizures. Arch Neural

28:334-8.

Bross R. (1957) Near fatality with combined ECT andreserpine. Am J Psychiatry 773:933.

Brown GL. (1975) Parkinsonism, depression, and ECT.Am J Psychiatry 732:1084.

Brown GL, Wilson WP, Green RL, Jr. (1973) Mentalaspects of Parkinsonism and their management. In: SJBern, ed. Parkinson's Diseaseâ!” Rigidity, Akinesia,

Behavior , vol 2; Selected Communications on Topic.Toronto: Hans Huber, pp. 265-78.

Brumback RA. (1983) EEC monitoring of ECT. Br J


P.261

Brumback RA. (1987) EEG monitoring. Convuls Ther

3:151-7.

Buckholtz NS, Davies AO, Rudorfer MV, Golden RN,Potter WZ. (1988) Lymphocyte beta adrenergic receptorfunction versus catecholamines in depression. Biol


Burke WJ, Rubin EH, Zorumski CF, Wetzel RD. (1987)The safety of ECT in geriatric psychiatry. J Am Geriatr

Soc 35:516-21.

Burke WJ, Rutherford JL, Zorumski CF, Reich T. (1985)Electroconvulsive therapy and the elderly. Compr


Burrows SG. (1828) Commentaries on the Causes,Forms, Symptoms and Treatment, Moral and Medical, ofInsanity. London: T&G Underwood, pp. 656-7. [as citedin Sandford, 1966]

Burton TM. (1991) Medical flap. The Wall Street Journal ,April 19.

Calev A, Ben-Tzvi E, Shapira B, Drexler H, Carasso R,Lerer B. (1989) Distinct memory impairments followingelectroconvulsive therapy and imipramine. PsycholMed

79:111-9.

Calev A, Cohen R, Tubi N, Nigal D, Shapira B,Kugelmass, S, Lerer B. (1991b) Disorientation andbilateral moderately suprathreshold ECT. Convuls Ther 7:99-110.

Calev A, Drexler H, Tubi N, Nigal D, Shapira B,Kugelmass S, Lerer B. (199la) Atropine and cognitiveperformance after electroconvulsive therapy. Convuls

Ther 7:92-8.

Calev A, Nigal D, Shapira B. (1991) Early and long-termeffects of electroconvulsive therapy and depression onmemory and other cognitive functions. J Nerv MentDis

779:526-33

Calloway SP, Dolan RJ, Jacoby RJ, Levy R. (1981) ECTand cerebral atrophy. A computed tomographic study.Acta Psychiatr Scand 64:442-5.

Cammermeyer J. (1961) The importance of avoiding â!œdarkâ! ! neurons in experimental neuropathology. Acta

Neuropathol 7:245-70.

Cammermeyer J. (1972) Nonspecific changes of thecentral nervous system in normal and experimentalmaterial. In GH Bourne, ed. The Structure and Func tion

of Nervous Tissue , vol VI. New York: Academic Press.

Campbell D. (1960) The psychological effects of cerebralelectroshock. In: HJ Eysenck, ed. Handbook of Abnormal

Psychology: An Experimental Ap proach . London: Pitnam,pp. 611-33.

Cannicott SM. (1962) Unilateral electro-convulsivetherapy. Postgrad Med J 38: 451-9.

Cannicott SM, Waggoner RW. (1967) Unilateral andbilateral electroconvulsive therapy. Arch Gen Psychiatry

16:229-32.

Cantor C. (1986) Carbamazepine and ECT: A paradoxicalcombination. J Clin Psychiatry 47:216-1.

Caplan G. (1946) Electrical convulsion therapy in thetreatment of epilepsy. J Ment Sci 92:784.

Cardno AG, Simpson CJ. (1991) Electroconvulsivetherapy in Paget's disease and hydrocephalus. Convuls

Ther 7:48-51.

Carl C, Engelhardt W, Teichmann G, Fuchs G. (1988)Open comparative study with treatment -refractorydepressed patients: Electroconvulsive therapyâ!”anes-thetic therapy with isoflurane (preliminary report).Pharmacopsychiatry 21: 432-3.

Carlson GA, Goodwin FR. (1973) The stages of mania.Arch Gen Psychiatry 28: 221-8.

Carney MV, Sheffield BE (1973) Electroconvulsiontherapy and the diencephalon. Lancet 7:150-6.

Carney MW, Sheffield BF. (1972) Depression andNewcastle scales. Their relationship to Hamilton's scale.Br J Psychiatry 727:35-40.

Carney MWP, Roth M, Garside RF. (1965) The diagnosisof depressive syndromes and the prediction of ECTresponse. Br J Psychiatry 777:659-74.

Carney RM, Freedland KE, Jaffe AS. (2001) Depressionas a risk factor for coronary heart disease mortality.Arch Gen Psychiatry 58: 229-30.

Carr V, Dorrington C, Schader G, Wale J. (1983) The useof ECT in childhood bipolar disorder. Br J Psychiatry

143:411-5.

Carter C. (1977) Neurological considerations with ECT.Convulsive Therapy Bulletin and Tardive Dyskinesia

Notes 2:61-9.

Casey DA. (1991) Electroconvulsive therapy and

P.262

Friedrich's ataxia. Convuls Ther 7:45-7.

Casey DA. (1987) Electroconvulsive therapy in theneuroleptic malignant syndrome. Convuls Ther 3:278-83.

Castelli I, Steiner LA, Kaufmann MA, Alfille PH, SchoutenR, Welch CA, Drop LJ. (1995) Comparative effects ofesmolol and labetalol to attenuate hyperdynamic statesafter electroconvulsive therapy. Anesth Analg 80:557-61.

Cerletti U. (1940) L'Elettroshock. Rivista Sperimentale

Freniatria 64:2093-10.

Cerletti U. (1950) Old and new information aboutelectroshock. Am J Psychiatry 707:87-94.

Cerletti U. (1956) Electroshock therapy In: F Marti -Ibanez, AM Sackler, RR Sackler, eds. The Great

Physiodynamic Therapies in Psychiatry . New York:Hoeber-Harper, pp. 91-120.

Cerletti U, Bini L. (1938) Un nuevo metodo dishockterapie â!œL'elettroshock.â! ! Bol lettino Accademia

Medica Roma 64:136-8.

Chacko RC, Root L. (1983) ECT and tardive dyskinesia:Two cases and a review. J Clin Psychiatry 44:265-6.

Chanpattanna W, Buppanhavun W, Chakrabhand MLS.(2000) Seizure threshold in ECT: Differences betweeninstruments. American Psychiatric Association, NewResearch Abstracts, #NR406.

Chapman AH. (1961) Aortic dacron graft surgery andelectroshock: Report of a case. Am J Psychiatry 7/7:93-7.

Charatan FB, Oldham AJ. (1954) Electroconvulsivetreatment in pregnancy. J Obst Gynecol Br Empire

67:665-7.

Chater SN, Simpson KH. (1988) Effect of passivehyperventilation on seizure duration in patientsundergoing electroconvulsive therapy. Compr Psychiatry

60:70-3.

Chatrian GE, Petersen MC. (1960) The convulsivepatterns provoked by Indoklon, Metrazol andelectroshock: some depth electrographic observations inhuman patients. Electroencephalogr Clin Neurophysiol

72:7152-5.

Chaudhry HR, Hofmann P, Loimer N et al. (2000)Prolactin and beta-endorphin serum elevations after ECTin manic patients. Acta Psychiatr Scand 702: 386-9.

Checkley SA, Meldrum BS, McWilliam JR. (1984)Mechanism of action of ECT: neuroendocrine studies. In:B Lerer, RD Weiner, RH Belmaker, eds. ECT: Basic

Mechanisms . London: John Libbey, pp. 101-6.

Chen JJ, Velamati S, Stewart C. (1990) Detection ofprolonged seizure by audible EEC. Convuls Ther 6:248-50.

Chen R, Classen J, Gerloff C et al. (1997) Depression ofmotor cortex excitability by low-frequency transcranialmagnetic stimulation. Neurology 48:1398-403.

Christie JE, Whalley LJ, Brown NS, Dick H. (1982) Effectof ECT on the neuroendocrine response to apomorphinein severely depressed patients. Br J Psychiatry 740:268-73.

P.263

Cinca J, Evangelista A, Montoyo J, Barutell C, Figueras J,Valle V, Rius J, Soler-Soler J. (1985) Electrophysiologiceffects of unilateral right and left stellate ganglion blockon the human heart. Am Heart J 709:46-54.

Cizlado BC, Wheaton A. (1995) Case study: ECTtreatment of a young girl with catatonia. J Am Acad

Child Adolesc Psychiatry 34:332-5.

Clark CP, Alexopoulos GS, Kaplan J. (1995) Prolactinrelease and clinical response to electroconvulsivetherapy in depressed geriatric inpatients: A preliminaryreport. Convuls Ther 77:24-31.

Clement AJ. (1962) Atropine premedication forelectroconvulsive therapy. Br Med J n228-9.

Clinical Research Centre Division of Psychiatry (1984)The Northwick Park ECT trial: Predictors of response toreal and simulated ECT. Br J Psychiatry 144: 227-37.

Clyma EA. (1975) Unilateral electroconvulsive therapy:how to determine which hemisphere is dominant. Br J


Coffey CE, Figiel GS, Djang WT, Cress M, Saunders WB,Weiner RD. (1988b) Leukoencephalopathy in elderlydepressed patients referred for ECT. Biol Psychiatry

24:143-61.

Coffey CE, Figiel GS, Djang WT, Weiner RD. (1990a)Subcortical hyperintensity on magnetic resonanceimaging: A comparison of normal and depressedsubjects. Am J Psychiatry 45:187-9.

Coffey CE, Figiel GS, Weiner RD, Saunders WB. (1990b)Caffeine augmentation of ECT. Am J Psychiatry 147:579-

85.

Coffey CE, Lucke J, Weiner RD, Krystal AD, Aque M.(1995a) Seizure threshold in electroconvulsive therapy:I. Initial seizure threshold. Biol Psychiatry 37: 713-20.

Coffey CE, Lucke J, Weiner RD, Krystal AD, Aque M.(1995b) Seizure threshold in electroconvulsive therapy(ECT) II. The anticonvulsant effect of ECT. Biol


Coffey CE, McCall WV, Hoelscher TJ, Carroll BJ, HinklePE, Saunders WB, Erwin CW, Marsh OR, Weiner RD.(1988a) Effects of ECT on polysomnographic sleep: Aprospective investigation. Convuls Ther 4:269-79.

Coffey CE, Weiner RD, Djang WT, Figiel GS, Soady SAR,Patterson LJ, Holt PD, Spritser CE, Wilkinson WE. (1991)Brain anatomic effects of ECT: A prospective magneticresonance imaging study. Arch Gen Psychiatry 48: 1013-21.

Coffey CE, Weiner RD, Hinkle PE, Cress M, Daughtry G,Wilson WH. (1987a) Augmentation of ECT seizures withcaffeine. Biol Psychiatry 22:637-49.

Coffey CE, Weiner RD, McCall WV, Heinz ER. (1987b)Electroconvulsive therapy in multiple sclerosis: Amagnetic resonance imaging study of the brain. Con vuls

Ther 3:137-44.

Cohen BD, Noblin CD, Silverman AJ, Penick SB. (1968)Functional asymmetry of the human brain. Science

162:475-7.

Coleman EA, Sackeim HA, Prudic J et al. (1996)Subjective memory complaints prior to and followingelectroconvulsive therapy. Biol Psychiatry 59:346-56

P.264

Colenda CC, McCall WV. (1996) A statistical modelpredicting the seizure threshold for right unilateral ECTin 106 patients. Convuls Ther 72:3-12.

Conca A, Koppi S, Konig P et al. (1996) Transcranialmagnetic stimulation: A novel antidepressant strategy?Neuropsychobiology 34:204-7

Cooper AJ, Finlayson R, Velamoor VR, Magnus RV,Cernovsky Z. (1989) Effects of ECT on prolactin, LH, FSHand testosterone in males with major depressive illness.Can J Psychiatry 34:814-7.

Cooper SJ, Leahey W, Green DF, King DJ. (1988) Theeffect of electroconvulsive therapy on CSF aminemetabolites in schizophrenic patients. Br J Psychiatry

752:59-63.

Coote M, Wilkins A, Werstiuk ES et al. (1998) Effects ofelectroconvulsive therapy and desipramine onneuroendocrine responses to the clonidine challengetest. J Psychiatry Neurosci 23:172-8

Coppen A, Milln P, Harwood J, Wood K. (1985) Does thedexamethasone suppression test predict antidepressanttreatment success? Br J Psychiatry 146: 294-6.

Coppen A, Rao VA, Bishop M, Abou-Saleh MT, Wood K.(1980a) Neuroendocrine studies in affective disorders.Part 1. Plasma prolactin response to thyrotropin-releasing hormone in affective disorders: effect of ECT. JAffect Disord 2:311-5.

Coppen A, Rao VA, Bishop M, Abou-Saleh MT, Wood K.(1980b) Neuroendocrine studies in affective disorders.Part 2. Plasma thyroid-stimulating hormone response tothyrotropin-releasing hormone in affective disorders:

effect of ECT. J Affect Disord 2:317-20.

Coryell W. (1978) Intrapatient responses to ECT andtricyclic antidepressants. Am J Psychiatry 735:1108-10.

Coryell W. (1982) Hypothalamic-pituitary-adrenal axisabnormality and ECT response. Psychiatr Res 6:283-91.

Coryell W, Pfohl B, Zimmerman M. (1985) Outcomefollowing electroconvulsive therapy: A comparison ofprimary and secondary depression. Convuls Ther 7:10-4.

Coryell W, Zimmerman M. (1983) The dexamethasonesuppression test and ECT outcome: A six -month follow-up. Biol Psychiatry 78:21-7.

Coryell W, Zimmerman M. (1984) Outcome following ECTfor primary unipolar depression: A test of newlyproposed response predictors. Am J Psychiatry 747:862-7.

Costain DW, Cowen PJ. (1982) ECT and the growthhormone response to apomorphine [letter]. Br J

Psychiatry 747:213.

Costain DW, Cowen PJ, Gelder MG, Grahame-Smith DG.(1982) Electroconvulsive therapy and the brain:evidence for increased dopamine-mediated responses.Lancet 2:400-4.

Costello CG, Belton GP, Abra JC, Dunn BE. (1970) Theamnesic and therapeutic effects of bilateral andunilateral ECT. Br J Psychiatry 776:69-78.

Coull DC, Crooks J, Dingwall-Fordyce I, Scott AM, WeirRD. (1970) Amitriptyline and cardiac disease: risk ofsudden death identified by monitoring system. Lancet

2:590-1.

Couture LJ, Lucas LF, Lippmann SB, Shaltout T,Paloheimo MPJ, Edmonds HL, Jr. (1988b) Monitoringseizure duration during electroconvulsive therapy. Con

vuls Ther 4:2061-4.

Couture LJ, Thomas DR, Lippman SB, Edmonds HL, LucasLF. (1988a) Monitoring seizure duration in patientsundergoing electroconvulsive therapy. Anesth An alg

67:842.

Crider BA, Hansen-Grant S. (1995) Nonconvulsive statusepilepticus as a cause for delayed emergence afterelectroconvulsive therapy. Anesthesiology 82: 591-3.

Cronholm B. (1969) Post-ECT amnesias. In: GA Talland,N Waugh, eds. The Pa thology of Memory . New York:Academic Press, pp. 818-9.

Cronholm B, Blomquist C. (1959) Memory disturbancesafter electroconvulsive therapy: II. Conditions one weekafter a series of treatments. Acta Psychiatr Scand

32:182-5.

Cronholm B, Lagergren A. (1959) Memory disturbancesafter electroconvulsive therapy. Acta Psychiatr Scand

34:283-310.

Cronholm B, Molander L. (1957) Memory disturbancesafter electroconvulsive therapy: I. Conditions six hoursafter electroshock treatment. Acta Psychiatr Scand

32:280-306.

Cronholm B, Molander L. (1961) Memory disturbancesafter electroconvulsive therapy. IV. Influence of aninterpolated electroconvulsive shock on retention ofmemory material. Acta Psychiatr Scand 36:83-90.

P.265

Cronholm B, Molander L. (1964) Memory disturbancesafter electroconvulsive therapy. V. Conditions one monthafter a series of treatments. Acta Psychiatr Scand

40:212-6.

Cronholm B, Ottosson J-O. (1960) Experimental studiesof the therapeutic action of electroconvulsive therapy inendogenous depression. Acta Psychiatr Neurol Scand

55:69-102.

Cronholm B, Ottosson J-O. (1961a) Memory functions inendogenous depression before and afterelectroconvulsive therapy. Arch Gen Psychiatry 5:193-9.

Cronholm B, Ottosson J-O. (1961b) â!œCountershockâ! !in electroconvulsive therapy. Influence on retrogradeamnesia. Arch Gen Psychiatry 4:254-8.

Cronholm B, Ottosson J-O. (1963a) The experience ofmemory function after electroconvulsive therapy. Br J


Cronholm B, Ottosson J-O. (1963b) Ultrabrief stimulustechnique in electroconvulsive therapy. I. Influence onretrograde amnesia of treatments with the Either ESelectroshock apparatus, Siemens Konvulsator III and oflidocaine-modified treatment. J Nerv Ment Dis 737:117-23.

Cronin D, Bodley P, Potts L, Mather MD, Gardner RK,Tobin JC. (1970) Unilateral and bilateral ECT: A study ofmemory disturbance and relief from depression. J Neurol

Neurosurg Psychiatry 33:705-11.

Cropper CFJ, Hughes M. (1964) Cardiac arrest (withapnoea) after ECT. Br J Psychiatry 110:222-5.

Crow TJ, Johnstone EC. (1986) Controlled trials ofelectroconvulsive therapy. Ann N Y Acad Sci 462:12-29.

Cuche J-L, Brochier P, Klioua N, Poirier M-F, Cuche H,Benimould M, Loo H, Safer M. (1990) Conjugatedcatecholamines in human plasma: where are they comingfrom? J Lab Clin Med 7/6:681-6.

Culver CM, Ferrell RB, Green RM. (1980) ECT and specialproblems of informed consent. Am J Psychiatry 737:586-91.

Cummings JL. (1993) The neuroanatomy of depression. JClin Psychiatry 54 Suppl: 14-20.

Dalby NO, Tonder N, Wolby DP et al. (1996) No loss ofhippocampal hilar somatostatinergic neurons afterrepeated electroconvulsive shock: A combinedstereological and in situ hybridization study. Br J

Psychiatry 769:68-74

Dam AM, Dam M. (1986) Quantitative neuropathology inelectrically induced generalized convulsions. Convuls

Ther 2:77-89.

Daniel WE (1985) ECT-induced hyperactive delirium andbrain laterality. Am J Psychiatry 742:521-2.

Daniel WF, Crovitz HE (1983a) Acute memory impairmentfollowing electroconvulsive therapy. 1. Effects ofelectrical stimulus waveform and number of treatments.Acta Psychiatr Scand 67:1â!“7.

Daniel WF, Crovitz HE. (1983b) Acute memoryimpairment following electroconvulsive therapy. 2.Effects of electrode placement. Acta Psychiatr Scand 67:57-68.

P.266

Daniel WF, Crovitz HE (1986) Disorientation duringelectroconvulsive therapy: Technical, theoretical, andneuropsychological issues. Ann N Y Acad Sci 462: 293-306.

Daniel WF, Crovitz HE, Weiner RD, Swartzwelder HS,Kahn EM. (1985) ECT-induced amnesia and postictal EEGsuppression. Biol Psychiatry 20:344-8.

Daniel WF, Weiner RD, Crovitz HE (1983)Autobiographical amnesia with ECT: An analysis of theroles of stimulus wave form, electrode placement,stimulus energy, and seizure length. Biol Psychiatry

75:121-6.

Davidson J, McLeod M, Law-Yone B, Linnoila M. (1978)Comparison of electroconvulsive therapy and combinedphenelzine-amitriptyline in refractory depression. Arch


Davis JM, Janicak PG, Sakkas P, Gilmore C, Wang Z.(1991) Electroconvulsive therapy in the treatment of theneuroleptic malignant syndrome. Convuls Ther 7:111-20.

de Curtis M, Avanzini G. (1994) Thalamic regulation ofepileptic spike and wave discharges. Fund Neural 9:307-26

De Montigny C, Cournoyer G, Morissette R, Langlois R,Caille G. (1983) Lithium carbonate addition in tricyclicantidepressant resistant unipolar depression. Arch Gen


Deakin JF, Ferrier IN, Crow TJ, Johnstone EC, Lawler P.(1983) Effects of ECT on pituitary hormone release:relationship to seizure, clinical variables and outcome.Br J Psychiatry 143:618-24.

Dec GW, Stern TA, Welch C. (1985) The effects ofelectroconvulsive therapy on serial electrocardiogramsand serum cardiac enzyme values: A prospective studyof depressed hospitalized inpatients. JAMA 253:2525-9.

Decina P, Malitz S, Sackeim HA, Holzer J, Yudofsky S.(1984) Cardiac arrest during ECT modified by beta-adrenergic blockade. Am J Psychiatry 141:298-300.

Decina P, Sackeim HA, Kahn DA, Pierson D, Hopkins N,Malitz S. (1987) Effects of ECT on the TRH stimulationtest. Psychoneuroendocrinology 12:29-34.

d'Elia G. (1970) Comparison of electroconvulsive therapywith unilateral and bilateral stimulation: II. Therapeuticefficiency in endogenous depression. Acta Psychiatr

Scand Suppl 275:30-43.

d'Elia G, Raotma H. (1975) Is unilateral ECT lesseffective than bilateral ECT? Br J Psychiatry 726:83-9.

d'Elia G, Widepalm K. (1974) Comparison offrontoparietal and temporoparietal unilateralelectroconvulsive therapy. Acta Psychiatr Scand 50:225-32.

Deliyiannis S, Eliakim M, Bellet S. (1962) Theelectrocardiogram during electroconvulsive therapy asstudied by radioelectrocardiography. Am J Cardiol 10:187-92.

Delva NJ, Brunei D, Hawken ER et al. (2000) Electricaldose and seizure threshold: relations to clinical outcomeand cognitive effects in bifrontal, bitemporal, and rightunilateral ECT. J ECT 16:361-9

Delva NJ, Brunei DG, Hawken ER et al. (2001)

P.267

Characlerislics of responders and nonresponders lo brief-pulse righl unilateral ECT in a conlrolled clinical trial. yÂ£C7-77:118-23

Demulh GW, Rand BS. (1980) Alypical major depressionin a palienl with severe primary degeneration demenlia.Am J Psychiatry 737:1609-10.

Dennison S, French RN. (1989) Cardiac problems in ECT.Am J Psychiatry 146: 939.

Department of Heallh and Social Securily: Heallh NoticeHN(82)18. (1982) Health Service Management,

Psychiatric Services, Electro-convulsive Therapy:

Equipment. London: DHSS Store, Heallh Publicalions Unit

DeQuardo JR, Tandon R. (1988a) Concurrenl lithiumiherapy prevents ECT-induced switch to mania. J Clin


DeQuardo JR, Tandon R. (1988b) ECT in post-strokemajor depression. Convuls Ther 4:221-4.

Devanand DP, Bowers MB, Hoffman FJ, Jr, Sackeim HA.(1989) Acute and subacute effects of ECT on plasmaHVA, MHPG, and prolactin. Biol Psychiatry 26: 408-12.

Devanand DP, Briscoe KM, Sackeim HA. (1989) Clinicalfeatures and predictors of postictal excitement. Convuls

Ther 5:140-6.

Devanand DP, Decina P, Sackeim HA, Hopkins N,Novacenko H, Malitz S. (1987) Serial dexamethasonesuppression tests in initial suppressors andnonsuppresssors treated with electroconvulsive therapy.Biol Psychiatry 22:463-72.

Devanand DP, Decina P, Sackeim HA, Prudic J. (1988a)Status epilepticus following ECT in a patient receivingtheophylline. J Clin Psyche/pharmacology 5:153.

Devanand DP, Dwark AJ, Hutchinson ER, Bolwig TG,Sackeim HA. (1994) Does ECT alter brain structure? Am


Devanand DP, Lisanby SH, Nobler MS et al. (1998) Therelative efficiency of altering pulse frequency or trainduration when determining seizure threshold. J ECT

74:227-35.

Devanand DP, Sackeim HA, Decina P, Prudic J. (1988b)The development of mania and organic euphoria duringECT. J Clin Psychiatry 49:69-1.

Devanand DP, Shapira B, Petty F et al. (1995) Effects ofelectroconvulsive therapy on plasma GABA. Convuls Ther

77:3-13.

Devanand DP, Verma AK, Tirumalasetti F, Sackeim HA.(1991) Absence of cognitive impairment after more than100 lifetime ECT treatments. Am J Psychiatry 148:929-32.

Devanand DP, Lisanby S, Lo ES et al. (1998) Effects ofelectroconvulsive therapy on plasma vasopressin andoxytocin. Biol Psychiatry 44:610-6.

Dewald PA, Margolis NM, Weiner H. (1954) Vertebralfractures as complications of electroconvulsive therapy.JAMA 754:981-4.

di Michele V, Giordano L, de Cataldo S, Sabatini MD,Petruzzi C, Casacchia M, Rossi A. (1992)Electroencephalographic seizure duration inelectroconvulsive therapy: A clinical study. Convuls Ther

8:258-61.

Dinan TG, Barry S. (1989) A comparison ofelectroconvulsive therapy with a combined lithium andtricyclic combination among depressed tricyclicnonresponders. Acta Psychiatr Scand 80:97-100.

Dodwell D, Goldberg D. (1989) A study of factorsassociated with response to electroconvulsive therapy inpatients with schizophrenic symptoms. Br J Psychiatry

154:635-9.

Donahue AB. (2000) Electroconvulsive therapy andmemory loss: A personal journey. J ECT 16:133-43.

Dored G, Stefansson S, d'Elia G, Kagedal B, Karlberg E,Ekman R. (1990) Corticotropin, cortisol and beta-endorphin responses to the human corticotropin -releasing hormone during melancholia and afterunilateral electroconvulsive therapy. Acta Psychiatr

Scand 82:204-9.

Dornbush RL. (1972) Memory and induced ECTconvulsions. Seminars in Psychi atry 4:47-54.

Dornbush RL, Williams M. (1974) Memory and ECT. In: MFink, S Kety, J. McGaugh, T Williams, eds. Psychobiology

of Convulsive Therapy . Washington, DC: VH Winston andSons, pp. 199-205.

Douglas CJ, Schwartz HI. (1982) ECT for depressioncaused by lupus cerebritis: A case report. Am J

Psychiatry 139:1631â!“2.

Douyon R, Serby M, Kluteh K, Rotroseu J. (1989) ECTand Parkinson's disease revisited: A â!œnaturalisticâ! !study. Am J Psychiatry 146:1451-5.

P.268

Dressier DM, Folk J. (1975) The treatment of depressionwith ECT in the presence of brain tumor. Am J


Drop LJ, Bouckoms AJ, Welch CA. (1988) Arterialhypertension and multiple cerebral aneurysms in apatient treated with electroconvulsive therapy. J Clin


Drop LJ, Viguera A, Welch CA. (2000) ECT in patientswith intracranial aneurysm [letter]. J ECT 16:71-2.

Drop LJ, Welch CA. (1989) Anesthesia forelectroconvulsive therapy in patients with majorcardiovascular risk factors. Convuls Ther 5:88-101.

Dubin WR, Jaffe RL, Roemer RA, Lipschutz L, Spencer M.(1989) Maintenance ECT in coexisting affective andneurologic disorders. Convuls Ther 5: 162-7.

Dubovsky SL, Gay M, Franks RD, Haddenhorst A. (1985)ECT in the presence of increased intracranial pressureand respiratory failure: case report. J Clin Psychiatry

46:489-91.

Duman RS, Vaidya VA. (1998) Molecular and cellularactions of chronic electroconvulsive seizures. J ECT

74:181-93.

Durrant BW. (1966) Dental care in electroplexy. Br J

Psychiatry 7/2:1173-6.

Dwyer R, McCaughey W, Lavery J, McCarthy G, DundeeJW. (1988) Comparison of propofol and methohexitoneas anaesthetic agents for electroconvulsive therapy.Anaesthesia 43:459-62.

Dykes S, Scott AIF, Gow SM, Whalley LJ. (1987) Effectsof seizure duration on serum TSH concentration afterECT. Psychoneuroendocrinology 72:477-82.

Ebadi M, Pfeiffer RE, Murrin LC. (1990) Pathogenesis andtreatment of neuroleptic malignant syndrome. Gen

Pharmacol 27:367-86.

Edwards RM, Stoudemire A, Vela MA, Morris R. (1990)Intraocular pressure changes in nonglaucomatouspatients undergoing electroconvulsive therapy. Convuls

Ther 6:209-13.

Egbert LD, Wolfe S. (1960) Evaluation of methohexitalfor premedication in electroshock therapy. Anesth Analg

39:416-9.

Eitzman DT, Bach DS, Rubenfire M. (1994) Managementof myocardial stunning associated with electroconvulsivetherapy guided by hyperventilation echocardiography. Am

Heart J 727:928-9.

el -Ganzouri AR, Ivankovich AD, Braverman B, McCarthyR. (1985) Monoamine oxidase inhibitors: should they bediscontinued preoperatively? Anesth Analg 64:592-6.

Elithorn A, Bridges PK, Hodges JR, Jones MT. (1968)Adrenocortical responsiveness during courses ofelectroconvulsive therapy. Br J Psychiatry 774:575-80.

Elliot DL, Linz DH, Kane JA. (1982) Electroconvulsivetherapy: Pretreatment medical evaluation. Arch Intern

Med 742:979-81.

el -Mallakh RS. (1988) Complications of concurrentlithium and electroconvulsive therapy: A review ofclinical material and theoretical considerations. Biol


P.269

Emrich HM, Holt V, Kissling W, Fischler M, Lapse H,Heinemann H, von Zerssen D, Herz A. (1979) (3-endorphin-like immunoreactivity in cerebrospinal fluidand plasma of patients with schizophrenia and otherneuropsychiatric disorders. Pharmakopsychiatrie 72:269-76.

Ende G, Braus DF, Walter S et al. (2000) Thehippocampus in patients treated with electroconvulsivetherapy: A proton magnetic resonance spectroscopicimaging study. Arch Gen Psychiatry 57:937-43.

Endler NS. (1988) The origins of electroconvulsivetherapy (ECT). Convuls Ther 4: 5-23.

Endler NS, Persad E. (1988) Electroconvulsive Therapy:

The Myths and The Real ities. Toronto: Hans Huber.

Engelhardt W, Carl G, Hartung E. (1993) Intra-individualopen comparison of burst-suppression-isoflurane-anesthesia versus electroconvulsive therapy in thetreatment of severe depression. Eur J Anaesthesiol

70:113-8.

Engle J, Jr, Duhl DE, Phelps ME. (1982) Patterns ofhuman local cerebral glucose metabolism during epilepticseizures. Science 218:64-6.

Enns M, Peeling J, Sutherland GR. (1996) Hippocampalneurons are damaged by caffeine -augmentedelectroshock seizures. Biol Psychiatry 40:642-7.

Enns N, Karvelas L. (1995) Electrical dose titration forelectroconvulsive therapy: A comparison with doseprediction methods. Convuls Ther 77:86-93.

Epstein, CM, Figiel GS, McDonald WM et al. (1999) Rapidrate transcranial magnetic stimulation in young andmiddle-aged refractory depressed patients. Psychiatry

Annals 28:36-9.

Epstein, HM, Fagman W, Bruce DL, Abram A. (1975)Intraocular pressure changes during anesthesia forelectroshock therapy. Anesth Analg 54:479-81.

Eriksson PS, Perfilieva E, Bjork-Eriksson T et al. (1998)Neurogenesis in the adult human hippocampus. Nat Med

4:1313-7.

Erman MK, Welch CA, Mandel MR. (1979) A comparisonof two unilateral ECT electrode placements: efficacy andelectrical energy considerations. Am J Psychiatry

736:1317-9.

Escalona PR, Coffey CE, Maus-Feldman J. (1991)Electroconvulsive therapy in a depressed patient with anintracranial arachnoid cyst: A brain magnetic resonanceimaging study. Convuls Ther 7:133-8.

Everman PD, Kellner CH, Deale MD et al. (1999) Modifiedelectrode placement in patients with neurosurgical skulldefects, [letter]. J ECT 75:237-9.

Faber R. (1983) Dental fracture during ECT [letter]. Am


Fall PA, Ekman R, Granerus AK. et al. (1995) ECT inParkinson's disease. Changes in motor symptoms,monoamine metabolites and neuropeptides. J Neural

Transm Park Dis Dement Sect 10:129-40.

Farah A, McCall WV. (1993) Electroconvulsive therapystimulus dosing: A survey of contemporary practices.Convuls Ther 9:90-4.

P.270

Farah A, McCall WV, Amundson RH. (1996) ECT aftercerebral aneurysm repair. Convuls Ther 72:165-70.

Fear C, Littlejohns CS, Rouse E, McQuail P. (1994)Propofol anaesthesia in electroconvulsive therapy.Reduced seizure duration may not be relevant. Br J


Feldman MJ. (1951) A prognostic scale for shocktherapy. Psychological Monograph No. 327.

Ferraro A, Roizin L, Helfand M. (1946) Morphologicchanges in the brain of monkeys following convulsionselectrically induced. J Neuropath Exper Neural 5:285-308.

Ferraro TN, Golden GT, Hare TA. (1990) Repeatedelectroconvulsive shock selectively alters -/-aminobutyricacid levels in rat brain: effect of electrode placement.Convuls Ther 6:199-208.

Fetterman JL. (1942) Electrocoma therapy of thepsychoses. Ann Intern Med 17: 775-89.

Figiel GS, DeLeo B, Zorumski CF, Baker K, Goewert A,Jarvis M, Smith DS, Mattingly G, Ruwitch J. (1993)Combined use of labetalol and nifedipine in controllingthe cardiovascular response from ECT. J Geriatr

Psychiatry Neu rol 6:20-4.

Figiel GS, Stoudemire A. (1994) The use of ECT forelderly patients with cardiac disease. Psychiatric Times,13-7. December.

Fink M. (1979) Convulsive Therapy: Theory and Practice .New York: Raven Press.

Fink M. (1983) Missed seizures and the bilateral-unilateral electroconvulsive therapy controversy [letter].Am J Psychiatry 740:198-9.

Fink M. (1984) Meduna and the origins of convulsivetherapy. Am J Psychiatry 141: 1034-41.

Fink M. (1986a) Neuroendocrine predictors ofelectroconvulsive therapy outcome: dexamethasonesuppression test and prolactin release. Ann N Y Acad Sci

462: 30-6.

Fink M. (1986b) Training in convulsive therapy[editorial]. Convuls Ther 2: 227-9.

Fink M. (1987) Douglas Goldman (1906-1986) [obituary].Convuls Ther 5:163.

Fink M. (1988) ECT for Parkinson's Disease? [editorial].Convuls Ther 4:189-91.

Fink M. (1990) How does convulsive therapy work?Neuro Psychopharmacology 3: 73-82.

Fink M. (1999a) ECT in delirious states [editorial]. J ECT

75:175-6.

Fink M. (1999b) Delirious mania. Bipolar Disord 7:54-60.

Fink M. (2001) Modal ECT is effective: response to letterby Sackeim et al. [letter]. J ECT 17:222-4.

Fink M, Abrams R, Bailine S et al. (1996) Ambulatoryelectroconvulsive therapy: report of a task force of theAssociation for Convulsive Therapy. Convuls Ther 72:42-55.

Fink M, Abrams, R. (1998) EEG monitoring in ECT: Aguide to treatment efficacy. Psychiatric Times, May,1998 (pp. 70-2).

Fink M, Bender MB, Green M. (1951) Patterns inperception of simultaneous tests of face and hand. Arch

Neural Psychiatry 66:355-62.

Fink M, Green M. (1950) Subdural hematoma developingduring hospitalization. Am J Psychiatry 707:381-383.

Fink M, Gujavarty K, Greenberg L. (1987) Serialdexamethasone suppression tests and clinical outcome inECT. Convuls Ther 3:111-20.

Fink M, Johnson L. (1982) Monitoring the duration ofelectroconvulsive therapy seizures: â!œcuffâ! ! and EEGmethods compared. Arch Gen Psychiatry 39: 1189-91.

Fink M, Kahn RL. (1957) Relation of EEG delta activity tobehavioral response in electroshock: Quantitative serialstudies. Arch Neural Psychiatry 78:516-25.

Fink M, Kahn RL. (1961) Behavioral patterns inconvulsive therapy. Arch Gen Psychiatry 5:30-6.

Fink M, Kahn RL, Korin H. (1959a) Relation of tests ofaltered brain function to behavioral change followinginduced convulsions. In: L van Bogaert, J Radermaker,eds. First International Congress of Neurological

Sciences . London: Pergamon Press, pp. 613-9.

Fink M, Nemeroff CB. (1989) A neuroendocrine view ofECT. Convuls Ther 5: 296-304.

Fink M, Ottosson JO. (1980) A theory of convulsive

P.271

therapy in endogenous depression: significance ofhypothalamic functions. Psychiatr Res 2:49-61.

Fink M, Stein JM. (1952) A clinical evaluation of carotidangiography. Confin Neu rologica 72:181-95.

Finlayson AJ, Vieweg WV, Wilkey WD, Cooper AJ. (1989)Hyponatremic seizure following ECT. Can J Psychiatry

34:463-4.

Fisher KA. (1949) Changes in test performance ofambulatory depressed patients undergoing ECT. J Gen

Psychol 41:195-232.

Fisman M. (1988) Intractable depression andpseudodementia: A report of two cases. Can J Psychiatry

35:628-30.

Flaherty JA, Naidu J, Dysken M. (1984) ECT, emergentdyskinesia, and depression. Am J Psychiatry 747:808-9.

Fleming GWTH, Golla FC, Walter WG. (1939) Electricconvulsion therapy of schizophrenia. Lancet 2:1353-5.

Fleminger JJ, de Home DJ, Nair NPV, Nott PN. (1970)Differential effect of unilateral and bilateral ECT. Am J


Florkowski CM, Crozier IG, Nightingale S et al. (1996)Plasma cortisol, PRL, ACTH, AVP and corticotrophinreleasing hormone responses to direct currentcardioversion and electroconvulsive therapy. Clin

Endocrinol 44:163-8.

Fochtmann LJ. (1994a) Animal studies ofelectroconvulsive therapy: foundations for futureresearch. Psychopharmacol Bull 30:321-444.

Fochtmann LJ. (1994b) What do rodents and test tubesteach us about ECT? Convuls Ther 70:287-9.

Folkerts H. (1996) The ictal electroencephalogram as amarker for the efficacy of electroconvulsive therapy. Eur

Arch Psych Clin Neurosci 246:155-64.

Folstein MF, Folstein SW, McHugh PR. (1975) â!œMiniMental State,â! ! a practical method of grading thecognitive state of patients for the clinician. J Psychiatr

Res 72:189-98.

Forssman H. (1955) Follow-up study of 16 childrenwhose mothers were given electric convulsive therapyduring gestation. Acta Psychiatr Neurol Scand 30:437-41.

Foster MW, Gayle RF. (1955) Dangers in combiningreserpine (Serpasil) with electroconvulsive therapy. JAMA

159:1520-2.

Foster S, Ries R. (1988) Delayed hypertension withelectroconvulsive therapy. J Nerv Ment Dis 776:374-6.

Foulds GA. (1952) Temperamental differences in mazeperformance: II. The effect of distraction and ofelectroconvulsive therapy on psychomotor retardation. Br

J Psychol 43:33-41.

Fox HA, Rosen A, Campbell RJ. (1989) Are brief pulseand sine wave ECT equally efficient? J Clin Psychiatry

50:432-5.

Fox P, Ingham R, George MS et al. (1997) Imaginghuman intra-cerebral connectivity by PET during TMS.Neuroreport 8:2787-91.

P.272

Frankel FH. (1973) Electro-convulsive therapy inMassachusetts: A task force report. Massachusetts J

Ment Health 3:3-29.

Frankel FH. (1977) Current perspectives on ECT: Adiscussion. Am J Psychiatry 734:1014-9.

Frankel FH. (1982) Medicolegal and ethical aspects oftreatment. In: R Abrams, WB Essman, eds.Electroconvulsive Therapy: Biological Foundations and

Clini cal Applications . New York: Spectrum Publications,pp. 245-58.

Frankel FH. (1986) Informed ECT for Health

Professionals, with Dr. Max Fink [videotape review].Convuls Ther 2:30-3.

Fraser RM, Glass IB. (1980) Unilateral and bilateral ECTin elderly patients. A comparative study. Acta Psychiatr

Scand 62:13-31.

Frederiksen SO, d'Elia G. (1979) Electroconvulsivetherapy in Sweden. Br J Psychiatry 734:283-7.

Freeman CP, Kendell RE. (1980) ECT: I. Patients'experiences and attitudes. Br J Psychiatry 737:8-16.

Freeman CP, Weeks D, Kendell RE. (1980) ECT: II:Patients who complain. Br J Psychiatry 737:17-25.

Freeman CPL, Basson JV, Crighton A. (1978) Double-blind controlled trial of electroconvulsive therapy (ECT)and simulated ECT in depressive illness. Lancet 7:738-40.

Freeman CPL, Cheshire KE. (1988) Attitude studies on

electroconvulsive therapy. Convuls Ther 2:31-42.

Freese KJ. (1985) Can patients safely undergoelectroconvulsive therapy while receiving monoamineoxidase inhibitors? Convuls Ther 7:190-4.

Frey R, Heiden A, Scharfetter J et al. (2001) Inverserelation between stimulus intensity and seizure duration:implications for treatment procedure. J ECT 77:102-8.

Fricchione GL, Kaufman LD, Gruber BL, Fink M. (1990)Electroconvulsive therapy and cyclophosphamide incombination for severe neuropsychiatric lupus withcatatonia. Am J Med 88:442-3.

Fried D, Mann JH. (1988) Electroconvulsive treatment ofa patient with known intracranial tumor. Biol Psychiatry

23:176-80.

Friedberg J. (1976) Shock Treatment Is Not Good For

Your Brain . San Francisco: Glide Publications.

Friedberg J. (1977) Shock treatment, brain damage, andmemory loss: A neurological perspective. Am J


Friedel RO. (1986) The combined use of neuroleptics andECT in drug resistant schizophrenic patients.Psychopharmacol Bull 22:928-30.

Friedman E. (1942) Unidirectional electrostimulatedconvulsive therapy. I: The effect of wave form andstimulus characteristics on the convulsive dose. Am J


Friedman E, Wilcox PH. (1942) Electrostimulatedconvulsive doses in intact humans by means ofunidirectional currents. J Nerv Ment Dis 96:56-63.

P.273

Fromholt P, Christensen AL, Stromgren LS. (1973) Theeffects of unilateral and bilateral electroconvulsivetherapy on memory. Acta Psychiatr Scand 49: 466-78.

Fromm GH. (1959) Observation on the effects ofelectroshock treatment in patients with parkinsonism.Bulletin of Tulane University 78:71-3.

Funkenstein DH, Greenblatt M, Solomon HC. (1952) Anautonomic nervous system test of prognostic significancein relation to electroshock treatment. Psycho som Med

74:347-62.

Gaitz CM, Essa M. (1991) Propranolol in ECT. Convuls

Ther 7:60-1.

Gaitz CM, Pokorny AD, Mills M, Jr. (1956) Deathfollowing electroconvulsive therapy. Arch Neural


Galen RS, Gambino SR. (1975) Beyond Normality: The

Predictive Value and Efficiency of Medical Diagnoses.New York: John Wiley, p. 17.

Gallinek A. (1952b) Organic sequelae of electricconvulsive therapy including facial and body dysgnosias.J Nerv Ment Dis 7/5:377-93.

Gallinek A, Kalinowsky LB. (1958) Psychiatric aspects ofmultiple sclerosis. Dis Nerv Syst 79:77-80.

Galynker II, Weiss J, Ongseng F et al. (1997) ECTtreatment and cerebral perfusion in Catalonia. J Nucl

Med 38:251-4.

Gangadhar BN, Girish K, Janakiramiah N et al. (1998)Formula method for stimulus setting in bilateralelectroconvulsive therapy: relevance of age. J ECT 14:259-65.

Gangadhar BN, Kapur RL, Kalyanasundaram S. (1982)Comparison of electroconvulsive therapy with imipraminein endogenous depression: A double blind study. Br J


Gangadhar BN, Lakshmanna G, Subba Krishna DK,Channabasavanna SM. (1985) Impedance measurementsduring electroconvulsive therapy. NIMHANS Jour nal

3:135-9.

Gardner MW, Kellner CH. (1998) Safe use of ECT with anintracranial aneurysm [letter]. J ECT 14:290-1.

Gassell MM. (1960) Deterioration after electroconvulsivetherapy in patients with intracranial meningioma. Arch


Geddes LA. (1987) Optimal stimulus duration forextracranial cortical stimulation. Neurosurgery 20: 94-9.

George MS, Lisanby SH, Sackeim HA. (1999)Transcranial magnetic stimulation: applications inpsychiatry. Arch Gen Psychiatry 56:300-11.

George MS, Nahas Z, Molloy M et al. (2000) A controlledtrial of daily left prefrontal cortex TMS for treatingdepression. Biol Psychiatry 48:962-10.

George MS, Wasserman EM, Kimbrell TA. (1997) Moodimprovement following daily left prefrontal repetitivetranscranial magnetic stimulation in patients withdepression: A placebo-controlled crossover trial. Am J

Psychiatry 154: 1752-6.

Geretsegger C, Rochowanski E, Kartnig C et al. (1998)Propofol and methohexital as anesthetic agents forelectroconvulsive therapy (ECT) a comparison of seizure-quality measures and vital signs. J ECT 74:28-35.

Gerring JP, Shields HM. (1982) The identification andmanagement of patients with a high risk for cardiacarrhythmias during modified ECT. J Clin Psychiatry

43:140-3.

Ghadirian AM, Gianoulakis C, Nair NP. (1988) The effectof electroconvulsive therapy on endorphins indepression. Biol Psychiatry 23:459-64.

Gibbons JL, McHugh PR. (1962) Plasma cortisol indepressive illness. J Psychiatr Res 7:162-71.

Gibson TC, Leaman DM, Devors J, Lepeschkin BE. (1973)Pacemaker function in relation to electroconvulsivetherapy. Chest 63:1025-7.

Gilbert DT. (1981) Shock therapy and informed consent.Illinois Bar Journal January: 272-87.

Gill D, Lambourn J. (1979) Indications for electricconvulsion therapy and its use by senior psychiatrists. Br

Med J 7:1169-71.

Gilmore JH, Isley MR, Evans DL, Kong LS, Ekstrom D,Kafer ER, Golden RN, et al. (1991) The reliability ofcomputer -processed EEG in the determination of ECTseizure duration. Convuls Ther 7:166-74.

Glassman A, Kantor SJ, Shostak M. (1975) Depression,delusions and drug response. Am J Psychiatry 732:716-9.

P.274

Glassman AH, Perel JM, Shostak M, Kantor SJ, Fleiss JL.(1977) Clinical implications of imipramine plasma levelsfor depressive illness. Arch Gen Psychiatry 34:197-204.

Gleiter CH, Deckert J, Nutt DJ, Marangos PJ. (1989)Electroconvulsive shock (ECS) and the adenosineneuromodulatory system: effect of single and repeatedECS on the adenosine Al and A2 receptors, adenylatecyclase, and the adenosine uptake site. J Neurochem

52:641-6.

Glenisson L, Auriacombe M, Fernandez T et al. (1998)Adverse seizure reactions after electroconvulsivetherapy. Study of personal cases and review of theliterature [French]. Encephale 24:1-8.

Globus JH, van Harreveld A, Wiersma CAG. (1943) Theinfluence of electric current applications on the structureof the brain of dogs. J Neumpathol Exp Neural 2:263-76.

Glynn RJ, Field TS, Rosner B, Hebert PR, et al. (1995)Evidence for a positive linear relation between bloodpressure and mortality in elderly people. Lancet

345:825-29.

Goetz CG, Olanow W, Roller WC, Penn RD, Cahill D,Morantz R, Stebbins G, Tonner CT, Klawans HL, ShannonKM, Cornelia CL, Witt T, Cox C, Waxhan M, Gauger L.(1989) Multicenter study of autologous adrenal medullarytransplantation to the corpus striatum in patients withadvanced Parkinson's disease. N Engl J Med 320:337-41.

Goldman D. (1949) Brief stimulus electric shock therapy.J Nerv Ment Dis 110: 36-45.

Goldstein MZ, Jensvold MF. (1989) ECT treatment of anelderly mentally retarded man. Psychosomatics 30:104-6.

Goldstein MZ, Richardson C. (1988) Meningioma withdepression: ECT risk or benefit ? Psychosomatics 29:349-51.

Gombos Z, Spiller A, Cottrell GA et al. (1999) Mossyfiber sprouting induced by repeated electroconvulsiveshock seizures. Brain Res $44:28-33.

Gomez J. (1975) Subjective side-effects of ECT. Br J


Goodman JA, Krahn LE, Smith GE et al. (2000) Patientsatisfaction with electroconvulsive therapy. Biol


Gordon D. (1981) The electrical and radiological aspectsof ECT. In: RL Palmer, ed. Electroconvulsive Therapy: An

Appraisal . Oxford: Oxford University Press, pp. 79-96.

Gordon D. (1982) Electro-convulsive therapy withminimum hazard. Br J Psychiatry 747:12-18.

Gosek E, Weller RA. (1988) Improvement of tardivedyskinesia associated with electroconvulsive therapy. JNerv Ment Dis 776:120-2.

Goswami U, Dutta S, Kuruvilla K, Papp. E, Perenyi A.(1989) Electroconvulsive therapy in neuroleptic-inducedparkinsonism. Biol Psychiatry 26:234-8.

Gottlieb G, Wilson I. (1965) Cerebral dominance:temporary disruption of verbal memory by unilateralelectroconvulsive shock treatment. J Comp Physiol

Psychol 60:368-70.

Gould E, Tanapat P, Rydel T et al. (2000) Regulation of

P.275

hippocampal neurogenesis in adulthood. Biol Psychiatry

48:715-20.

Goumeniouk AD, Fry PD, Zis AP. (1990) Abdominal aorticaneurysm and ECT. J ECT 6:55-7.

Gravenstein JS, Anton AH, Weiner SM, Tetlow AG. (1965)Catecholamine and cardiovascular response toelectroconvulsion therapy in man. Compr Psychiatry

37:833-9.

Graybar G, Goethe J, Levy T, Phillips J, Youngberg J,Smith D. (1983) Transient large upright T -waves on theelectrocardiogram during multiple monitoredelectroconvulsive therapy. Anesthesiology 59:467-9.

Green M, Fink M. (1957) Clinical andelectroencephalographic effects of megimide in patientswithout cerebral disease. Neurology 8:682-5.

Green R, Woods A. (1955) Effects of modified electro-convulsive therapy on the electrocardiogram. Br Med J

7:1503.

Greenbank RK. (1958) Aortic homograft surgery andelectroshock: case report. Am J Psychiatry 115:469.

Greenberg LB. (1985) Detection of prolonged seizuresduring electroconvulsive therapy: A comparison ofelectroencephalogram and cuff monitoring. Convuls Ther

1:32-7.

Greenberg LB, Anand A, Roque CT, Grinberg Y. (1986)Electroconvulsive therapy and cerebral venous angioma.Convuls Ther 2:197-202.

Greenberg LB, Gage J, Vitkun S, Fink M. (1987)

Isoflurane anesthesia therapy: A replacement for ECT indepressive disorders? Convuls Ther 3:269-71.

Greenberg LB, Gujavarty K. (1985) The neurolepticmalignant syndrome: review and report of three cases.Compr Psychiatry 26:63-70.

Greenberg LB, Mofson R, Fink M. (1988) Prospectiveelectroconvulsive therapy in a delusional depressedpatient with a frontal meningioma. Br J Psychiatry

755:105-7.

Greenblatt M, Grosser GH, Wechsler H. (1964)Differential response of hospitalized depressed patientsin somatic therapy. Am J Psychiatry 120:935-43.

Greffe J, Lemoine P, Lacroix C et al. (1996) Increasedserum levels of neuron-specific enolase in epilepticpatients and after electroconvulsive therapyâ!”apreliminary report. Clin Chim Acta 244:199-208.

Griesemer DA, Kellner CH, Beale MD et al. (1997)Electroconvulsive therapy for treatment of intractableseizures. Initial findings in two children. Neurology

49:1389-92.

Griffiths EJ, Lorenz RP, Baxter S, Talon NS. (1989) Acuteneurohumoral response to electroconvulsive therapyduring pregnancy. A case report. J Reprod Med 4:907-11.

Grigg JR. (1988) Neuroleptic malignant syndrome andmalignant hyperthermia. Am J Psychiatry 745:1175.

Grinspoon L, Greenblatt M. (1963) Pharmacotherapycombined with other treatment methods. Compr


P.276

Grisaru N, Chudakov B, Yaroslavsky Y et al. (1998)Transcranial magnetic stimulation in mania: a controlledstudy. Am J Psychiatry 755:1608-10

Grogan R, Wagner DR, Sullivan T, Labar D. (1995)Convuls Ther 77:51-6.

Grunhaus L. (1991) The technique of ECT [videotapereview]. Convuls Ther 7: 143-4.

Grunhaus L, Pande AC, Haskett RF. (1990) Full andabbreviated courses of maintenance electroconvulsivetherapy. Convuls Ther 6:130-8.

Grunhaus L, Shipley JE, Eiser A et al. (1997) Sleep-onsetrapid eye movement after electroconvulsive therapy ismore frequent in patients who respond less well toelectroconvulsive therapy. Biol Psychiatry 42:191-200.

Grunhaus L, Dannon PN, Schreiber S et al. (2000)Repetitive transcranial magnetic stimulation is aseffective as electroconvulsive therapy in the treatmentof nondelusional major depressive disorder: an openstudy. Biol Psychiatry 15: 314-24.

Gujavarty K, Greenberg LB, Fink M. (1987)Electroconvulsive therapy and neuroleptic medication intherapy-resistant positive-symptom psychosis. Convuls

Ther 3:185-95.

Guttmacher LB, Cretella H. (1988) Electroconvulsivetherapy in one child and three adolescents. J Clin

Psychiatry 4:20-3.

Guttmacher LB, Greenland P. (1990) Effects ofelectroconvulsive therapy on the electrocardiogram ingeriatric patients with stable cardiovascular diseases.

Convuls Ther 6:5-12.

Guze BH, Baxter LR, Schwartz JM, Szuba MP, Listen EH.(1991) Electroconvulsive therapy and brain glucosemetabolism. Convuls Ther 7:20-7.

Guze BH, Listen EH, Baxter LR Jr, Richeimer SH, GoldME. (1989) Poor interrater reliablity of MECTA EEGrecordings of ECT seizure duration. J Clin Psychiatry

50:140-2.

Guze BH, Weinman B, Diamond RP. (1987) Use of ECT totreat bipolar depression in a mental retardate withcerebral palsy. Convuls Ther 3:60-4.

Guze SB. (1967) The occurrence of psychiatric illness insystemic lupus erythematosus. Am J Psychiatry

723:1562-70.

Halliday AM, Davison K, Browne MW, Kreeger LC. (1968)A comparison of the effects on depression and memoryof bilateral ECT and unilateral ECT to the dominant andnon-dominant hemispheres. Br J Psychiatry 774:997-1012.

Halsall PJ, Carr CM, Stewart KG. (1988) Propofol reducesseizure duration in patients having anaesthesia forelectroconvulsive therapy. Compr Psychiatry 61: 343-4.

Hamilton M. (1960) A rating scale for depression. JNeurol Neurosurg Psychiatry 23:56-62.

Hamilton M. (1982) Prediction of the response ofdepressions to ECT. In: R Abrams, WB Essman, eds.Electroconvulsive Therapy: Biological Foundations and

Clinical Applications . New York: Spectrum Publications,pp. 113-28.

Hamilton M, Stacker MJ, Spencer CM. (1979) Post-ECTcognitive defect and elevation of blood pressure. Br J


Hamilton M, White JM. (1960) Factors related to theoutcome of depression treated with ECT. J Men Sci

706:1031-41.

Handforth A. (1982) Postseizure inhibition of kindledseizures by electroconvulsive shock. Exp Neurol 78:483-91.

Hardman JB, Morse RM. (1972) Early electroconvulsivetreatment of a patient who had artificial aortic andmitral valves. Am J Psychiatry 728:895-7.

Harms E. (1956) The origin and early history ofelectrotherapy and electroshock. Am J Psychiatry

777:933-4.

Harper RG, Wiens AN. (1975) Electroconvulsive therapyand memory. J Nerv Ment Dis 767:245-54.

Hartelius H. (1952) Cerebral changes followingelectrically induced convulsions. An experimental studyof cats. Acta Psychiatry Neurol Scand Suppl 77.

Hartmann SJ, Saldivia A. (1990) ECT in an elderlypatient with skull defects and shrapnel. Convuls Ther

6:165-71.

Haskett RF, Zis AP, Albala AA. (1985) Hormone responseto repeated electroconvulsive therapy: effects ofnaloxone. Biol Psychiatry 20:623-33.

Hauser WA. (1983) Status epilepticus: frequency,etiology, and neurological sequelae. Adv Neurol 34:3-14.

P.277

Hay D. (1989) ECT in the medically ill elderly. Convuls

Ther 5:8-16.

Heath ES, Adams A, Wakeling PLG. (1964) Short coursesof ECT and simulated ECT in chronic schizophrenia. Br J


Heikman P, Salmelin R, Makela JP et al. (2001) Relationbetween frontal 3-7 Hz MEG activity and the efficacy ofECT in major depression. J ECT 17: 136-40.

Heilbrunn G, Weil J. (1942) Pathologic changes in thecentral nervous system in experimental electric shock.Arch Neural Psychiatry 47:918-30.

Hemphill RE. (1940) Studies in certain pathophysiologicaland psychological phenomena in convulsive therapy. JMent Sci 86:199.

Heninger GR, Charney DS, Sternberg DE. (1983) Lithiumcarbonate augmentation of antidepressant treatmentâ!”an effective prescription for treatment of refractorydepression. Arch Gen Psychiatry 40:1335-42.

Henry ME, Schmidt ME, Matochik JA et al. (2001) Theeffects of ECT on brain glucose: A pilot FDG PET study. JECT 77:33-40.

Hermann RC, Dorwart RA, Hoover CW, Brody J. (1995)Variation in ECT use in the United States. Am J


Hermesh H, Aizenberg D, Lapidot M, Munitz H. (1988)Risk of malignant hyperthermia among patients withneuroleptic malignant syndrome and their families. Am J


Herrington RM, Bruce A, Johnstone EC. (1974)Comparative trial of L-tryptophan and ECT in severedepressive illness. Lancet 2:731-4.

Heshe J, Roeder E. (1976) Electroconvulsive therapy inDenmark. Br J Psychiatry 728:241-5.

Heshe J, Roeder E, Theilgaard A. (1978) Unilateral andbilateral ECT. A psychiatric and psychological study oftherapeutic effect and side effects. Acta Psychiatr Scand

Suppl 275:1-180.

Hickey DR, O'Connor JP, Donati F. (1987) Comparison ofatracurium and succinylcholine for electroconvulsivetherapy in a patient with atypical plasma cholinesterase.Can J Anaesthesiology 34:280-3.

Hickie I, Scott E, Mitchell P et al. (1995) Subcorticalhyperintensities on magnetic resonance imaging: clinicalcorrelates and prognostic significance in patients withsevere depression. Biol Psychiatry 37:151-60.

Hicks EG. (1987) ECT modified by atracurium. Casereport. Convuls Ther 3:54-9.

Hill GE, Wong KC, Hodges MR. (1976) Potentiation ofsuccinylcholine neuromuscular blockade by lithiumcarbonate. Anesthesiology 44:439-42.

Hillard JR, Folger R. (1977) Patients' attitudes andattributions to electroconvulsive shock therapy. J Clin

Psychol 33:855-61.

Hobson RF. (1953) Prognostic factors in electricconvulsive therapy. J Neurol Neu rosurg Psychiatry

76:275-81.

P.278

Hodges JR, Jones M, Elithorn A, Bridges P. (1964) Effectof electroconvulsive therapy on plasma cortisol -levels.Nature 204:154-6.

Hoenig J, Chaulk R. (1977) Delirium associated withlithium therapy and electroconvulsive therapy. Can Med

Assoc J 7/6:837-8.

Hoffman G, Linkowski P, Kerkhofs M, Desmedt D,Mendlewicz J. (1985) Effects of ECT on sleep and CSFbiogenic amines in affective illness. Psychiatry Res

76:199-206.

Hofmann P, Gangadhar BN, Probst C, Koinig G, HatzingerR. (1994) TSH response to TRH and ECT. J Affect Disord

32:127-31.

Holcomb HH, Sternberg DE, Heninger GR. (1983) Effectsof electroconvulsive therapy on mood, parkinsonism, andtardive dyskinesia in a depressed patient: ECT anddopamine systems. Biol Psychiatry 78:865-73.

Hollender MH, Sleekier PP. (1972) Multiple sclerosis andschizophrenia: A case report. Psychiatr Med 3:251-7.

Holmberg G. (1953a) The influence of oxygenadministration on electrically induced convulsions inman. Acta Psychiatr Neurol Scand 28:365-86.

Holmberg G. (1953b) The factor of hypoxemia inelectroshock therapy. Am J Psychiatry 7/0:115-8.

Holtzman JL, Finley D, Johnson B, Berry DA, Sirgo MA.(1986) The effects of single-dose atenolol, labetalol, andpropranolol on cardiac and vascular function. Clin

Pharmacol Ther 40:268-73.

Hood DD, Mecca RS. (1983) Failure to initiateelectroconvulsive seizures in a patient pretreated withlidocaine. Anesthesiology 58:379-81.

Hordern A, Holt HP, Burt CG, Gordon WF. (1963)Amitriptyline in depressive cases. Br J Psychiatry

709:815-25.

Howie MB, Black HA, Zvara D, McSweeney TD, Martin DJ,Coffman JA. (1990) Esmolol reduces autonomichypersensitivity and length of seizures induced byelectroconvulsive therapy. Anesth Analg 77:384-8.

Howie MB, Hiestand DC, Zvara DA, Kim PY, McSweeneyTD, Coffman JA. (1992) Defining the dose range foresmolol used in electroconvulsive therapy hemodynamicattenuation. Anesth Analg 75:805-10.

Hoyle NR, Pratt RT, Thomas DG. (1984) Effect ofelectroconvulsive therapy on serum myelin basic proteinimmunoreactivity. Br Med J 288:1110-1.

Hrdlicka M, Moran M, Vachutka J et al. (1996) EEG inelectroconvulsive therapy: Is more intensive paroxysmalactivity associated with a higher therapeutic response?Neuropsychobiology 33:138-41.

Hsiao JK, Evans DL. (1984) ECT in a depressed patientafter craniotomy. Am J Psychiatry 141:442-4.

Huang KC, Lucas LF, Tsueda K, Thomas M, Lippmann SB.(1989) Age-related changes in cardiovascular functionassociated with electroconvulsive therapy. Convuls Ther

5:17-25.

Hughes J, Barraclough BM, Reeve W. (1981) Are patientsshocked by ECT? J Roy Soc Med 74:283-5.

Hughes JR. (1986) ECT during and after the neurolepticmalignant syndrome: case report. J Clin Psychiatry

47:42-3.

Hunt SA, Kaplan E. (1998) ECT in the presence ofcerebral aneurysm. J ECT 14: 123-4.

Huston PE, Strother CH. (1948) The effect of ECT onmental efficiency. Am J Psychiatry 104:701.

Husum B, Vester-Andersen T, Buchmann G, Bolwig TG.(1983) Electroconvulsive therapy and intracranialaneurysm. Prevention of blood pressure elevation in anormotensive patient by hydralazine and propranolol.Anaesthesia 38: 1205-7.

Hyrman V, Palmer LH, Cernik J, Jetelina J. (1985) CT:The search for the perfect stimulus. Biol Psychiatry

20:634-5.

Hyrman, V. (1999) Pulse width and frequency in ECT[letter). J ECT 75:285-7.

Impastato DJ. (1966) Tendon reflexes as a guide to thesafe use of succinylcholine in medicine. Can Psychiatr

Assoc J 77:67-77.

Impastato DJ, Almansi R. (1942) The electron! in thetreatment of mental disease. J Nerv Ment Dis 96:395-409.

Impastato DJ, Berg S, Pacella BL. (1953) Electroshocktherapy: focal spread technique. A new form oftreatment of psychiatric illness. Confinia Neuralogica

13:266-10.

Impastato DJ, Karliner W. (1966) Control of memory

P.279

impairment in EST by unilateral stimulation of the non-dominant hemisphere. Diseases of the Nervous System

27:182-8.

Impastato DJ, Pacella BL. (1952) Electrically producedunilateral convulsions. Dis eases of the Nervous System

73:368-9.

Inturrisi CE, Alexopoulos G, Lipman R, Foley K, RossierJ. (1982) (3-Endorphin immunoreactivity in the plasmaof psychiatric patients receiving electroconvulsivetreatment. Ann N Y Acad Sci 395:413-23.

Irving AD, Dray son AM. (1984) Bladder rupture duringECT. Br J Psychiatry 144: 670.

Isaac RJ. (1990) FDA's shocking treatment of a valuabledevice. The Wall Street Journal , Dec. 5.

Isenberg KE, Dinwiddie SH, Heath AC, et al. (1996)Effect of stimulus parameters on seizure threshold andduration, [abstract]. Convuls Ther 72:68.

Ives JO, Weaver LA, Williams R. (1976) Portableelectromyograph monitoring of unilateral ECT. Am J


Jacobs BL, Praag H, Gage FH. (2000) Adult brainneurogenesis and psychiatry: A novel theory ofdepression. Mol Psychiatry 5:262-9.

Jaffe R, Brubaker G, Dubin WR, Roemer R. (1990b)Caffeine-associated cardiac dysrhythmia during ECT:report of three cases. Convuls Ther 6:308-13.

Jaffe R, Dubin W, Shoyer B, Roemer R, Sharon D,Lipschutz L. (1990a) Outpatient electroconvulsive

therapy: efficacy and safety. Convuls Ther 6:231-8.

Janakiramaiah N, Channabasavanna SM, Murthy NS.(1982) ECT/chlorpromazine combination versuschlorpromazine alone in acutely schizophrenic patients.Acta Psychiatr Scand 66:464-70.

Janis IL. (1950a) Psychologic effects of electricconvulsive treatments (posttreatment amnesias). J Nerv

Ment Dis 777:359-82.

Janis IL. (1950b) Psychologic effects of electricconvulsive treatments (changes in word associationreactions). J Nerv Ment Dis 777:383-97.

Jauhar P, Weller M, Hirsch SR. (1979) Electroconvulsivetherapy for patient with cardiac pacemaker. Br Med J

7:901.

Jeffries BF, Kishore PRS, Singh KS, Ghatak NR, KrempaJ. (1980) Postoperative computed tomographic changesin the brain. Radiology 735:751-3.

Jensen AV, Becker RF, Windle WE (1948) Changes inbrain structure and memory after intermittent exposureto simulated altitude of 30,000 feet. Arch Neural


Jephcott G, Kerry RJ. (1974) Lithium: An anesthetic risk.Compr Psychiatry 46: 389-90.

Jessee SS, Anderson GF. (1983) ECT in the neurolepticmalignant syndrome: case report. / Clin Psychiatry

44:186-8.

Johansson F, von Knorring L. (1987) Changes in serumprolactin after electroconvulsive and epileptic seizures.

P.280

Ear Arch Psychiatry Neurol Sci 236:312-8.

Johnstone EC, Deakin JF, Lawler P, Frith CD, Stevens M,McPherson K, et al. (1985) The Northwick Parkelectroconvulsive therapy trial. Lancet 2:1317-20.

Jones RM, Knight PR. (1981) Cardiovascular andhormonal responses to electroconvulsive therapy.Modification of an exaggerated response in anhypertensive patient by (3-receptor blockade.Anaesthesia 36:795-9.

Kahn RL, Fink M. (1960) Prognostic value of Rorschachcriteria in clinical reponse to convulsive therapy. JNeuropsychiatry 7:242-5.

Kahn RL, Pollack M, Fink M. (1959) Sociopsychologicaspects of psychiatric treatment in a voluntary mentalhospital: duration of hospitalization, discharge ratingsand diagnosis. Arch Gen Psychiatry 7:565-74.

Kalayam B, Alexopoulos GS. (1989) Nifedipine in thetreatment of blood pressure rise after ECT. Convuls Ther

5:110-3.

Kalayam B, Steinhart MJ. (1981) A survey of attitudeson the use of electroconvulsive therapy. Hasp

Community Psychiatry 32:185-8.

Kales H, Raz J, Tandon R et al. (1997) Relationship ofseizure duration to antidepressant efficacy inelectroconvulsive therapy. Psychol Med 27:1373-80.

Kalinowsky L. (1945) Organic psychotic syndromesoccurring during electric convulsive therapy. Arch Neural


Kalinowsky LB. (1939) Electric-convulsion therapy inschizophrenia. Lancet 2: 1232-3.

Kalinowsky LB. (1956a) The danger of various types ofmedication during electric convulsive therapy. Am J


Kalinowsky LB. (1956b) Additional remarks on thedanger of premedication in electric convulsive therapy.Am J Psychiatry 773:79-80.

Kalinowsky LB. (1982) The history of electroconvulsivetherapy. In: R Abrams, WB Essman, eds.Electroconvulsive Therapy: Biological Foundations and

Clini cal Applications . New York: Spectrum Publications,p. 16.

Kalinowsky LB. (1986) History of convulsive therapy. Ann

N Y Acad Sci 462:1-4.

Kalinowsky LB, Hippius H, Klein HE. (1982) Biological

Treatments in Psychiatry . New York: Grune and Stratton.

Kalinowsky LB, Kennedy F. (1943) Observations inelectric shock therapy applied to problems of epilepsy. JNerv Ment Dis 98:56-61.

Kane FJ. (1963) Transient neurological symptomsaccompanying ECT. Am J Psychiatry 779:786-7.

Kandel ER, Schwartz JH, Jessel TM. (1991) Principles ofNeural Science (3d. ed.), New York, Elsevier, pp. 109-11.

Kapur S, Mann JJ. (1993) Antidepressant action and theneurobiologic effects of ECT: human studies. In: CECoffey, ed. The Clinical Science of Electrocon vulsive

Therapy . Washington DC: American Psychiatric Press, pp.235-50.

Karliner W. (1978) ECT for patients with CNS disease.Psycho somatics 79:781-3.

Katona CL, Aldridge CR, Roth M, Hyde J. (1987) Thedexamethasone suppression test and prediction ofoutcome in patients receiving ECT. Br J Psychiatry

750:315-8.

Katona CLE, Aldridge CR. (1984) Prediction of ECTresponse. Neuropharmacology 23:281-3.

Kaufman KR. (1994) Asystole with electroconvulsivetherapy. J Intern Med 235: 275-7.

Kaufman KR, Finstead BA, Kaufman ER. (1986) Statusepilepticus following electroconvulsive therapy. Mt Sinai

J Med 53:119-22.

Kearns A. (1987) Cotard's syndrome in a mentallyhandicapped man. Br J Psychiatry 750:112-4.

Kellner CH. (1997) Left unilateral ECT; still a viableoption? [editorial]. Convuls Ther 73:65-67.

Kellner CH. (2000) High-dose right unilateral ECT[editorial!. J ECT 76:209-210.

Kellner CH. (2001) Towards the modal ECT treatment[editorial]. J ECT 77:1-2.

Kellner CH, Bachman DL. (1992) Hallucination afterintravenous caffeine. Am J Psychiatry 149:422.

P.281

Kellner CH, Pritchett JT, Beale MD, et al. (1997)Handbook of ECT . Washington, DC: Amer Psych Press.

Kelway B, Simpson KH, Smith RJ, Halsall P. (1986)Effects of atropine and glycopyrrolate on cognitivefunction following anesthesia and electroconvulsivetherapy (ECT). Int J Clin Pharmacol 7:296-302.

Kendell B, Pratt RTC. (1983) Brain damage and ECT. Br J


Kendell RE, Cooper JE, Gourlay AJ, Copeland JR. (1971)Diagnostic criteria of American and British psychiatrists.Arch Gen Psychiatry 25:123-30.

Kerr RA, McGrath JJ, O'Kearney RT, Price J. (1982) ECT:misconceptions and attitudes. Aust N Z J Psychiatry

76:43-9.

Kety S, Javoy F, Thierry AM, Julou L, Glowinski J. (1967)A sustained effect of electroconvulsive shock on theturnover of norepinephrine in the central nervous systemof the rat. Proc Natl Acad Sci 58:1249-54.

Kety SS, Woodford RB, Harmel MH, Freyhand FA, AppelKE, Schmidt CF. (1948) Cerebral blood flow andmetabolism in schizophrenia: The effects of barbituratesemi-narcosis, insulin coma and electroshock. Am J


Khoury GF, Benedetti C. (1989) T -wave changesassociated with electroconvulsive therapy. Anesth Analg

69:677-9.

Kimbrell TA, Little JT, Dunn RT. (1999) Frequencydependence of antidepressant response to left prefrontalrepetitive transcranial magnetic stimulation (rTMS) as a

function of baseline cerebral glucose metabolism. Biol


King BH, Listen EH. (1990) Proposals for the mechanismof action of convulsive therapy: A synthesis. Biol


King PD. (1960) Chlorpromazine and electroconvulsivetherapy in the treatment of newly hospitalizedschizophrenics. J Clin Expl Psychopathol 27:101-5.

Kirkegaard C. (1987) Effects of ECT on the TRHstimulation test. Psychoneuroen docrinology 72:491-3.

Kirkegaard C, Carroll BJ. (1980) Dissociation of TSHadrenocortical disturbances in endogenous depression.Psychiatry Res 3:253-64.

Kirkegaard C, Norlem N, Lauridsen UB, Bjorum N. (1977)Prognostic value of thyrotropin-releasing hormonestimulation test in endogenous depression. Acta

Psychiatr Scand 52:170-7.

Kitamura T, Page AJ. (1984) Electrocardiographicchanges following electroconvulsive therapy. Eur Arch

Psychiatry Neural Sci 254:147-8.

Klein E, Kreinin I, Christakov A et al. (1999) Therapeuticefficacy of right prefrontal slow repetitive transcranialmagnetic stimulation in major depression: A double-blindcontrolled study. Arch Gen Psych 56:315-20.

Klimes I, Vigas M, Jurcovicova J, Wiedermann V. (1978)Serum prolactin after electroconvulsive therapy.Endokrinologie 72:371-3.

Kolb LC, Vogel VH. (1942) The use of shock therapy in305 mental hospitals. Am J Psychiatry 99:90-100.

P.282

Kolbeinsson H, Arnaldsson OS, Petursson H, Skulason S.(1986) Computed tomographic scans in ECT patients.Acta Psychiatr Scand 73:28-32.

Kolbeinsson H, Petursson H. (1988)Electroencephalographic correlates of electroconvulsivetherapy. Acta Psychiatr Scand 75:162-8.

Korin H, Fink M, Kwalwasser S. (1956) Relation ofchanges in memory and learning to improvement inelectroshock. Confinia Neurologica 76:88-96.

Kovac AL, Goto H, Arakawa K, Pardo MP. (1990) Esmololbolus and infusion attenuates increases in blood pressureand heart rate during electroconvulsive therapy. Can J

Anesth 37:58-62.

Kramer BA. (1985) Use of ECT in California, 1977-1983.Am J Psychiatry 142: 1190-2.

Kramer BA. (1987) Maintenance ECT: A survey ofpractice (1986). Convuls Ther 3:260-8.

Kramer BA. (1990) Maintenance electroconvulsivetherapy in clinical practice. Con vuls Ther 6:279-86.

Kramer BA. (1999a) Use of ECT in California, Revisited:1984-1994. J ECT 15: 245-51.

Kramer BA. (1999b) A naturalistic view of maintenanceECT at a university setting. J ECT 75:262-9.

Kramer BA, Afrasiabi A. (1991) A typical cholinesteraseand prolonged apnea during electroconvulsive therapy.Convuls Ther 7:129-32.

Kramer BA, Allen RE, Friedman B. (1986) Atropine andglycopyrrolate as ECT preanesthesia. J Clin Psychiatry

47:199-200.

Kramer BA, Pollock VE, Schneider LS, Gray GE. (1989)Interrater reliability of MECTA SR-1 seizure duration.Biol Psychiatry 25:642-4.

Kramp P, Bolwig TG. (1981) Electroconvulsive therapy inacute delirious states. Compr Psychiatry 22:368-71.

Kriss A, Blumhardt LD, Halliday AM, Pratt RT. (1978)Neurological asymmetries immediately after unilateralECT. J Neural Neurosurg Psychiatry 41: 1135-44.

Kronfol Z, Hamsher KD, Digre K, Waziri R. (1978)Depression and hemispheric functions: changesassociated with unilateral ECT. Br J Psychiatry 132: 560-7.

Krueger RB, Fama JM, Devanand DP, Prudic J, SackeimHA. (1993) Does ECT permanently alter seizurethreshold? Biol Psychiatry 33:272-6.

Krystal AD. (1998) The clinical utility of ictal EEG seizureadequacy models. Psych Ann 25:30-5.

Krystal AD, Coffey CE, Weiner RD et al. (1998) Changesin seizure threshold over the course of electroconvulsivetherapy affect therapeutic response and are detectableby ictal EEG ratings. J Neuropsychiatry Clin Neurosci 10:178-86.

Krystal AD, Dean MD, Weiner RD et al. (2000) ECTstimulus intensity: Are present ECT devices too limited?Am J Psychiatry 757:963-7.

P.283

Krystal AD, Weiner RD. (1994) ECT seizure therapeuticadequacy. Convuls Ther 70:153-64.

Krystal AD, Weiner RD. (1999) EEG correlates of theresponse to ECT: A possible antidepressant role of brain-derived neurotrophic factor. J ECT 75:27-38.

Krystal AD, Weiner RD, Coffey CE. (1995) The ictal EEGas a marker of adequate stimulus intensity withunilateral ECT. J Neuropsychiatry 7:295-303.

Krystal AD, Weiner RD, Coffey CE, et al. (1992) EEGevidence of more â!œintenseâ! ! seizure activity withbilateral ECT. Biol Psychiatry 37:617-21.

Krystal AD, Weiner RD, Gasseert D, et al. (1996) Therelative ability of three ictal EEG frequency bands todifferentiate ECT seizures on the basis of electrodeplacement, stimulus intensity, and therapeutic response.Convuls Ther 12: 13-24.

Krystal AD, Weiner RD, Lindahl V et al. (2000) Thedevelopment and retrospective testing of anelectroencephalographic seizure quality -based stimulusdosing paradigm with ECT. J ECT 76:338-49.

Krystal AD, Weiner RD, McCall WV, et al. (1993) Theeffects of ECT stimulus dose and electrode placement onthe ictal electroencephalogram: An intra-individualcross-over study. Biol Psychiatry 34:759-61.

Kupfer D. (1986) The sleep EEG in diagnosis andtreatment of depression. In: AJ Rush, KZ Altshuler, eds.Depression: Basic Mechanisms, Diagnosis, and

Treatment. Guilford Press: New York, pp. 102-25.

Kurokawa Y, Ueno T, Obara T, Gotohda T, Fukatsu R,

Yamashita I. (1989) Hyperkinetic mutism within thescope of consciousness disorder in a case of systemiclupus erythematosus. Jpn J Psychiatry Neural 43:89-96.

Kwentus JA, Hart RP, Calabrese V, Hekmati A. (1986)Mania as a symptom of multiple sclerosis.Psychosomatics 27:729-31.

Kwentus JA, Schulz SC, Hart RP. (1984) Tardivedystonia, catatonia, and electroconvulsive therapy. JNerv Ment Dis 772:171-3.

Laird DM. (1955) Convulsive therapy in psychosesaccompanying pregnancy. N Engl J Med 252:934-6.

Lalla FR, Milroy T. (1996) The current status of seizureduration in the practice of electroconvulsive therapy.Can J Psychiatry 41:299-304.

Lambourn J, Gill D. (1978) A controlled comparison ofsimulated and real ECT. Br J Psychiatry 133:54-9.

Lancaster NP, Steinert RR, Frost I. (1958) Unilateralelectro-convulsive therapy. / Ment Sci 704:221-7.

Lane R, Novelly R, Cornell C, Zeitlin S, Schwartz G.(1988) Asymmetric hemispheric control of heart rate[abstract]. Psychophysiology 25:464.

Lane RD, Zeitlin SB, Abrams R, Swartz CM. (1989)Differential effects of right unilateral and bilateral ECTon heart rate. Am J Psychiatry 746:1041-3.

Langer G, Karazman R, Neumark J, Saletu B, SchonbeckG, Grunberger J, Dittrich R, Petriceck W, Hoffman P,Linzmayer L, Anderer P, Steinberger K. (1995) Isofluranenarcotherapy in depressive patients refractory to

conventional antidepressant drug treatment.Neuropsychobiology 37:182-94.

Langer G, Neumark J, Koining G, Graf M, Schonbeck G.(1985) Rapid psychotherapeutic effects of anesthesiawith isoflurane (ES narcotherapy) in treatment -refractorydepressed patients. Neuropsychobiology 14:118-20.

Langsley DG, Enterline JD, Hickerson GX. (1959)Comparison of chlorpromazine and EST in treatment ofacute schizophrenic and manic reactions. Arch Psychiatry

57:384-91.

Larson G, Swartz C, Abrams R. (1984) Duration of ECT-induced tachycardia as a measure of seizure length. Am


Larson G, Swartz CM. (1986) Differences between firstand second electroconvulsive treatments given in thesame session. Convuls Ther 2:191-6.

Latey RH, Fahy TJ. (1985) Electroconvulsive therapy inthe Republic of Ireland 1982: A summary of findings. Br


Lauritzen L, Odgaard K, Clemmesen L et al. (1996)Relapse prevention by means of paroxetine in ECT-treated patients with major depression: A comparisonwith imipramine and placebo in medium-termcontinuation therapy. Acta PsychiatrScand 94:241-5l.

Lawson JS, Inglis J, Delva NJ, Rodenburg M, Waldron JJ,Letemendia FJ. (1990) Electrode placement in ECT:cognitive effects. Psychol Med 20:335-44.

Lee JT, Erbguth PH, Stevens WC, Sack RL. (1985)Modification of electroconvulsive therapy inducedhypertension with nitroglycerine ointment. Anesthesi

P.284

ology 62:793-6.

Leechuy I, Abrams R. (1987) Postictal delirium (andrecovery from melancholia) after left -unilateral ECT.Convuls Ther 3:65-8.

Leechuy I, Abrams R, Kohlhaas J. (1988) ECT-inducedpostictal delirium and electrode placement. Am J


Lehmann L, Liddell J. (1969) Human cholinesterase(pseudocholinesterase): genetic variants and theirrecognition. Compr Psychiatry 41:235-44.

Lerer B. (1987) Neurochemical and other neurobiologicalconsequences of ECT: implications for the pathogenesisand treatment of affective disorders. In: HY Meltzer, ed.Psyche/pharmacology: The Third Generation of Progress .Raven Press: New York.

Lerer B (1999) The neurobiology of ECT: The roadahead. J ECT 75:1-4

Lerer B, Belmaker RH. (1982) Receptors and themechanism of action of ECT. Biol Psychiatry 77:497-511.

Lerer B, Shapira B, Calev A, Tubi N, Drexler H, KindlerS, Lidsky D, Schwartz JE. (1995) Antidepressant andcognitive effects of twiceversus three-times-weekly ECT.Am J Psychiatry 752:564-70.

Letemendia JF, Delva NJ, Rodenberg M, Lawson JS, InglisJ, Waldron JJ, Lywood DW. (1993) Therapeutic advantageof bifrontal electrode placement in ECT. Psychol Med

2J:349-60.

Levy R. (1968) The clinical evaluation of unilateral

electroconvulsive therapy. Br J Psychiatry 774:459-63.

Levy SD. (1988) â!œCuffâ! ! monitoring, osteoporosis,and fracture. Convuls Ther 4: 248-9.

Levy SD, Levy SB. (1987) Electroconvulsive therapy intwo former neurosurgical patients: skull prosthesis andventricular shunt. Convuls Ther 3:46-8.

Lewis WH, Jr, Richardson DJ, Gahagan LH. (1955)Cardiovascular disturbances and their management inmodified electrotherapy for psychiatric illness. N Engl J

Med 252:1016-20.

Liang RA, Lam RW, Aneill RJ. (1988) ECT in thetreatment of mixed depression and dementia. Br J


Liberson WT. (1944) New possibilities in electricconvulsive therapy: brief stimulus technique. Digest of

Neurology and Psychiatry 72:368.

Liberson WT. (1945) Time factors in electric convulsivetherapy. Yale J Biol Med icine 77:571-8.

Liberson WT. (1948) Brief stimulus therapy. Physiologicaland clinical observations. Am J Psychiatry 705:28-9.

Liberson WT. (1953) Current evaluation of electricconvulsive therapyâ!”correlation of the parameters ofelectric current with physiologic and psychologicchanges. Res Publ Assoc Res Nerv Ment Dis 37:199-231.

Liberson WT, Kaplan JA, Sherer IW, Trehub A. (1956)Correlations of EEG and psychological findings duringintensive brief stimulus therapy. Confinia Neu rologica

16:116-25.

P.285

Liberson WT, Wilcox P. (1945) Electric convulsivetherapy: Comparison of â!œbrief stimulus techniqueâ! !with Friedman-Wilcox-Reiter technique. Digest of Neu

rology and Psychiatry 8:292-302.

Liberzon I, DeQuardo JR, Sidell G, Mazzara C, Tandon R.(1992) Post-ECT dyskinesia. Convuls Ther 7:40-4.

Lidbeck WL. (1944) Pathologic changes in the brain afterelectric shock: An experimental study on dogs. JNeuropathol Exp Neural 3:81-6.

Liebowitz NR, El-Mallakh RS. (1993) Cardiac arrestduring ECT: A cholinergic phenomenon? J din Psychiatry

54:219-80.

Lindefors N, Brodin E, Metsis M. (1995) Spatiotemporalselective effects on brain-derived neurotrophic factorand trkB messenger RNA in rat hippocampus byelectroconvulsive shock. Neuroscience 65:661-70.

Lingley JR, Robbins LL. (1947) Fractures followingelectroshock therapy. Radiology 45:124-8.

Linkowski P, Mendlewicz J, Kerkhofs M, Leclercq R,Golstein J, Brasseur M, et al. (1987) 24-hour profiles ofadrenocorticotropin, cortisol, and growth hormone inmajor depressive illness: effect of antidepressanttreatment. J Clin Endo crinol Metab 65:141-52.

Linnoila M, Karoum F, Potter WZ. (1983) Effects ofantidepressant treatments on dopamine turnover indepressed patients. Arch Gen Psychiatry 40:1015-7.

Linnoila M, Miller TL, Bartko J, Potter WZ. (1984) Fiveantidepressant treatments in depressed patients. Effects

on urinary serotonin and 5-hydroxyindoleacetic acidoutput. Arch Gen Psychiatry 41:688-92.

Lipkin KM, Dyrud J, Meyer GG. (1970) The many faces ofmania. Arch Gen Psychiatry 22:262-7.

Lipman RS, Backup C, Bobrin Y, Delaplane JM, Doeff J,Gittleman S, Joseph R, Kanefield, M. (1986a)Dexamethasone suppression test as a predictor ofresponse to electroconvulsive therapy. I. Inpatienttreatment. Convuls Ther 2: 151-60.

Lipman RS, Uffner W, Schwalb N, Ravetz R, Lief B, LevyS. (1986b) Dexamethasone suppression test as apredictor of response to electroconvulsive therapy. II.Six-month follow-up. Convuls Ther 2:161-7.

Lippmann S, Manshadi M, Wehry M, Byrd R, Past W,Keller W, Schuster J, Elams, Meyer D, O'Daniel R. (1985)1,250 electroconvulsive treatments without evidence ofbrain injury. Br J Psychiatry 147:203-4.

Lisanby SH, Belmaker RH. (2000) Animal models of themechanisms of action of repetitive transcranial magneticstimulation (RTMS): comparisons with electroconvulsiveshock (ECS). Depress Anxiety 72:178-87.

Lisanby SH, Devanand DP, Nobler MS, Prudic J, Mullen L,Sackeim HA. (1996) Exceptionally high seizurethresholds: ECT device limitations. Convuls Ther 72:156-64.

Lisanby SH, Devanand DP, Prudic J et al. (1998)Prolactin response to electroconvulsive therapy: effectsof electrode placement and stimulus dosage. Biol


Lisanby SH, Devanand DP, Prudic J, Pierson D, Nobler M,

P.286

Fitzsimons L, Sackeim HA. (in press) Prolactin responseto ECT: effects of electrode placement and stimulusdosage. Biol Psychiatry .

Lisanby SH, Gutman D, Luber B et al. (2001b) ShamTMS: intracerebral measurement of the induced electricalfield and the induction of motor -evoked potentials. Biol


Lisanby SH, Luber B, Barroilhet L. et al. (2001d)Magnetic Seizure Therapy (MST)-Acute Cognitive Effectsof MST Compared with ECT. Presented at the annualmeeting of the Association for Convulsive Therapy, NewOrleans, May 6, 2001.

Lisanby SH, Luber B, Sackeim HA et al. (2001a)Deliberate seizure induction with repetitive transcranialmagnetic stimulation in nonhuman primates. Arch Gen


Lisanby SH, Maddox JH, Prudic J et al. (2000) Theeffects of electroconvulsive therapy on memory forautobiographical and public events. Arch Gen Psychiatry

57:581-90.

Lisanby SH, Schlaepfer TE, Fisch, H-U et al. (2001c)Magnetic seizure therapy of major depression [letter].Arch Gen Psychiatry 55:303-5.

Liskow BI. (1985) Relationship between neurolepticmalignant syndrome and malignant hyperthermia[letter]. Am J Psychiatry 142:390.

Listen EH, Guze BH, Baxter LR, Jr, Richeimer SH, GoldME. (1988) Motor versus EEG seizure duration in ECT.Biol Psychiatry 24:94-6.

Listen EH, Salk JD. (1990) Hemodynamic responses toECT after bilateral adrenalectomy. Convuls Ther 6:160-4.

Liston EH, Sones DE. (1990) Postictal hyperactivedelirium in ECT: management with midazolam. Convuls

Ther 6:192-5.

Liu JC. (1949) Nerve cell changes resulting fromstarvation. Anat Rec 103:68.

London SW, Glass DD. (1985) Prevention ofelectroconvulsive therapy-induced dysrhythmias withatropine and propranolol. Anesthesiology 62:819-22.

Loo C, Mitchell P, Sachdev P et al. (1999) Double-blindcontrolled investigation of transcranial magneticstimulation for the treatment of resistant majordepression. Am J Psychiatry 756:946-8.

Loo CK, Taylor JL, Gandevia SC et al. (2000)Transcranial magnetic stimulation (TMS) in controlledtreatment studies: Are some â!œshamâ! ! forms active?Biol Psychiatry 47:325-31.

Loo H, Galinowski A, Boccara I, Richard A. (1988)Interet de la sismotherapie d'entretien dans lesdepressions recurrentes: A propos de 4 observations.Encephale 74:39-41.

Loo H, Galinowski A, De Carvalho W, Bourdel MC, PoirierME (1991) Use of maintenance ECT for elderly depressedpatients. Am J Psychiatry 748:810.

Loo H, Kuche H, Benkelfat C. (1985) Electroconvulsivetherapy during anticoagulant therapy. Convuls Ther

1:258-62.

Lotstra F, Linkowski P, Mendlewicz J. (1983) General

P.287

anesthesia after neuroleptic malignant syndrome. Biol


Lovett-Doust JW, Raschka LB. (1975) Enduring effects ofmodified ECT on the cerebral circulation in man. Acomputerized study by cerebral impedanceplethysmography. Psychiatr Clin 8:293-303.

Lowinger P, Huston PE. (1953) Electric shock inpsychosis with cerebral spastic paralysis. Diseases of the

Nervous System 14:2-4.

Lown B. (1979) Sudden cardiac death: The majorchallenge confronting contemporary cardiology. Am J

Cardiol 43:313.

Luber B, Nobler MS, Moeller JR. (2000) Quantitative EEGduring seizures induced by electroconvulsive therapy:relations to treatment modality and clinical features. II.Topographic analyses. J ECT 76:229-43.

Lunn V, Trolle E. (1949) On the initial impairment ofconsciousness following electric convulsive therapy. Acta

Psychiatr Scand 24:33-58.

Lykouras L, Markianos M, Augoustides A, Papakostas Y,Stefanis C. (1993) Evaluation of TSH and prolactinresponses to TRH as predictors of the therapeutic effectof ECT in depression. Eur Neuropsychopharmacol 3:81-3.

Lykouras L, Markianos M, Hatzimanolis J, Malliaras D,Stefanis C. (1990) Biogenic amine metabolites duringelectroconvulsive therapy of melancholic patients.Convuls Ther 6:266-72.

Mac DS, Pardo MR (1983) Systemic lupus erythematosusand catatonia: A case report. J Clin Psychiatry 44:155-6.

Malek-Ahmadi P, Beceiro JB, McNeil BW, Weddige RL.(1990) Electroconvulsive therapy and chronic subduralhematoma. Convuls Ther 6:38-41.

Malek-Ahmadi P, Sedler RR. (1989) Electroconvulsivetherapy and asymptomatic meningioma. Convuls Ther

5:168-70.

Malek-Ahmadi P, Weddige RL. (1988) Tardive dyskinesiaand electroconvulsive therapy. Convuls Ther 4:328-31.

Maletzky BM. (1978) Seizure duration and clinical effectin electroconvulsive therapy. Compr Psychiatry 79:541-50.

Maletzky BM. (1981) Multiple-Monitored Electroconvulsive

Therapy . Boca Raton, FL: CRC Press.

Maletzky BM. (1986) Conventional and multiple-monitored electroconvulsive therapy: A comparison inmajor depressive episodes. J Nerv Ment Dis 174: 257-64.

Malmivuo, Plonsey (1995) Bioelectromagnetism . NewYork: Oxford University Press, p. 64.

Malsch E, Gratz I, Mani S, Backup C, Levy S, Allen E.(1994) Efficacy of electroconvulsive therapy afterpropofol and methohexital anesthesia. Convuls Ther

70:212-19.

Maltbie AA, Wingfield MS, Volow MR, Weiner RD, SullivanJL, Cavenar JO, Jr. (1980) Electroconvulsive therapy inthe presence of brain tumor. Case reports and anevaluation of risk. J Nerv Ment Dis 768:400-5.

Mandel M. (1980) Review of Breggin P. Electroshock: Its

Brain Disabling Effects. N Engl J Med 303:402.

Mandel MR, Miller AL, Baldessarini RJ. (1980)Intoxication associated with lithium and ECT. Am J


Mander AJ, Whitfield A, Kean D, Smith MA, Douglas RH,Kendell RE. (1987) Cerebral and brain stem changesafter ECT revealed by nuclear magnetic resonanceimaging. Br J Psychiatry 757:69-71.

Maneksha FR. (1991) Hypertension and tachycardiaduring electroconvulsive therapy: to treat or not totreat? Convuls Ther 7:283-5.

Manly DT, Swartz CM. (1994) Asymmetric bilateral rightfrontotemporal left frontal stimulus electrode placement:comparisons with bifrontotemporal and unilateralplacements. Convuls Ther 70:267-70.

Mann JJ. (1998) Neurobiological correlates of theantidepressant action of electroconvulsive therapy. J ECT

74:172-80.

Mann JJ, Mahler JC, Wilner PJ, Halper JP, Brown RP,Johnson KS, et al. (1990) Normalization of bluntedlymphocyte p-adrenergic responsivity in melancholicinpatients by a course of electroconvulsive therapy. Arch


Mann SC, Caroff SN, Bleier HR, Welz WK, Kling MA,Hayashida M. (1986) Lethal catatonia. Am J Psychiatry

143:1374-81.

Manning EL, Hollander WM. (1954) Glaucoma andelectroshock therapy. Am J Ophthalmol 37:857-9.

P.288

Mansheim P. (1983) ECT in the treatment of a depressedadolescent with meningomyelocele, hydrocephalus, andseizures. J Clin Psychiatry 44:385-6.

Marco LA, Randels PM. (1979) Succinylcholine druginteractions during electroconvulsive therapy. Biol


Markianos M, Papakostas Y, Stefanis C. (1987) Thepatterns of prolactin release by ECT and TRH compared.Life Sci 47:1273-6.

Martensson B, Bartfai A, Hallen B, Hellstrom C, Junthe T,Olander M. (1994) A comparison of propofol andmethohexital as anesthetic agents for ECT: effects onseizure duration, therapeutic outcome, and memory. Biol


Martin BA, Kramer PM. (1982) Clinical significance of theinteraction between lithium and a neuromuscularblocker. Am J Psychiatry 739:1326-8.

Martin RD, Flegenheimer WV. (1971) Psychiatric aspectsof the management of the myasthenic patients. Mt Sinai

J Med 3S:594-601.

Martin W, Ford HF, McDonald EC et al. (1965) Clinicalevaluation of unilateral EST. Am J Psychiatry 727:1087-90.

Mathe AA, Bergman P, Aperia B, Wetterberg L. (1987)Electroconvulsive therapy and plasma prostaglandin E2metabolite in major depressive disorder. Prog

Neuropsychopharmacol Biol Psychiatry 77:701-7.

Mattes JA, Pettinati HM, Stephens S, Robin SE, WillisKW. (1990) A placebo-controlled evaluation of

vasopressin for ECT-induced memory impairment. Biol


Matthew JR, Constan E. (1964) Complications followingECT over a three-year period in a state institution. Am J


Maxwell RD. (1968) Electrical factors in electroconvulsivetherapy. Acta Psychiatr Scand 44:436-48.

Mayur PM, Gangadhar BN, Janakiramaiah N et al. (1999)Motor seizure monitoring during electroconvulsivetherapy. Br J Psychiatry 774:270-2.

McAllister DA, Perri MG, Jordan MC, Rauscher FP, SattinA. (1987) Effects of ECT given two vs three timesweekly. Psychiatry Res 27:63-9.

McAllister TW, Price TR. (1982) Severe depressivepseudodementia with and without dementia. Am J


McAndrew J, Berkey B, Matthews C. (1967) The effectsof dominant and nondominant unilateral ECT ascompared to bilateral ECT. Am J Psychiatry 124: 483-90.

McAndrew J, Hauser G. (1967) Preventilation of oxygenin electroconvulsive treatment: A suggested modificationof technique. Am J Psychiatry 724:251-2.

McCabe MS. (1976) ECT in the treatment of mania: Acontrolled study. Am J Psychiatry 733:688-91.

McCabe MS, Morris B. (1977) ECT versus chlorpromazinein mania. Biol Psychiatry 72:245-54.

McCall WV. (1996) Asystole in electroconvulsive therapy:

P.289

report of four cases. J Clin Psychiatry 57:199-203.

McCall WV, Farah A. (1995) Greater ictal EEC regularityduring RUL ECT is associated with greater treatmentefficacy [abstract]. Convuls Ther 77:69.

McCall WV, Farah BA, Reboussin D, Colenda CC. (1995)Comparison of the efficacy of titrated, moderate-doseand fixed, high-dose right unilateral ECT in elderlypatients. Am J Geriatr Psychiatry 3:317-24.

McCall WV, Reboussin DM, Weiner RD et al. (2000)Titrated moderately supra-threshold vs fixed high-doseright unilateral electroconvulsive therapy. Arch Gen

Psych 57:438-44.

McCall WV, Reid S, Rosenquist P, Foreman A, Kiesow-Webb N. (1993b) A reappraisal of the role of caffeine inECT. Am J Psychiatry 750:1543-5.

McCall WV, Reid S, Ford M. (1994) Electrocardiographicand cardiovascular effects of subconvulsive stimulationduring titrated right unilateral ECT. Convuls Ther 70:25-33.

McCall WV, Shelp FE, Weiner RD, Austin S, Norris J.(1993a) Convulsive threshold differences in rightunilateral and bilateral ECT. Biol Psychiatry 54:606-11.

McCall WV, Shelp FE, Weiner RD, Austin S, Harrill A.(1991) Effects of labetalol on hemodynamics and seizureduration during ECT. Convuls Ther 7:5â!”14.

McCall WV, Weiner RD, Carroll BJ, Shelp FE, Ritchie JC,Austin S, Norris RN. (1996) Serum prolactin, electrodeplacement, and the convulsive threshold during ECT.Convuls Ther 72:81-5.

McClelland R, McAllister G. (1988) Comparison ofelectrical measurements on constant voltage andconstant current ECT machines. Br J Psychiatry 153:126-7.

McConnell KA, Nahas Z, Shastri A et al. (2001) Thetranscranial magnetic stimulation motor thresholddepends on the distance from coil to underlying cortex:A replication in healthy adults comparing two methods ofassessing the distance to cortex. Biol Psychiatry 49:454-9.

McCormick WF, Acosta-Rua GJ. (1965) The size ofintracranial saccular aneurysms. An autopsy study. JNeurosurg 22:155-9.

McCreadie RG, Phillips K, Robinson AD, Gilhooly G,Crombie W. (1989) Is electroencephalographicmonitoring of electroconvulsive therapy useful ? Br J


McElroy SL, Keck PE, Friedman LR. (1995) Minimizingand managing antidepresant side effects. J Clin

Psychiatry 56 (Suppl 2,1:49-55.

McKenna G, Engle RP, Jr, Brooks H, Dalen J. (1970)Cardiac arrhythmias during electroshock therapy:significance, prevention, and treatment. Am J Psychi atry

727:530-3.

McKinney PA, Beale MD, Kellner CH. (1998)Electroconvulsive therapy in a patient with a cerebellarmeningioma. J ECT 14:49-52.

Meco G, Casacchia M, Carchedi F, Falaschi P, Rocco A,Frajese G. (1978) Prolactin response to repeatedelectroconvulsive therapy in acute schizophrenia. Lancet

2:999.

P.290

Medical Research Council. (1965) Clinical trial of thetreatment of depressive illness. Br Med 75459:881-6.

Meduna L. (1985) Autobiography. Part 1. Convuls Ther

7:43-57.

Meduna LJ (1932) Klinische und anatomische Beitrag ziirFrage der genuinen Epilepsie. Deutsche Zeitung der

Nervenkrankheiten 729:17-42.

Meduna LJ (1934) Uber experimentelle Campherepilepsie.Arch Psychiatr Nervenkr 102:333-9.

Meisel A, Roth LH, Lidz CW. (1977) Toward a model ofthe legal doctrine of informed consent. Am J Psychiatry

134:285-9.

Meldrum BS. (1986) Neuropathological consequences ofchemically and electrically induced seizures. Ann N Y

Acad Sci 462:186-93.

Mendels J. (1965a) Electroconvulsive therapy anddepression: I. The prognostic significance of clinicalfactors. Br J Psychiatry 777:675-81.

Mendels J. (1965b) Electroconvulsive therapy anddepression: II. Significance of endogenous and reactivesyndromes. Br J Psychiatry 777:682-6.

Mendels J. (1965c) Electroconvulsive therapy anddepression: III. A method for prognosis. Br J Psychiatry

777:687-90.

Mendels J. (1967) The prediction of response toelectroconvulsive therapy. Am J Psychiatry 724:153-9.

Menken M, Safer J, Goldfarb C, Varga E. (1979) MultipleECT: morphologic effects. Am J Psychiatry 3(5:453.

Mensah GA, Schoen RE, Devereux RB. (1990)Intracardiac thrombi in patients undergoingelectroconvulsive therapy. Am Heart J 779:684-5.

Merritt HH, Putnam TJ. (1938) New series ofanticonvulsant drugs tested by experiments on animals.Arch Neural Psychiatry 39:1003-15.

Messina AG, Paranicas M, Katz B, Markowitz J, Yao F-S,Devereux RB. (1992) Effect of electroconvulsive therapyon the electrocardiogram and echocardiogram. Anesth

Analg 75:511-4.

Miller AL, Faber RA, Hatch JP, Alexander HE. (1985)Factors affecting amnesia, seizure duration, and efficacyin ECT. Am J Psychiatry 742:692-6.

Miller DH, Clancy J, Cummings E. (1953) A comparisonbetween unidirectional current nonconvulsive electricalstimulation given with Reiter's machine, standardalternating current electroshock and pentothal in chronicschizophrenia. Am J Psychiatry 709:617-20.

Miller E. (1970) The effect of ECT on memory andlearning. Br J Med Psychol 43: 57-62.

Milo TJ, Kaufman GE, Barnes WE et al. (2001) Changesin regional cerebral blood flow after electroconvulsivetherapy for depression. J ECT 77:15-21.

Milstein V, Small JG, Klapper MH, Small IF, Miller MJ,Kellams JJ. (1987) Universus bilateral ECT in thetreatment of mania. Convuls Ther 3:1-9.

Misiaszek J, Cork RC, Hameroff SR, Finley J, Weiss IL.(1984) The effect of electroconvulsive therapy on plasmaB-endorphin. Biol Psychiatry 79:450-5.

Mitchell P, Smythe G, Torda T. (1990) Effect of theanesthetic agent propofol on hormonal responses to ECT.Biol Psychiatry 25:315-24.

Mitchell P, Torda T, Hickie I, Burke C. (1991) Propofol asan anaesthetic agent for ECT: effect on outcome andlength of course. Aust N Z J Psychiatry 25: 255-61.

Modigh K, Balldin J, Eriksson E, Granerus AK, Walinder J.(1984) Increased responsiveness of dopamine receptorsafter ECT: A review of experimental and clinicalevidence. In: B Lerer, RD Weiner, RH Belmaker, eds.ECT: Basic Mechanisms . London: John Libbey, pp. 18-27.

Moir DC, Crooks J, Cornwell WB, O'Malley K, Dingwall-Fordyce I, Turnbull MJ, et al. (1972) Cardiotoxicity ofamitriptyline. Lancet 2:561-4.

Monke JV. (1952) Electroconvulsive therapy followingsurgical correction of aortic coarctation by implantationof an aortic isograft. Am J Psychiatry 109: 378-9.

Moore MB. (1960) Electroconvulsive therapy and theaorta. Can Med Assoc J 83: 1258-9.

Moore RA. (1977) The electroconvulsive therapy fight inCalifornia. J Forensic Sci 22:845-50.

Moriarty JD, Siemens JC. (1947) Electroencephalographicstudy of electric shock therapy. Arch Neural Psychiatry

57:693-711.

Mowbray RM. (1959) Historical aspects of electric

P.291

convulsant therapy. Scott Med J 4:373-8.

Mulgaokar GD, Dauchot PJ, Duffy JP, Anton AH. (1985)Noninvasive assessment of electroconvulsive-inducedchanges in cardiac function. J Clin Psychiatry 46:479-82.

Muller DJ. (1971) Unilateral ECT. (One year's experienceat a city hospital). Dis eases of the Nervous System

32:422-4.

Murray GB, Shea V, Conn DK. (1986) Electroconvulsivetherapy for poststroke depression. J din Psychiatry

47:258-60.

Najjar F, Guttmacher LB (1998) ECT in the presence ofaneurysm. [case report]. J ECT 14:266-1â!"1.

Nakajima T, Daval JL, Gleiter CH, Deckert J, Post RM,Marangos PJ. (1989) C-fos mRNA expression followingelectrical -induced seizure and acute nociceptive stress inmouse brain. Epilepsy Res 4:156-9.

National Institutes of Health. (1985) ConsensusConference: Electroconvulsive therapy. JAMA 254:2103-8.

Nelson JP, Rosenberg DR. (1991) ECT treatment ofdemented elderly patients with major depression: Aretrospective study of efficacy and safety. Convuls Ther

7:157-65.

Nemeroff CB, Bissette G, Akil H, Fink M. (1991)Neuropeptide concentrations in the cerebrospinal fluid ofdepressed patients treated with electroconvulsivetherapy: corticotrophin-releasing factor, (3-endorphinand somatostatin. Br J Psychiatry 158:59-63.

Nerozzi D, Graziosi S, Melia E, Aceti F, Magnani A, FiumeS, Fraioli F, Frajese G. (1987) Mechanism of action ofECT in major depressive disorders: A neuroendocrineinterpretation. Psychiatry Res 20:207-13.

Neuberger K, Whitehead RH, Rutledge EK, Ebaugh FG.(1942) Pathologic changes in the brains of dogs givenrepeated electric shocks. Am J Merit Sci 204: 381-7.

Newman ME, Lerer B. (1988) Chronic electroconvulsiveshock and desipramine reduce the degree of inhibitionby 5-HT and carbachol of forskolin-stimulated adenylatecyclase in rat hippocampal membranes. Eur J Pharmacol

148: 257-60.

News and Notes (1992) Dick Cavett and ECT. Convuls

Ther 8:300-1.

Ng C, Schweitzer I, Alexopoulos et al. (2000) Efficacyand cognitive effects of right unilateral electroconvulsivetherapy. J ECT 76:370-79.

Nibuya M, Morinobu S, Duman RS. (1995) Regulation ofBDNF and trkB mRNA in rat brain by chronicelectroconvulsive seizure and antidepressant drugtreatments. J Neurosci 75:7539-47.

Niedermeyer E. (1996) Primary (idiopathic) generalizedepilepsy and underlying mechanisms. Clin

Electroencephalogr 27:1-21.

Nobler MS, Luber B, Moeller JR. (2000) Quantitative EECduring seizures induced by electroconvulsive therapy:relations to treatment modality and clinical features I.Global analyses. J ECT 76:211-28.

Nobler MS, Oquendo MA, Kegeles LS et al. (2001)Decreased regional brain metabolism after ECT. Am J

P.292

Psychiatry / 53:305-8.

Nobler MS, Sackeim HA, Prohovnik I, et al. (1994)Regional cerebral blood flow in mood disorders, III:Treatment and clinical response. Arch Gen Psychiatry

57:884-97.

Nobler MS, Sackeim HA, Solomou M, et al. (1993) EEGmanifestations during ECT: effects of electrodeplacement and stimulus intensity. Biol Psychiatry

34:321-30.

Nolen WA, Zwaan WA. (1990) Treatment of lethalcatatonia with electroconvulsive therapy and dantrolenesodium: A case report. Acta Psychiatr Scand 82: 90-2.

Nutt DJ, Gleiter CH, Glue P. (1989)Neuropharmacological aspects of ECT: in search of theprimary mechanism of action. Convuls Ther 5:250-60.

Nyiro J, Jablonsky A. (1929) Einige daten zur prognoseder Epilepsie, mit besonderer rucksicht auf diekonstitution. Psychiatr Neural Wochenschr 31:541-9.

Nystrom S. (1964) On relation between clinical factorsand efficacy of ECT in depression. Acta Psychiatr Scand

Suppl 181:115-8.

O'Brien PD, Morgan DH. (1991) Bladder rupture duringECT. Convuls Ther 7: 56-9.

O'Connell BK, et al. (1988) Neuronal lesions inmercaptopropionic acid-induced status epilepticus. Acta

Neuropathol (Berl) 77:47-54.

O'Connor MK, Knapp R, Husain M et al. (in preparation)The influence of age on the response of major

depression to electroconvulsive therapy: A C O R Ereport.

O'Dea JPK, Gould D, Hallberg M, Wieland RG. (1978)Prolactin changes during electroconvulsive therapy. Am J


O'Dea JPK, Llerna LA, Hallberg M, Wieland RG. (1979)Specificity of pituitary responses to electroconvulsivetherapy. / Irish Med Assoc 72:490-2.

O'Donnell MP, Webb MGT. (1986) Post-ECT bloodpressure rise and its relationships to cognitive andaffective change. Br J Psychiatry 749:494-7.

O'Flaherty D, Husain MM, Moore M, Wolff TR, Sills S,Giesecke AH. (1992) Circulatory responses duringelectroconvulsive therapy: The comparative effects ofplacebo, esmolol and nitroglycerin. Anesthesia 47:563-7.

Ohman R, Walinder J, Balldin J, Wallin L, Abrahamsson L.(1976) Prolactin response to electroconvulsive therapy.Lancet 2:936-8.

Olesen AC, Lolk A, Christensen P. (1989) Effect of asingle nighttime dose of oxazepam on seizure duration inelectroconvulsive therapy. Convuls Ther 5: 3-7.

Olfson M, Marcus S, Sackeim HA et al. (1998) Use ofECT for the inpatient treatment of recurrent majordepression. Am J Psychiatry 755:22-9.

Oliviero A, Di Lazzaro V, Piazza O et al. (1999) Cerebralblood flow and metabolic changes produced by repetitivemagnetic brain stimulation. J Neurol 246: 1164-8.

Ontario Medical Association and the Ontario Psychiatric

P.293

Association. (1985) Position Paper on Electro-ConvulsiveTherapy of the Ontario Medical Association and theOntario Psychiatric Association [excerpt]. Schedule â

!œHâ! ! of Report of the Electroconvulsive Therapy

Review Committee . Ontario: Ontario GovernmentBookstore, p. 87, December.

Osborne RG, Tunakan B, Barmore J. (1963) Anaestheticagent in electroconvulsive therapy: A controlledcomparison. J Nerv Ment Dis 737:297-300.

Ottosson J-O. (1960) Experimental studies of the modeof action of electroconvulsive therapy. Acta Psychiatr

Neurol Scand Suppl 745:1-141.

Ottosson J-O. (1962a) Electroconvulsive therapy ofendogenous depression: An analysis of the influence ofvarious factors on the efficacy of therapy J Ment Sci

708:694-703.

Ottosson J-O. (1968) Psychological theories of ECT: Areview. Psychological or physiological theories of ECT.Int J Psychiatry 5:170-4.

Ottosson J-O. (1992) Ethics of electroconvulsive therapy.Convuls Ther 8:233-6.

Ottosson J-O. (1995) Ethical aspects of research andpractice of ECT [lecture]. Convuls Ther 77:288-99.

Owens DG, Johnstone EC, Crow TJ, Frith CD, Jagoe JR,Kreel L. (1985) Lateral ventricular size in schizophrenia:relationship to the disease process and its clinicalmanifestations. Psychol Med 15:27-41.

Packman PM, Meyer DA, Verdun RM. (1978) Hazards ofsuccinylcholine administration during electrotherapy.

Arch Gen Psychiatry 35:1137-41.

Padberg F, Zwanzger P, Thoma H et al. (1999) Repetitivetranscranial magnetic stimulation (rTMS) inpharmacotherapy-refractory major depression:comparative study of fast, slow and sham rTMS.Psychiatry Res 55:163â!”71.

Palmer RL, Mani C, Abdel-Kariem MAA, Brandon S.(1990) Dexamethasone suppression tests in the contextof a double-blind trial of electroconvulsive therapy andsimulated ECT. Convuls Ther 6:13-8.

Pande AC, Grunhaus LJ, Aisen AM, Haskett RF. (1990) Apreliminary magnetic resonance imaging study of ECT-treated depressed patients. Biol Psychiatry 27:102-4.

Pande AC, Krugler T, Haskett RF, Greden JF, GrunhausLJ. (1988) Predictors of response to electroconvulsivetherapy in major depressive disorder. Biol Psychiatry

24:91-3.

Papakostas Y, Fink M, Lee J, Irwin P, Johnson L. (1981)Neuroendocrine measures in psychiatric patients: courseand outcome with ECT. Psychiatr Res 4: 55-64.

Papakostas Y, Markianos M, Papadimitriou G, Stefanis C.(1986a) Prolactin response induced by ECT and TRH. Br J


Papakostas Y, Markianos M, Papadimitriou G, Stefanis C.(1990) Thyrotropin and prolactin secretion during ECT:implications for the mechanism of ECT action. Convuls

Ther 6:214-20.

Papakostas Y, Markianos M, Stefanis C. (1988)Methysergide reduces the prolactin response to ECT. Biol


Papakostas Y, Stefanis C, Markianos M, Papadimitriou G.(1986b) Electrode placement and prolactin response toelectroconvulsive therapy. Convuls Ther 2: 99-107.

Papakostas Y, Stefanis C, Sinouri A, Trikkas G,Papadimitriou G, Pittoulis S. (1984) Increases inprolactin levels following bilateral and unilateral ECT.Am J Psychiatry 141:1623-4.

Papakostas YG, Stefanis CS, Markianos M, PapadimitriousGN. (1985) Naloxone fails to block ECT-induced prolactinincrease. Biol Psychiatry 20:1326-7.

Parab AL, Chaudhari LS, Apte J. (1992) Use ofnitroglycerine ointment to prevent hypertensive responseduring electroconvulsive therapyâ!”study of 50 cases. JPostgrad Med 38:55-7.

Partap M, Jos CJ, Dye CJ. (1983) Vasopressin -8-lysine inprevention of ECT-induced amnesia. Am J Psychiatry

140:946-7.

Pascual-Leone A, Rubio B, Pallardo F et al. (1996) Rapid-rate transcranial magnetic stimulation of left dorsolateralprefrontal cortex in drug-resistant depression. Lancet

348:233-7.

Pascual-Leone A, Valls-Sole J, Brasil -Neto JP et al.(1992) Seizure induction and transcranial magneticstimulation [letter] Lancet 339:997.

Patkar AA, Hill KP, Weinstein SP et al. (2000) ECT in thepresence of brain tumor and increased pressure:evaluation and reduction of risk [case report]. J ECT

76:189-97.

Paulson GW. (1967) Exacerbation of organic brain

P.294

disease by electroconvulsive treatment. North Carolina

Medical Journal 25:328-31.

Paus T, Jech R, Thompson CJ et al. (1997) Transcranialmagnetic stimulation during positron emissiontomography: A new method for studying connectivity ofthe human cerebral cortex. J Neurosci 77:3178-84.

Paus T, Jech R, Thompson CJ et al. (1998) Dose-dependent reduction of cerebral blood flow during rapid-rate transcranial magnetic stimulation of the humansensorimotor cortex. J Neurophysiol 79:1102-7.

Pearlman C. (1990) Neuroleptic malignant syndrome andelectroconvulsive therapy. Convuls Ther 6:251-4.

Pearlman C, Richmond J. (1990) New data on themethohexital -thiopental-arrhythmia issue. Convuls Ther

6:221-3.

Pearlman T, Loper M, Tillery L. (1990) Shouldpsychiatrists administer anesthesia for ECT? Am J


Penfield W, Jasper H. (1954) Epilepsy and FunctionalAnatomy of the Human Brain. Boston: Little Brown.

Penney JF, Dinwiddie SH, Zorumski CF, Wetzel RD.(1990) Concurrent and close temporal administration oflithium and ECT. Convuls Ther 6:139-45.

Peretti CS, Danion JM, Grange D et al. (1996) BilateralECT and autobiographical memory of subjectiveexperiences related to melancholia: A pilot study. JAffect Disord 47:9-15.

Perrin GM. (1961) Cardiovascular aspects of electric

shock therapy. Acta Psychiatr Neural Scand 36:1-45.

Perry GF. (1983) ECT for dementia and catatonia[letter]. J Clin Psychiatry 44:117.

Peters SG, Wochos DN, Peterson GC. (1984) Statusepilepticus as a complication of concurrentelectroconvulsive and theophylline therapy. Mayo Clin

Proc 59:568-70.

Peterson G. (1991) San Francisco official attacks ECT.The Psychiatric Times, May.

Petrides G, Fink M. (1996) The â!œhalf -ageâ! !stimulation strategy for ECT dosing. Convuls Ther

72:138-46.

Petrides G, Fink M, Husain M et al. (2001) ECTRemission in psychotic versus non-psychotic depressedpatients: A Report from CORE. J ECT 77:244-53.

Petrides G, Kellner C, Knapp. R et al. (2000) Can IctalEEG Indices Predict Response to ECT? [poster].Presented at the annual NCDEU meeting, January, 2000.

Pettinati HM. (1994) Speed of ECT? [letter]. Convuls

Ther 70:69-71.

Pettinati HM, Nilsen S. (1985) Missed and brief seizuresduring ECT: differential response between unilateral andbilateral electrode placement. Biol Psychi atry 20:506-14.

Pettinati HM, Rosenberg J. (1984) Memory self-ratingsbefore and after electroconvulsive therapy: depressivevsECT-induced. Biol Psychiatry 70:539-48.

P.295

Pettinati HM, Stephens RN, Willis KM, Robin S. (1990)Evidence for less improvement in depression in patientstaking benzodiazepines during unilateral ECT. Am J


Pettinati HM, Tamburello BA, Ruetsch CR, Kaplan FN.(1994) Patient attitudes toward electroconvulsivetherapy. Psychopharmacol Bull 30:471-5.

Petty F. (1995) GAB A and mood disorders: A briefreview and hypothesis. J Affect Disord 34:275-81.

Petty F, Kramer GL, Gullion CM et al. (1992) Low plasmagamma-aminobutyric acid levels in male patients withdepression. Biol Psychiatry 32:354-63.

Pinel JP, Van Oot PH. (1975) Generality of the kindlingphenomenon: some clinical implications. Can J Neural Sci

2:467-75.

Pinel JP, Van Oot PH. (1977) Intensification of thealcohol withdrawal syndrome following periodicelectroconvulsive shocks. Biol Psychiatry 72:479-86.

Pippard J, Ellam L. (1981) Electroconvulsive Treatment in

Great Britain, 1980 . London: Gaskell.

Pisvejc J, Hyrman V, Sikora J et al. (1998) A comparisonof brief and ultrabrief pulse stimuli in unilateral ECT. JECT 74:68-75.

Pitts FM, Jr, Desmarias GM, Stewart W, Schaberg K.(1965) Induction of anesthesia with methohexital andthiopental in electroconvulsive therapy. N Engl J Med

273:353-60.

Pitts FN, Jr. (1982) Medical physiology of ECT. In: R

Abrams, WB Essman, eds. Electroconvulsive Therapy:

Biological Foundations and Clinical Applica tions. NewYork: Spectrum Publications, pp. 57-90.

Pollard BJ, O'Leary J. (1981) Guedel airway and toothdamage. Anaesthesia and Intensive Care 9:395.

Pollitt JD. (1965) Suggestions for a physiologicalclassification of depression. Br J Psychiatry 777:489-95.

Pomeranze J, Karliner W, Triebel WA, King EJ. (1968)Electroshock therapy in presence of serious organicdisease. Depression and aortic aneurysm. Geri atrics

23:122-4.

Pope H, Lipinski JF, Cohen BM, Axelrod DT. (1980) â!œSchizoaffective disorderâ! !: An invalid diagnosis? Acomparison of Schizoaffective disorder, schizophrenia,and affective disorder. Am J Psychiatry 737:921-7.

Pope HG, Lipinski JF. (1978) Diagnosis in schizophreniaand manic -depressive illness: A reassessment of thespecificity of â!œschizophrenicâ! ! symptoms in the lightof current research. Arch Gen Psychiatry 35:811-28.

Posner JB, Plum F, Van Poznak A. (1969) Cerebralmetabolism during electrically induced seizures in man.Arch Neural 20:388-95.

Post RM, Kimbrell TA, McCann UD et al. (1999)Repetitive transcranial magnetic stimulation asneuropsychiatric tool: present status and futurepotential. J ECT 75:39-59.

Post RM, Putnam F, Contel NR, Goldman B. (1984)Electroconvulsive seizures inhibit amygdala kindling:implications for mechanisms of action in affective illness.Epilepsia 25:234-9.

P.296

Powell JC, Silveira WR, Lindsay R. (1988) Pre-pubertaldepressive stupor: A case report. Br J Psychiatry

753:689-92.

Pratt RT, Warrington EK, Halliday AM. (1971) UnilateralECT as a test for cerebral dominance, with a strategy fortreating left -handers. Br J Psychiatry 119: 79-83.

Price TR. (1982a) Short-and long-term cognitive effectsof ECT: Part Iâ!”Effects on memory. Psychopharmacol

Bull 78:81-91.

Price TR. (1982b) Short-and long-term cognitive effectsof ECT: Part IIâ!”Effects on nonmemory associatedcognitive functions. Psychopharmacol Bull 18: 91-101.

Price TR, Levin R. (1978) The effects of electroconvulsivetherapy on tardive dyskinesia. Am J Psychiatry 735:991-3.

Price TR, Mackenzie TB, Tucker GJ, Culver C. (1978) Thedose-response ratio in electroconvulsive therapy: Apreliminary study. Arch Gen Psychiatry 35: 1131-6.

Price TR, McAllister TW. (1989) Safety and efficacy ofECT in depressed patients with dementia: A review ofclinical experience. Convuls Ther 5:1-74.

Price TR, Tucker GJ. (1977) Psychiatric and behavioralmanifestations of normal pressure hydrocephalus. J Nerv

Ment Dis 164:51-5.

Pridmore S, Ryback M. (1999) Supraorbital and d'Eliaelectrode positions in unilateral ECT? [letter]. J ECT

75:170-1.

Pridmore S. (2001) A failure to find large visible effextsof ECT-induced seizures on SPECT scans. J ECT 77:155-6.

Proctor LD, Goodwin JE. (1943) Comparativeelectroencephalographic observations followingelectroshock therapy using raw 60 cycle alternating andunidirectional fluctuating current. Am J Psychiatry

99:525-30.

Prohovnik I, Sackeim HA, Decina P, Malitz S. (1986)Acute reductions of regional cerebral blood flow followingelectroconvulsive therapy: interactions with modality andtime. Ann N Y Acad Sci 462:249-62.

Prudic J, Devanand DP, Sackeim HA, Decina P, Kerr B.(1989) Relative response of endogenous and non-endogenous symptoms to electroconvulsive therapy. JAffect Disord 16:59-64.

Prudic J, Fitzsimons L, Nobler MS et al. (1999) Naloxonein the prevention of the adverse cognitive effects ofECT: A within-subject, placebo controlled study.NeuroPsychopharmacology 27:285-93.

Prudic J, Olfson M, Sackeim HA. (2001) Electro-convulsive therapy practices in the community. Psychol

Med 37:929-34.

Prudic J, Peyser S, Sackeim HA. (2000) Subjectivememory complaints: A review of patient self-assessmentof memory after electroconvulsive therapy. J ECT

76:121-32.

Prudic J, Sackeim HA. (1999) Electroconvulsive therapyand suicide risk. J Clin Psychiatry 60 (Suppl 2): 104-10.

Prudic J, Sackeim HA, Decina P, Hopkins N, Ross FR,

Malitz S. (1987) Acute effects of ECT on cardiovascularfunctioning: relations to patient and treatment variables.Acta Psychiatr Scand 75:344-51.

Prudic J, Sackeim HA, Devanand DP, Krueger RB,Settembrino JM. (1994) Acute cognitive effects ofsubconvulsive stimulation. Convuls Ther 10:4-24.

Railton R. (1987) Comparison of electrical measurementson constant voltage and constant current ECT machines[letter]. Br J Psychiatry 757:701.

Railton R, Fisher J, Sinclair A, Shrigmankar JM. (1987)Comparison of electrical measurements on constantvoltage and constant current ECT machines. Br J


Rampton AJ, Griffin RM, Stuart CS, Durcan JJ, Huddy NC,Abbott MA. (1989) Comparison of methohexital andpropofol for electroconvulsive therapy: effects onhemodynamic responses and seizure duration.Anesthesiology 70: 412-7.

Ranck JB (1975) Which elements are excited in electricalstimulatin of mammalian central nervous sytem: Areview. Brain Res 98:417-40.

Rao KM, Gangadhar BN, Janakiramaiah N. (1993)Nonconvulsive status epilepticus after the ninthelectroconvulsive therapy. Convuls Ther 9:128-34.

Rasmussen KG, Abrams R. (1991) Treatment ofParkinson's disease with electroconvulsive therapy.Psychiatr Clin North Am 14:925-33.

Rasmussen KG, Abrams R (1992) The role ofelectroconvulsive therapy in Parkinson's disease. In:Huber S, Cummings J, eds. Neurobehavioral Aspects of

P.297

Parkinson's Disease . New York: Oxford University Press,pp. 252-70.

Rasmussen KG, Pandurangi AL. (1999) Naloxone fails toprolong seizure length in ECT. J ECT 75:258-61.

Rasmussen KG, Zorumski CF. (1993) Electroconvulsivetherapy in patients taking theophylline. J Clin Psychiatry

54:427-31.

Rasmussen KG, Zorumski CF, Jarvis MR. (1994) Possibleimpact of stimulus duration on seizure threshold in ECT.Convuls Ther 70:177-80.

Rausch JL, Rich CL, Risch SC. (1988) Platelet serotonintransport after a single ECT. Psychopharmacology

95:139-41.

Ray I. (1975) Side effects from lithium. Can Med Assoc J

7/2:417-9.

Reed RR, Ciesel C, Owens G. (1971) Induced seizures astherapy of experimental strokes in dogs. J Neurosurgery

34:178-84.

Regestein QR, Kahn CB, Siegel AJ, Blacklow RS, GenackA. (1971) A case of catatonia occurring simultaneouslywith severe urinary retention. J Nerv Ment Dis 752:432-5.

Regestein QR, Murawski BJ, Engle RP. (1975) A case ofprolonged, reversible dementia associated with abuse ofelectroconvulsive therapy. J Nerv Ment Dis 161:200-3.

Regestein QR, Reich P. (1985) Electroconvulsive therapyin patients at high risk for physical complications.Convuls Ther 7:101-14.

Reid AH. (1972) Psychoses in adult mental defectives. I.Manic depressive psychosis. Br J Psychiatry 120:205-12.

Reiter-Theil S. (1992) Autonomy and beneficence:ethical issues in electroconvulsive therapy. Convuls Ther

5:237-44.

Remick RA, Maurice WL. (1978) ECT in pregnancy[letter]. Am J Psychiatry 135: 761-2.

Repke JT, Berger NG. (1984) Electroconvulsive therapyin pregnancy. Obstetr Gy necol 63:39S-41S.

Rice EH, Sombrotto LB, Markowitz JC, Leon AC. (1994)Cardiovascular morbidity in high-risk patients. Am J


Rich CL, Black NA. (1985) The efficiency of ECT: II.Correlation of specific treatment variables to responserate in unilateral ECT. Psychiatry Res 16: 147-54.

Rich CL, Cunningham LA, Maher CC, Woodruff RA.(1975a) The effect of modified ECT on serum creatinephophokinase, I. With intravenous atropine. Diseases of

the Nervous System 36:653-5.

Rich CL, Cunningham LA, Maher CC, Woodruff RA.(1975b) The effect of modified ECT on serum creatinephosphokinase. II. With subcutaneous atropine. Dis

eases of the Nervous System 36:655â!”6.

Rich CL, Spiker DG, Jewell SW, Neil JF. (1984) DSM-III,RDC, and ECT: depressive subtypes and immediateresponse. J Clin Psychiatry 45:14-8.

Rich CL, Woodruff RA, Jr, Cadoret R, Craig A, Pitts FN,

P.298

Jr. (1969) Electrotherapy: The effects of atropine onEKG. Diseases of the Nervous System 30:622â!”6.

Ries R, Bokan J. (1979) Electroconvulsive therapyfollowing pituitary surgery. J Nerv Ment Dis 767:767-8.

Ries RK. (1985) Poor interrater reliability of MECTA EEGseizure duration measurement during ECT. Biol

Psychiatry 20:94-8.

Ries RK. (1987) Informed ECT for Patients and Families:Informed ECT for Health Professionals; Shock Therapy[videotape review]. Hasp Community Psychi atry 38:131-8.

Rifkin A. (1988) ECT versus tricyclic antidepressants indepression: A review of the evidence. J Clin Psychiatry

49:3-1.

Roberts AH. (1963) The value of ECT in delirium. Br J


Roberts JM. (1959a) Prognostic factors in theelectroshock treatment of depressive states: I. Clinicalfeatures from history and examination. J Ment Sci 105:693-702.

Roberts JM. (1959b) Prognostic factors in theelectroshock treatment of depressive states: II. Theapplication of specific tests. J Ment Sci 705:703-13.

Roberts R, Owens G, Vilisenkas J, Thomas DD. (1972)Induced seizures as therapy for experimental stroke inmonkeys. J Neurosurgery 37:111â!“4.

Robertson AD, Inglis J. (1978) Memory deficits afterelectroconvulsive therapy: cerebral asymmetry and dual-

encoding. Neuropsychologia 76:179-87.

Robin A, Binnie CD, Copas JB. (1985)Electrophysiological and hormonal responses to threetypes of electroconvulsive therapy. Br J Psychiatry

747:707-12.

Robin A, DeTissera S. (1982) A double-blind controlledcomparison of the therapeutic effects of low and highenergy electroconvulsive therapies. Br J Psychiatry

141:351-66.

Robin A A, Harris JA. (1962) A controlled trial ofimipramine and electroplexy. J Ment Sci 106:211-9.

Robins MA, Attah JR. (1988) Carbamazepine and ECT. Br

J Psychiatry 753:418.

Roccaforte WH, Burke WJ. (1989) ECT followingcraniotomy. Psycho somatic s 30: 99-101.

Rochford G, Williams M. (1962) Development andbreakdown of the use of names. J Neural Neurosurg

Psychiatry 25:222.

Rogers MC, Battit G, McPeek B, Todd D. (1978)Lateralization of sympathetic control of the human sinusnode: ECG changes of stellate ganglion block.Anesthesiology 48:139-41.

Roith AI. (1959) Status epilepticus as a complication ofECT. Br J Clin Pract 13: 711-2.

Rollason WN, Sutherland MS, Hall DJ. (1971) Anevaluation of the effect of methohexitone and propanididon blood pressure, pulse rate and cardiac arrhythmiaduring electroconvulsive therapy. Compr Psychiatry

43:160-6.

Rosen AD, Gur RC, Sussman N, Gur RE, Hartig H. (1982)Hemispheric asymmetry in the control of heart rate.Abstracta Soc Neurosci 8:917.

Rosenbach ML, Hermann RC, Dorwart RA. (1997) Use ofelectroconvulsive therapy in the Medicare populationbetween 1987 and 1992. Psychiatr Serv 48: 1537-42.

Rosenberg R, Vorstrup S, Andersen A, Bolwig TG. (1988)Effect of ECT on cerebral blood flow in melancholiaassessed with SPECT. Convuls Ther 4:62-13.

Rosenfeld JE, Glassberg S, Sherrid M. (1988)Administration of ECT 4 years after aortic aneurysmdissection. Am J Psychiatry 145:128-9.

Rosenquist PB, McCall WV, Colenda CC et al. (1998) Acomparison of visual and computer -generated measuresof â!œseizure quality.â! ! J ECT 74:76-82.

Rosenquist PB, McCall WV, Farah A, Reboussin DM.(1994) Effects of caffeine pretreatment on measures ofseizure impact. Convuls Ther 70:181-5.

Rossi A, Stratta P, Nistico R, Sabatini MD, DiMichele V,Casacchia M. (1990) Visuospatial impairment indepression: A controlled ECT study. Acta Psychiatr Scand

81:245-9.

Roth LH, Meisel A, Lidz CW. (1977) Tests of competencyto consent to treatment. Am J Psychiatry 134:219-84.

Roth M. (1951) Changes in the EEG under barbiturateanesthesia produced by electro-convulsive treatment andtheir significance for the theory of ECT action.Electroencephalogr Clin Neurophysiol 3:261-80.

P.299

Roth SD, Mukherjee S, Sackeim HA. (1988)Electroconvulsive therapy in a patient with mania,parkinsonism, and tardive dyskinesia. Convuls Ther

4:92-7.

Rouse EC. (1988) Propofol for electroconvulsive therapy.A comparison with methohexitone. Preliminary report.Anaesthesia 43 Suppl: 61-4.

Royal College of Psychiatrists. (1989) The Practical

Administration of Electrocon vulsive Therapy (ECT) . ECTSub-committee of the Research Committee. London:Gaskell.

Royal College of Psychiatrists. (1995) The ECT

Handbook: The Second Report of the Royal College of

Psychiatrists' Special Committee on ECT . London: RoyalCollege of Psychiatrists.

Rudorfer MV, Linnoila M, Potter WZ. (1987) Combinedlithium and electroconvulsive therapy: pharmacokineticand pharmacodynamic interactions. Convuls Ther 3:40-5.

Ruedrich SL, Chu CC, Moore SL. (1983) ECT for majordepression in a patient with acute brain trauma. Am J


Rummans TA, Bassingthwaigte ME. (1991) Severemedical and neurologic complications associated withnear-lethal catatonia treated with electroconvulsivetherapy. Convuls Ther 7:121-4.

Ryan RJ, Swanson DW, Faiman C, Mayberry WE SpadoniAJ. (1970) Effects of convulsive electroshock on serumconcentrations of follicle stimulating hormone, luteinizinghormone, thyroid stimulating hormone and growth

hormone in man. J Clin Endocrinol Metab 30:51.

Sackeim HA. (199la) Optimizing unilateralelectroconvulsive therapy. Convuls Ther 7:201-12.

Sackeim HA. (1991b) Are ECT devices underpowered?Convuls Ther 7:233-6.

Sackeim HA. (1992) The cognitive effects ofelectroconvulsive therapy. In Thai et al., eds. Cognitive

Disorders: Pathophysiology and Treatment. New York,Marcel Dekker, pp. 183-228.

Sackeim HA. (1994a) Physical properties of the ECTstimulus [Response to commentaries]. Convuls Ther

70:140-52.

Sackeim HA. (1994b) Central issues regarding themechanisms of action of electroconvulsive therapy:directions for future research. Psychopharmacol Bull 30:281-308.

Sackeim HA. (1994c) Magnetic stimulation therapy andECT [commentary]. Con vuls Ther 70:255-8.

Sackeim HA. (1999b) The anticonvulsant hypothesis ofthe mechanism of action of ECT: current status. J ECT

15:5-26.

Sackeim HA. (2000) Memory and ECT: from polarizationto reconciliation [editorial]. J ECT 76:87-96.

Sackeim HA, Decina P, Kanzler M, Kerr B, Malitz S.(1987a) Effects of electrode placement on the efficacy oftitrated, low dosage ECT. Am J Psychiatry 144: 1449-55.

Sackeim HA, Decina P, Malitz S, Hopkins N, Yudofsky SC,

P.300

Prohovnik I. (1983b) Postictal excitement followingbilateral and right-unilateral ECT. Am J Psychiatry

740:1367-8.

Sackeim HA, Decina P, Portnoy S, Neeley P, Malitz S.(1987b) Studies of dosage, seizure threshold, andseizure duration in ECT. Biol Psychiatry 22:249-68.

Sackeim HA, Decina P, Prohovnik I, Malitz S. (1987d)Seizure threshold in electroconvulsive therapy: effects ofsex, age, electrode placement, and number oftreatments. Arch Gen Psychiatry 44:355-60.

Sackeim HA, Decina P, Prohovnik I, Malitz S, Resor SR.(1983a) Anticonvulsant and antidepressant properties ofelectroconvulsive therapy: A proposed mechanism ofaction. Biol Psychiatry 75:1301-10.

Sackeim HA, Decina P, Prohovnik I, Portnoy S, KanzlerM, Malitz S. (1986a) Dosage, seizure threshold, and theantidepressant efficacy of electroconvulsive therapy. Ann

N Y Acad Sci 462:398-410.

Sackeim HA, Devanand DP, Prudic J. (1991) Stimulusintensity, seizure threshold, and seizure duration:Impact on efficacy and safety of electroconvulsivetherapy. In: CH Kellner, ed. Electroconvulsive Therapy:

The Psychiatric Clinics of North America, Vol. 14,Philadelphia: WB Saunders, pp. 803-43.

Sackeim HA, Devanand DP, Prudic J. (1991) Stimulusintensity, seizure threshold, and seizure duration:Impact on the efficacy and safety of electroconvulsivetherapy. Psychiatr Clin North Am 74:803-43.

Sackeim HA, Greenberg MS, Weiman MA, Gur RC,Hungerbuhler JP, Geshwind N. (1982) Hemisphericasymmetry in the expression of positive and negative

emotions: Neurologic evidence. Arch Neural 39:210-8.

Sackeim HA, Haskett RF, Mulsant BH et al. (2001a)Continuation pharmacotherapy in the prevention ofrelapse following electroconvulsive therapy. JAMA 285:1299-307.

Sackeim HA, Long J, Luber B, et al. (1994) Physicalproperties and quantification of the ECT stimulus: I.Basic principles. Convuls Ther 70:93-123.

Sackeim HA, Luber BL, Katzman GP, Moeller JR, Prudic J,Devanand DP, Nobler MS. (1996) The effects ofelectroconvulsive therapy on quantitative EEG:relationship to clinical outcome. Arch Gen Psychiatry

53:814-24.

Sackeim HA, Mukherjee S. (1986) Neurophysiologicalvariability in the effects of the ECT stimulus. Convuls

Ther 2:267-76.

Sackeim HA, Portnoy S, Neeley P, Steif BL, Decina P,Malitz S. (1986b) Cognitive consequences of low-dosageelectroconvulsive therapy. Ann N Y Acad Sci 462:326-40.

Sackeim HA, Prudic J, Devanand DP et al. (2000) Aprospective, randomized, double-blind comparison ofbilateral and right unilateral electroconvulsive therapy atdifferent stimulus intensities. Arch Gen Psych 57:425-34.

Sackeim HA, Prudic J, Devanand DP, Kiersky JE,Fitzimons L, Moody BJ, Mc-Elhiney MC, Coleman EZ,Settembrino JM. (1993) Effects of stimulus intensity andelectrode placement on the efficacy and cognitive effectsof electroconvulsive therapy. N Engl J Med 328:839-46.

Sackeim HA, Prudic J, Nobler MS et al. (2001b) Ultra-

P.301

brief pulse ECT and the affective and cognitiveconsequences of ECT [abstract]. J ECT 17:16.

Sackeim, HA. (1999a) Response: Pulse width andfrequency of ECT [letter]. J ECT 75:288-90.

Salaris S, Szuba MP, Traber K. (2000) ECT and vascularmasses [case report]. J ECT 76:198-203.

Salmon JB, Hanna MH, Williams TB, Wheeler M. (1988)Thalamic painâ!”the effect of electroconvulsive therapy(ECT). Pain 33:67-71.

Salzman C. (1977) ECT and ethical psychiatry. Am J


Sandford JL. (1966) Electric and convulsive treatmentsin psychiatry. Diseases of the Nervous System 27:333-8.

Savitsky N, Karliner W. (1951) Electroshock therapy andmultiple sclerosis. N Y J Med 57:788.

Savitsky N, Karliner W. (1953) Electroshock in thepresence of organic disease of the nervous system.Journal of Hillside Hospital 2:3-22.

Schillinger D. (1987) Nifedipine in hypertensiveemergenciesâ!”a prospective study. J Emerg Med 5:463-73.

Schwartz M. (1985) Computed tomography and ECT.Convuls Ther 7:70-1.

Scott AI. (1989) Which depressed patients will respondto electroconvulsive therapy? The search for biologicalpredictors of recovery. Br J Psychiatry 754:81-7.

Scott AI, Douglas RH, Whitfield A, Kendell RE. (1990)Time course of cerebral magnetic resonance changesafter electroconvulsive therapy. Br J Psychiatry 756:551-3.

Scott AI, McCreadie RG. (1999) Prolonged seizuresdetectable by electroencephalogram in electroconvulsivetherapy [letter]. Br J Psychiatry 775:91-2.

Scott AI, Riddle W. (1989) Status epilepticus afterelectroconvulsive therapy. Br J Psychiatry 755:119-21.

Scott AI, Shering PA, Legros JJ et al. (1991)Improvement in depressive illness is not associated withaltered release of neurophysins over a course of ECT.Psychiatry Res 36:65-73.

Scott AI, Turnbull LW. (1990) Do repeated courses ofECT cause brain damage detectable by MRI? Am J


Scott AI, Whalley LJ, Bennie J, Bowler G. (1986)Oestrogen-stimulated neurophysin and outcome afterelectroconvulsive therapy. Lancet 7:1411-4.

Scott AI, Whalley LJ, Legros JJ. (1989) Treatmentoutcome, seizure duration, and the neurophysinresponse to ECT Biol Psychiatry 25:585-97.

Scott AIR (1995) Does ECT alter brain structure?[letter]. Am J Psychiatry 752: 140-3.

Scott AIF, Boddy H. (2000) The effect of repeatedbilateral electroconvulsive therapy on seizure threshold.J ECT 76:244-51.

Scott AIF, Milner JB, Shering PA. (1989b) DiminishedTSH release after a course of ECT: altered monoaminefunction or seizure activity? Psychoneuroendo crinology

74:425-31.

Scott AIF, Turnbull LW, Blane A, Douglas RHB. (1991)Electroconvulsive therapy and brain damage [letter].Lancet 338:264.

Scott BW, Wojtowicz JM, Burnham WM. (2000)Neurogenesis in the dentate gyrus of the rat followingelectroconvulsive shock seizures. Exp Neural 765: 231-6.

Selvin BL. (1987) Electroconvulsive therapy 1987.Anesthesiology 67:367-85.

Senter NW, Winslade WJ, Listen EH, Mills MJ. (1984)Electroconvulsive therapy: The evolution of legalregulation. Am J Soc Psychiatry 4:11-5.

Settle EC. (1998) Antidepressant drugs: disturbing andpotentially dangerous side effects. J Clin Psychiatry 59

(Suppl 76):25-30.

Shagass C, Jones AL. (1958) A neurophysiological testfor psychiatric diagnosis: results in 750 patients. Am J


Shankel LW, Dimassimo DA, Whittier JR. (1960) Changeswith age in electrical reactions in mental patients.Psychiatr Q 34:284-92.

Shapira B, Lerer B, Gilboa D, Drexler H, Kugelmass S,Calev A. (1987) Facilitation of ECT by caffeinepretreatment. Am J Psychiatry 744:1199-202.

Shapira B, Lerer B, Kindler S, Lichtenberg P, Gropp C,Cooper T, Calev A. (1992) Enhanced serotonergic

P.302

responsivity following electroconvulsive therapy inpatients with major depression. Br J Psychiatry 760:223-9.

Shapira B, Lidsky D, Gorfine M, Lerer B. (1996)Electroconvulsive therapy and resistant depression:clinical implications of seizure threshold. J Clin

Psychiatry 57:328.

Shapira B, Newman ME, Gelfin Y et al. (2000) Bluntedtemperature and cortisol responses to ipsapirone inmajor depression: lack of enhancement byelectroconvulsive therapy. Psychoneuroendocrinology

25:421-438.

Shapira B, Tubi N, Drexler H et al. (1998) Cost andbenefit in the choice of ECT schedule. Twice versus threetimes weekly ECT. Br J Psychiatry 772:44-48.

Shapira B, Zohar J, Newman M, Drexler H, Belmaker RH.(1985) Potentiation of seizure length and clinicalresponse to electroconvulsive therapy by caffeinepretreatment: A case report. Convuls Ther 7:58-60.

Shapiro MF, Goldberg HH. (1957) Electroconvulsivetherapy in patients with structural diseases of thecentral nervous system. Am J Med Sci 255:186-95.

Sheline YI, Sanghavi M, Mintun MA et al. (1999)Depression duration but not age predicts hippocampalvolume loss in medically healthy women with recurrentmajor depression. J Neurosci 79:5034-43.

Shellenberger W, Miller MJ, Small IF et al. (1982)Follow-up study of memory deficits after ECT. Can J


Shepley WH, McGregor JS. (1939) Electrically inducedconvulsions in treatment of mental disorders. Br Med J

2:1269-71.

Shettar MS, Grunhaus L, Pande AC, Tandon RC, KnonfolZA, Haskett RF. (1989) Protective effects ofintramuscular glycopyrrolate on cardiac conductionduring ECT. Convuls Ther 5:349-52.

Shiah IS, Yatham LN. (1998) GABA function in mooddisorders: An update and critical review. Life Sci

65:1289-303.

Shimamura AP, Squire LR. (1987) A neuropsychologicalstudy of fact memory and source amnesia. J Exp Psychol

[Learn Mem Cog] 15:464-73.

Siebner H, Peller M, Bartenstein P et al. (2001)Activation of frontal premotor areas duringsuprathreshold transcranial magnetic stimulation of theleft primary sensorimotor cortex: A glucose metabolicPET study. Hum Brain Map 12: 157-67.

Siebner HR, Peller M, Willoch F et al. (2000) Lastingcortical activation after repetitive TMS of the motorcortex: A glucose metabolic study. Neurology 54: 956-63.

Siekert RG, Williams SC, Windle WF. (1950) Histologicstudy of the brains of monkeys after experimentalelectroshock. Arch Neural Psychiatry 65:79-86.

Silverskiold P, Gustafson L, Risberg J, Ingmar R. (1986)Acute and late effects of electroconvulsive therapy:clinical outcome, regional cerebral blood flow, andelectroencephalogram. Ann N Y Acad Sci 462:236-48.

Simpson S, Baldwin RC, Jackson A et al. (1998) Is

P.303

subcortical disease associated with a poor response toantidepressants? Neurological, neuropsychological andneuroradiological findings in late-life depression. Psychol

Med 28: 1015-26.

Singh G, Bryce AB, Black DF. (2001) Electroconvulsivetherapy and Friedreich's ataxia [case report]. J ECT

77:53-4.

Siris SG, Glassman AH, Stetner F. (1982) ECT andpsychotropic medication in the treatment of depressionand schizophrenia. In: R Abrams, WB Essman, eds.Electroconvulsive Therapy: Biological Foundations and

Clinical Applica tions. New York: Spectrum Publications,pp. 91-112.

Skrabanek P, Balfe A, Webb M, Maguire J, Powell D.(1981) Electroconvulsive therapy increases plasmagrowth hormone, prolactin, luteinizing hormone andfollicle-stimulating hormone but not thyrotropin orsubstance P. Psychoneu roendocrinology 6:261-7.

Slade AP, Checkley SA. (1980) A neuroendocrine studyof the mechanism of action of ECT. Br J Psychiatry

137:211-21.

Slawson P. (1985) Psychiatric malpractice: Theelectroconvulsive therapy experience. Convuls Ther

7:195-203.

Slawson P. (1989) Psychiatric malpractice and ECT: Areview of national loss experience. Convuls Ther 5:126-30.

Slawson P. (1991) Psychiatric malpractice and ECT: Areview of 1,700 claims. Con vuls Ther 7:255-61.

Slawson PF, Guggenheim FG. (1984) Psychiatricmalpractice: A review of the national loss experience.Am J Psychiatry 747:979-81.

Small IF. (1974) Inhalant convulsive therapy. In: M Fink,S Kety, J McGaugh, TA Williams, eds. Psychobiology of

Convulsive Therapy . Washington DC: VH Winston andSons, pp. 65-77.

Small IF, Small JG, Milstein V, Moore JE. (1972)Neuropsychological observations with psychosis andsomatic treatment. Neuropsychological examinations ofpsychiatric patients. J Nerv Ment Dis 755:6-13.

Small IF, Small JG, Milstein V, Sharpley P. (1973)Interhemispheric relationships with somatic therapy.Diseases of the Nervous System 34:170-7.

Small JG, Kellams JJ, Milstein V, Small IF. (1980)Complications with electroconvulsive treatment combinedwith lithium. Biol Psychiatry 75:103-12.

Small JG, Klapper MH, Kellams JJ, Miller MJ, Milstein V,Sharpley PH, Small IF. (1988) ECT compared with lithiumin the management of manic states. Arch Gen Psychiatry

45:727-32.

Small JG, Small IF, Milstein V, Kellams JJ, Klapper MH.(1985) Manic symptoms: An indication for bilateral ECT.Biol Psychiatry 20:125-34.

Small JG, Small IF, Perez HC, Sharpley P. (1970)Electroencephalographic and neurophysiological studiesof electrically induced seizures. J Nerv Ment Dis

750:479-89.

Smith J, Williams K, Birkett S et al. (1994)Neuroendocrine and clinical effects of electroconvulsive

P.304

therapy and their relationship to treatment outcome. Psy

chol Med 24:547-55.

Smith JE, Williams K, Burkett S, Glue P, Nutt DJ. (1990)Oxytocin and vasopressin responses to ECT. Psychiatry

Res 32:201-2.

Smith MA, Zhang LX, Lyons WE, Mamounas LA. (1997)Anterograde transport of endogenous brain-derivedneurotrophic factor in hippocampal mossy fibers.Neuroreport 8:1829-34.

Smith S. (1956) The use of electroplexy (ECT) inpsychiatric syndromes complicating pregnancy. J Ment

Sci 702:796-800.

Snow SS, Wells CE. (1981) Case studies inneuropsychiatry: diagnosis and treatment of coexistentdementia and depression. J Clin Psychiatry 42:439-41.

Snowdon DA, Greiner LH, Markesbery WR. (2000)Linguistic ability in early life and the neuropathology ofAlzheimer's disease and cerebrovascular disease.Findings from the Nun Study. Ann N Y Acad Sci 903:34-8.

Sobel DE. (1960) Fetal damage due to ECT, insulincoma, chlorpromazine or reserpine. Arch Gen Psychiatry

2:606-11.

Sobin C, Prudic J, Devanand DP et al. (1996) Whoresponds to electroconvulsive therapy? A comparison ofeffective and ineffective forms of treatment. Br J


Sobin C, Sackeim HA, Prudic J, Devanand DP, Moody BJ,McElheiney MC. (1995) Predictors of retrograde amnesia

following ECT. Am J Psychiatry 152: 995-1001.

Solomons K, Holliday S, Illing M. (1998) Non-convulsivestatus epilepticus complicating electroconvulsive therapy[letter]. Int J Geriatr Psychiatry 13: 731-4.

Sommer BR, Satlin A, Friedman MS. (1989)Glycopyrrolate versus atropine in post-ECT amnesia inthe elderly. J Geriatr Psychiatry Neural 2:18-21.

Sorensen PS, Bolwig TG, Lauritsen B, Bengtson O.(1981) Electroconvulsive therapy: A comparison ofseizure duration as monitored withelectroencephalograph and electromyograph. Acta

Psychiatry Scand 64:193-8.

Speer AM, Kimbrell TA, Wasserman EM et al. (2000)Opposite effects of high and low frequency rTMS onregional brain activity in depressed patients. Biol


Sperling MR, Melmed S, McAllister T, Price TR. (1989)Lack of effect of naloxone on prolactin and seizures inelectroconvulsive therapy. Epilepsia 50:31-44.

Sperling MR, Pritchard PB, III, Engel J, Jr, Daniel C,Sagel J. (1986) Prolactin in partial epilepsy: An indicatorof limbic seizures. Ann Neural 20:716-22.

Sperling MR, Wilson CL. (1986) The effects of limbic andextralimbic electrical stimulations upon prolactinsecretions in humans. Brain Res 371:29-31.

Spiker DO, Dealy RS, Hanin I, Weiss JC, Kupfer DJ.(1986) Treating delusional depression with amitriptyline.J Clin Psychiatry 47:24-35.

Squire LR. (1982) Neuropsychological effects of ECT. In:

R Abrams, WB Essman, eds. Electroconvulsive Therapy:

Biological Foundations and Clinical Appli cations . NewYork: Spectrum Publications, pp. 169-86.

Squire LR, Chace PM. (1975) Memory functions six tonine months after electroconvulsive therapy. Arch Gen


Squire LR, Miller PL. (1974) Diminution of anterogradeamnesia following electroconvulsive therapy. Br J


Squire LR, Slater PC. (1983) Electroconvulsive therapyand complaints of memory dysfunction: A prospectivethree-year follow-up study. Br J Psychiatry 742:1-8.

Squire LR, Slater PC, Chace PM. (1975) Retrogradeamnesia: temporal gradient in very long-term memoryfollowing electroconvulsive therapy. Science 187: 77-9.

Squire LR, Slater PC, Chace PM. (1976) Anterogradeamnesia following electroconvulsive therapy: no evidencefor state-dependent learning. Behav Biol 17: 31-41.

Squire LR, Slater PC, Miller PL. (1981) Retrogradeamnesia and bilateral electroconvulsive therapy. Long-term follow-up. Arch Gen Psychiatry 38:89-95.

Squire LR, Wetzel CD, Slater PC. (1979) Memorycomplaint after electroconvulsive therapy: assessmentwith a new self-rating instrument. Biol Psychiatry 14:791-801.

Squire LR, Zouzounis JA. (1986) ECT and memory: briefpulse versus sine wave. Am J Psychiatry 743:596-601.

Srinivasaraghavan A, Abrams R. (1996) Court-ordered

P.305

electroconvulsive therapy: The Illinois experience. Procllt h World Congress on Medical Law, Vol 2, pp. 405-15.

Stack CG, Abernethy MH, Thacker M. (1988) Atracuriumfor ECT in plasma cholinesterase deficiency. Compr


Stain-Malmgren R, Tham A, Aberg-Wistedt A. (1998)Increased platelet 5-HT2 receptor binding afterelectroconvulsive therapy in depression. J ECT 14: 15-24.

Staton RD, Enderle JD, Gerst JW. (1988) Theelectroencephalographic pattern during electroconvulsivetherapy: V. Observations on the origin of phase III deltaenergy and the mechanism of action of ECT. Clin

Electroencephalogr 19: 176-98.

Staton RD, Hass PJ, Brumback RA. (1981)Electroencephalographic recording during bitemporal andunilateral non-dominant hemisphere (Lancaster position)electroconvulsive therapy. J Clin Psychiatry 42:264-9.

Stenfors C, Theodorsson E, Mathe AA. (1989) Effect ofrepeated electroconvulsive treatment on regionalconcentrations of tachykinins, neurotensin, vasoactiveintestinal peptide, neuropeptide Y, and galanin in ratbrain. J Neurosci Res 24:445-50.

Stern L, Dannon PN, Hirschmann s et al. (1999)Aminophylline increases seizure length duringelectroconvulsive therapy. J ECT 15:252-1.

Stieper DR, Williams M, Duncan CP. (1951) Changes inimpersonal and personal memory followingelectroconvulsive therapy. J Clin Psychol 7:361-6.

Stoker MJ, Spencer CM, Hamilton M. (1981) Blood-pressure elevation during ECT and associated cognitivedeficit. In: RL Palmer, ed. Electroconvulsive Ther apy:

An Appraisal . Oxford: Oxford University Press, pp. 106-12.

Stone AA. (1977) Recent mental health litigation: Acritical perspective. Am J Psychiatry 134:213-9.

Stoudemire A, Knos G, Gladson M, Markwalter H, SungYF, Morris R, Cooper R. (1990) Labetalol in the controlof cardiovascular responses to electroconvulsive therapyin high-risk depressed medical patients. J Clin Psychiatry

51: 508-12.

Stromgren LS. (1988) Electroconvulsive therapy inAarhaus, Denmark, in 1984: its application innondepressive disorders. Convuls Ther 4:306-13.

Strain JJ, Bidder TG. (1971) Transient cerebralcomplication associated with multiple monitoredelectroconvulsive therapy. Diseases of the Nervous

System 32: 95-100.

Strain JJ, Brunschwig L, Duffy JP, Agle DP, RosenbaumAL, Bidder TG. (1968) Comparison of therapeutic effectsand memory changes with bilateral and unilateral ECT.Am J Psychiatry 725:294-304.

Strober M, Rao U, DeAntonio M. (1998) Effects ofelectroconvulsive therapy in adolescents with severeendogenous depression resistant to pharmacotherapy.Biol Psychiatry 43:335-8.

Stromgren LS (1997) ECT in acute delirium and relatedclinical states. Convuls Ther 73:10-7.

Summers WK, Robins E, Reich T. (1979) The natural

P.306

history of acute organic mental syndrome after bilateralelectroconvulsive therapy. Biol Psychiatry 14: 905-12.

Summerskill J, Seeman W, Meals DW. (1952) Anevaluation of post-electroshock confusion with the Reiterapparatus. Am J Psychiatry 708:835-8.

Sundlblom DM, Heikman P, Naukkarinen H et al. (1999)Blood concentrations of vasopressin, neuropeptide FFand prolactin are increased by high-dose unilateral ECT.Peptides 20:319-26.

Suppes T, Webb A, Carmody T et al. (1996) Is postictalelectrical silence a predictor of response toelectroconvusive therapy? J Affect Dis 41:55-8.

Sutherland EM, Oliver JE, Knight DR. (1969) E.E.G.,memory and confusion in dominant, non-dominant andbi-temporal ECT. Br J Psychiatry 115: 1059-64.

Swartz C, Abrams R. (1984) Prolactin levels afterbilateral and unilateral ECT. Br J Psychiatry 144:643-5.

Swartz CM. (1985) The time course of post-ECT prolactinlevels. Convuls Ther 7:81-8.

Swartz CM. (1989a) Safety and ECT stimulus electrodes:I. Heat liberation at the electrode-skin interface.Convuls Ther 5:171-5.

Swartz CM. (1991a) Quantity of prolactin released byECT seizure [letter]. Convuls Ther 7:63.

Swartz CM. (1991b) Electroconvulsive therapy-inducedprolactin release as an epiphenomenon. Convuls Ther

7:85-91.

Swartz CM. (1993a) Editorial: Beyond seizure duration asa measure of treatment quality. Convuls Ther 9:1-7.

Swartz CM. (1993b) Clinical and laboratory predictors ofECT response. In: CE Coffey, ed. The Clinical Science of

Electroconvulsive Therapy . Washington DC: AmericanPsychiatric Press, pp. 53-72.

Swartz CM. (1994a) Asymmetric bilateral rightfrontotemporal left frontal stimulus electrode placementfor electroconvulsive therapy. Neuro-psychobiology 29:174-8.

Swartz CM. (1994b) Electroconvulsive therapy (ECT)stimulus charge rate and its efficacy. Ann Clin Psychiatry

6:205-6.

Swartz CM. (1995) Setting the ECT stimulus. Psych

Times, June, pp. 33-4.

Swartz CM. (1996) Case report: Disconnection ofelectroencephalographic, motoric, and cardiac evidenceof ECT seizure. Convuls Ther 72:25-30.

Swartz CM. (1997) Flawed model of caffeine effects onelectroconvulsive therapy. EM Psychiatry 47:991-2.

Swartz CM. (2000) Physiological response to ECTstimulus dose. Psych Res 97: 229-35.

Swartz CM. (2001a) Inclusion of patients who cannotrespond to ECT blocks observation of EEC findings[letter]. J ECT 77:153-6.

Swartz CM. (2001b) Stimulus dosing in electroconvulsivetherapy and the threshold multiple method [editorial]. JECT 77:87-90.

P.307

Swartz CM, Abrams R. (1986) An auditory representationof ECT-induced seizures. Convuls Ther 2:125-8.

Swartz CM, Abrams R. (199la) Electroconvulsive therapyapparatus and method for monitoring patient seizures.U.S. Patent Application 07/698,304.

Swartz CM, Abrams R. (2001) Electroconvulsive therapytesting and training device. U.S. Patent 6,200,331.

Swartz CM, Abrams R, Lane RD, DuBois MA,Srinivasaraghavan J. (1994b) Heart rate differencesbetween right and left unilateral ECT. J Neural

Neurosurg Psychiatry 57:97-9.

Swartz CM, Abrams R, Rasmussen K, Pavel J, ZorumskiCF, Srinivasaraghavan J. (1994a) Computer automatedversus visually determined electroencephalo graphic andelectromyographic seizure duration. Convuls Ther

70:165-70.

Swartz CM, Evans CM. (1996) Beyond bitemporal andright unilateral electrode placements. Psych Ann 26:705-8.

Swartz CM, Inglis AE. (1990) Blood pressure reductionwith ECT response. J Clin Psychiatry 57:414-6.

Swartz CM, Larson G. (1986) Generalization of theeffects of unilateral and bilateral ECT. Am J Psychiatry

743:1040-1.

Swartz CM, Larson G. (1989) ECT stimulus duration andits efficacy. Ann Clin Psychiatry 7:147-52.

Swartz CM, Lewis RK. (1991) Theophylline reversal ofelectroconvulsive therapy (ECT) seizure inhibition.Psychosomatics 32:47-51.

Swartz CM, Manly DT. (2000) Efficiency of the stimuluscharacteristics of ECT. Am J Psych 757:1504-6.

Swartz CM, Mehta R. (1986) Double electroconvulsivetherapy for resistant depression. Convuls Ther 2:55-7.

Swartz CM, Saheba N. (1990) Dose effect ondexamethasone suppression testing withelectroconvulsive therapy. Ann Clin Psychiatry 2:183-8.

Swartz CM, Saheba NC. (1989) Comparison of atropinewith glycopyrrolate for use in ECT. Convuls Ther 5:56-60.

Swift MR, LaDu BN. (1966) A rapid screening test foratypical cholinesterase. Lan cet 7:513-74.

Szirmai I, Boldizsar F, Fischer J. (1975) Correlationbetween blood gases, glycolytic enzymes and EEG duringelectroconvulsive treatment in relaxation. Acta Psychiatr

Scand 57:171-81.

Tancer ME, Evans DL. (1989) Electroconvulsive therapyin geriatric patients undergoing anticoagulation therapy.Convuls Ther 5:102-9.

Tancer ME, Golden RN, Ekstrom RD, Evans DL. (1989)Use of electroconvulsive therapy at a university hospital:1970 and 1980-81. Hasp Community Psychiatry 40:64-8.

Tancer ME, Pedersen CA, Evans DL. (1987) ECT andanticoagulation. Convuls Ther 3:222-7.

Tarachow S, Fink M. (1953) A statistical study of apsychoanalytichypothesis: absence of a parent as aspecific factor determining choice of neurosis. J Hill side

Hospital 2:67-71.

Tauboll E, Gjerstad L, Stokke KT, Lundervold A, Telle B.(1987) Effects of electroconvulsive therapy (ECT) onthyroid function parameters. Psychoneu roendocrinology

72:349-54.

Taylor JR, Kuhlengel BG, Dean RS. (1985) ECT, bloodpressure changes and neuropsychological deficit. Br J


Taylor MA. (1982) Indications for electroconvulsivetherapy. In: R Abrams, WB Essman, eds.Electroconvulsive Therapy: Biological Foundations and

Clini cal Applications . New York: Spectrum Publications,pp. 7-40.

Taylor MA, Abrams R. (1975) Acute mania: Clinical andgenetic study of responders and non-responders totreatments. Arch Gen Psychiatry 32:863-5.

Taylor MA, Abrams R. (1978) The prevalence ofschizophrenia: A reassessment using modern diagnosticcriteria. Am J Psychiatry 735:945-8.

Taylor MA, Abrams R. (1985) Short-term cognitiveeffects of unilateral and bilateral ECT. Br J Psychiatry

746:308-11.

Taylor P, Fleminger JJ. (1980) ECT for schizophrenia.Lancet 7:1380-2.

Taylor PJ, von Witt RJ, Fry AH. (1981) Serum creatinephosphokinase activity in psychiatric patients receivingelectroconvulsive therapy. J Clin Psychiatry 42: 103-5.

P.308

Tchou PJ, Piasecki E, Gutmann M, Jazayeri M, Axtell K,Akhtar M. (1989) Psychological support and psychiatricmanagement of patients with automatic implantablecardioverter defibrillators. Int J Psychiatry Med 79:393-407.

Tewfik GI, Wells BG. (1957) The use of Arfonad for thealleviation of cardiovascular stress following electro-convulsive therapy. J Ment Sci 70:636-44.

Thenon J. (1956) Electrochoque monolateral. Acta

Neuropsiquiatria Argentina 2: 292-6.

Thiagarajan AB, Gleiter CH, Nutt DN. (1988)Electroconvulsive shock does not increase plasma insulinin rats. Convuls Ther 4:292-6.

Thienhaus OJ, Margletta S, Bennett JA. (1990) A study ofthe clinical efficacy of maintenance ECT. J Clin


Thomas J, Reddy B. (1982) The treatment of mania. Aretrospective evaluation of the effects of ECT,chlorpromazine, and lithium. J Affect Disord 4:85-92.

Thompson JW, Weiner RD, Myers CP. (1994) Use of ECTin the United States in 1975, 1980, and 1986. Am J


Thorell JI, Adielsson G. (1973) Antidepressive effects ofelectroconvulsive therapy and thyrotrophin-releasinghormone. Lancet 2:43.

Thornton JE, Mulsant BH, Dealy R, Reynolds CF. (1990)A retrospective study of maintenance electroconvulsivetherapy in a university-based psychiatric practice.


Thorogood M, Cown P, Mann J, Murphy M, Vessey M.(1992) Fatal myocardial infarction and use ofpsychotropic drugs in young women. Lancet 340' . 1067-70.

Thorpe JG. (1959) Learning ability during a course of 20electroshock treatments. J Ment Sci 705:1017-21.

Tormos JM, Pascual AP, Catala MD et al. (inpreparation): Antidepressant effects of repetitivetranscranial magnetic stimulation to the dorsolateralpreforntal cortex: I. Hemispheric asymmetry.

Tortella FC, Long JB, Hong J et al. (1989) Modulation ofendogenous opioid systems by electroconvulsive shock.Convuls Ther 5:261-73.

Triggs WJ, McCoy KJM, Greer R et al. (1999) Effects ofleft frontal transcranial magnetic stimulation ondepressed mood, cognition, and corticomotor threshold.Biol Psychiatry 45:1440-6.

Troup PJ, Small JG, Milstein V, Small IF, Zipes DP.(1978) Effect of electroconvulsive therapy on cardiacrhythm, conduction and repolarization. PACE 1: 171-7.

Tsuang MT, Tidball JS, Geller D. (1979) ECT in adepressed patient with shunt in place for normalpressure hydrocephalus. Am J Psychiatry 736:1205-6.

Turner TH, Ur E, Grossman A. (1987) Naloxone has noeffect on hormonal responses to ECT in man. Psychiatry

Res 22:207-12.

Umlauf CW, Gunter RC, Tunnicliffe WW. (1951)Impedance of the human head as observed during

P.309

electro-shock treatment. Confinia Neurologica 77:129-38.

Ungerleider JT. (1960) Acute myocardial infarction andelectroconvulsive therapy. Diseases of the Nervous

System 27:149-53.

Usubiaga JE, Gustafson W, Moya F, Goldstein B. (1967)The effect of intravenous lignocaine on cardiacarrhythmias during electroconvulsive therapy. Compr


Vaidya VA, Siuciak JA, Du F, Duman RS. (1999)Hippocampal mossy fiber sprouting induced by chronicelectroconvulsive seizures. Neuroscience 89:157-66.

Vaidya VA, Terwilliger RZ, Duman RS. (2000) Alterationsin heavy and light neurofilament proteins in hippocampusfollowing chronic ECS administration. Synapse 35:137-43.

Valentine M, Keddie KM, Dunne D. (1968) A comparisonof techniques in electroconvulsive therapy. Br J


Varma NK, Lee SI. (1992) Nonconvulsive statusepilepticus following electroconvulsive therapy [letter].Neurology 42:263-4.

Varma SL, Lai N, Trivedi JK, Anand M. (1988) Post-dexamethasone plasma cortisol levels in depressivepatients receiving electro-convulsive therapy. Indian J

Med Res 57:86-91.

Videbech P. (1997) MRI. findings in patients withaffective disorder: A meta-analysis. Acta Psychiatr Scand

96:157-68.

Viguera A, Rordorf G, Schouten R et al. (1988)Intracranial haemodynamics during attenuated responsesto electroconvulsive therapy in the presence of anintracerebral aneurysm. J Neural Neurosurg Psychiatry

64:802-5.

Villalonga A, Bernardo M, Gomar C, Fita G, Escobar R,Pacheco M. (1993) Cardiovascular response andanesthetic recovery in electroconvulsive therapy withpropofol or thiopental. Convuls Ther 9:108-11.

Villalonga A, Planella T, Castillo J, Hernandez C, CabrerC, Manalich M, Towas A, Nalda MA. (1989) Nitroglicerinaen nebulizador en la profilaxis de la hipertensioninducida por la terapia electroconvulsiva. Rev Esp

Anestesiol Reanim 36:264-6.

Viparelli U, Di Lorenzo R, Capasso G, Manieri L, SciorioG, Viparelli G. (1976) Trattamento con e shock diinserma psicotica gia operata di commissurotomiamitralica. Ospedale Psichiatrico 5:1-10.

Viparelli U, Viparelli G. (1992) ECT and grand malepilepsy. Convuls Ther 8: 39-42.

Vitkun SA, Boccio RV, Poppers PJ. (1990) Anestheticmanagement of a patient with neuroleptic malignantsyndrome. J din Anesth 2:188-91.

Volavka J, Feldstein S, Abrams R, Fink M. (1972) EEGand clinical change after bilateral and unilateralelectroconvulsive therapy. Electroencephalogr Clin

Neurophysiol 32:631-9.

Volkow ND, Bellar S, Mullani N, Jould L, Dewey S. (1988)Effects of electroconvulsive therapy on brain glucosemetabolism: A preliminary study. Convuls Ther 4:199-205.

P.310

Walker R, Swartz CM. (1994) Electroconvulsive therapyduring high-risk pregnancy. Gen Hasp Psychiatry

76:348-53.

Walter G. (1998) Portrayal of ECT in movies fromAustralia and New Zealand. J ECT 14:56-60.

Walter G, Koster K, Rey JM. (1999a) Electroconvulsivetherapy in adolescents: experience, knowledge, andattitudes of recipients. J Am Acad Child Adolesc


Walter G, Koster K, Rey JM. (1999b) Views abouttreatment among parents of adolescents who receivedelectroconvulsive therapy. Psychiatr Serv 50: 701-2.

Warmflash VL, Stricks L, Sackeim HA, Decina P, NelleyP, Malitz S. (1987) Reliability and validity of measuresof seizure duration. Convuls Ther 3:18-25.

Warnecke L. (1975) A case of manic -depressive illness inchildhood. Can Psychiatr Assoc J 20:195-200.

Warren AC, Holroyd S, Folstein MF. (1989) Majordepression in Down's syndrome. Br J Psychiatry

155:202-5.

Wassermann EM. (1998) Risk and safety of repetitivetranscranial magnetic stimulation: report and suggestedguidelines from the International Workshop on theSafety of Repetitive Transcranial Magnetic Stimulation,June 5-7, 1996. Electroencephalogr Clin Neurophysiol

108:1â!” 16.

Watterson D. (1945) The effect of age, head resistance,

and other physical factors on the stimulus threshold ofelectrically induced convulsions. J Neurol Neu rosurg

Psychiatry 8:121-5.

Weatherly J, Villien LM. (1958) Treatment of a case ofpsychotic depression complicated by aortic homographreplacement. Am J Psychiatry 7/4:1120-1.

Weaver L, Williams R, Rush S. (1976) Current density inbilateral and unilateral ECT. Biol Psychiatry 77:303-12.

Weaver LA Jr, Ives J, Williams R. (1982) Studies inbrief-pulse electroconvulsive therapy: The voltagethreshold, interpulse interval, and pulse polarityparameters. Biol Psychiatry 777:131-43.

Weaver LA, Jr, Williams RW. (1982) Theelectroconvulsive therapy stimulus. In: R Abrams, WBEssman, eds. Electroconvulsive Therapy: Biological

Founda tions and Clinical Applications . New York:Spectrum Publications, pp. 129-56.

Webb MC, Coffey CE, Saunders WR, Cress MM, WeinerRD, Sibert TR. (1990) Cardiovascular response tounilateral electroconvulsive therapy. Biol Psychi atry

28:758-66.

Wechsler D. (1945) A standardized memory scale forclinical use. J Psychol 19: 87-95.

Weeks D, Freeman CP, Kendell RE. (1980) ECT: III:Enduring cognitive deficits? Br J Psychiatry 737:26-37.

Weil-Malherbe H. (1955) The effect of convulsive therapyon plasma adrenaline and noradrenaline. J Ment Sci

707:156-62.

Weinberger DR, Torrey EF, Neophytides AN, Wyatt RJ.

(1979) Structural abnormalities in the cerebral cortex ofchronic schizophrenic patients. Arch Gen Psychiatry

36:935-6.

Weiner RD. (1980a) ECT and seizure threshold: Effectsof stimulus wave form and electrode placement. Biol


Weiner RD. (1980b) The persistence of electroconvulsivetherapy-induced changes in the electroencephalogram. JNerv Ment Dis 168:224-8.

Weiner R. (1980c) Review of Breggin P. Electroshock: Its

Brain Disabling Effects. Am J Psychiatry 737:1144.

Weiner RD. (1981) ECT-induced status epilepticus andfurther ECT: A case report. Am J Psychiatry 738:1237-8.

Weiner RD. (1982) Electroencephalographic correlates ofECT. Psychopharmacol Bull 78:78-81.

Weiner RD. (1983b) ECT in the physically ill. J Psychiatr

Treat Eval 5:457-62.

Weiner RD. (1984) Does electroconvulsive therapy causebrain damage? Behavioral and Brain Sciences 7:1-53.

Weiner RD. (1988) The first ECT devices. Convuls Ther

4:50-61.

Weiner RD, Coffey CE. (1986) Differential ability of ECTdevices to produce seizures [letter]. Convuls Ther

2:134-5.

Weiner RD, Coffey CE. (1988) Constant current vsconstant voltage ECT devices [letter]. Br J Psychiatry

752:292-3.

P.311

Weiner RD, Coffey CE. (1989) Comparison of brief-pulseand sine wave stimuli [letter]. Convuls Ther 5:184-5.

Weiner RD, Coffey CE, Krystal AD. (1991) The monitoringand management of electrically induced seizures.Psychiatr Clin North Am 74:845-69.

Weiner RD, Rogers HJ, Davidson JRT, Kahn EM. (1986a)Effects of electroconvulsive therapy upon brain electricalactivity. Ann N Y Acad Sci 462: 270-81.

Weiner RD, Rogers HJ, Davidson JRT, Squire LR. (1986b)Effects of stimulus parameters on cognitive side effects.Ann N Y Acad Sci 462:315-25.

Weiner RD, Volow MR, Gianturco DT, Cavenar JO, Jr.(1980a) Seizures terminable and interminable with ECT.Am J Psychiatry 737:1416-8.

Weiner RD, Whanger AD, Erwin CW, Wilson WP. (1980b)Prolonged confusional state and EEG seizure activityfollowing concurrent ECT and lithium use. Am J


Weingartner H, Gold P, Ballenger JC, Smallberg SA,Summers R, Rubinow DR, Post RM, Goodwin FK. (1981)Effects of vasopressin on human memory functions.Science 277:601-3.

Weinger MB, Partridge BL, Hauger R, Mirow A. (1991)Prevention of the cardiovascular and neuroendocrineresponse to electroconvulsive therapy: I. Effectivenessof pretreatment regimens on hemodynamics. Anesth

Analg 73: 556-62.

Weinstein MR, Fischer A. (1967) Electroconvulsive

treatment of a patient with artificial mitral and aorticvalves. Am J Psychiatry 723:882-4.

Weiss DM. (1955) Changes in blood pressure withelectroshock therapy in a patient receivingchlorpromazine hydrochloride (Thorazine). Am J


Weiss SRB, Li X-L, Heynen T et al. (1997) Kindling andquenching: conceptual links to rTMS. CNS Spectrums

2:32-68.

Weizman A, Gil -Ad 1, Grupper D, Tyano S, Laron Z.(1987) The effect of acute and repeatedelectroconvulsive treatment on plasma beta-endorphin,growth hormone, prolactin and cortisol secretion indepressed patients. Psychophar macology 93:122-6.

Welch CA. (1982) The relative efficacy of unilateralnondominant and bilateral stimulation. Psychopharmacol

Bull 18:68-10.

Welch CA, Drop LJ. (1989) Cardiovascular effects of ECT.Convuls Ther 5:35-43.

Wells DG, Davies GG. (1987) Hemodynamic changesassociated with electroconvulsive therapy. Anesth Analg

66:1193-5.

Wells DG, Davies GG, Rosewarne F. (1989) Attenuationof electroconvulsive therapy induced hypertension withsublingual nifedipine. Anesthesia Intensive Care 77:31-3.

Wells DG, Zelcer J, Treadrae C. (1988) ECT-inducedasystole from a subconvulsive shock. Anesthesia

Intensive Care 16:368-71.

West ED. (1981) Electric convulsion therapy in

P.312

depression: A double-blind controlled trial. Br Med J

282:355-1.

Whalley LJ, Eagles JM, Bowler GM, Bennie JG, Dick HR,McGuire RJ, Fink G. (1987) Selective effects of ECT onhypothalamic-pituitary activity. Psychol Med 77:319-28.

Whalley LJ, Rosie R, Dick H, Levy G, Watts AG, ShewardWJ, et al. (1982) Immediate increases in plasmaprolactin and neurophysin but not other hormones afterelectroconvulsive therapy. Lancet 2:1064-8.

White RK, Shea JJ, Jonas MA. (1968) Multiple monitoredelectroconvulsive treatment. Am J Psychiatry 725:622-6.

Widepalm K. (1987) Comparison of fronto-frontal andtemporo-parietal unilateral non-dominant ECT. Aretrograde memory study Acta Psychiatr Scand 75: 441-4.

Widerlov E, Ekman R, Jensen L, Borglund L, Nyman K.(1989) Arginine vasopressin, but not corticotropinreleasing factor, is a potent stimulator of adrenocorticotropic hormone following electroconvulsivetreatment. J Neural Trans 75:101-9.

Wilcox KW. (1955) The pattern of cognitive reorientationfollowing loss of consciousness. Papers of the Michigan

Academy of Science, Art and Letters 41: 357.

Willerson JT. (1982) Acute myocardial infarction. In: JBWyngaarden, LH Smith, eds. Textbook of Medicine (16thed.). Philadelphia: WB Saunders pp. 247-5.

Williams KM, lacono WG, Remick RA, Greenwood P.(1990) Dichotic perception and memory followingelectroconvulsive treatment for depression. Br J


Williams M. (1966) Memory disorders associated withelectroconvulsive therapy. In: CWM Whitty, OL Zangwill,eds. Amnesia. London: Butterworths, pp. 139-49.

Wilson 1C, Gottlieb G. (1967) Unilateralelectroconvulsive shock therapy. Diseases of the Nervous

System 28:541-5.

Wilson 1C, Vernon JT, Guin T, Sandifer MG. (1963) Acontrolled study of treatments of depression. JNeuropsychiatry 4:331-7.

Wilson WP, Schieve JF, Durham NC, Scheinberg P.(1952) Effect of series of electric shock treatments oncerebral blood flow and metabolism. Archives of Neu

rology Psychiatry 68:651-4.

Windebank AJ, Bailey PA, Bell GN, Blexrud MD,Charmichael SW. (1990) Adrenal medullarytransplantation into the brain for treatment ofParkinson's disease: clinical outcome and neurochemicalstudies. Mayo Clin Proc 65:3052-8.

Windle WF, Groat RA, Fox CA. (1944) Experimentalstructural alterationsin brain during and afterconcussion. Surg Gynec Obstetr 79:561-72.

Windle WF, Krieg WJS, Arieff AJ. (1945) Failure to detectstructural changes in the brain after electrical shock.Quart Bull Northwestern U Med Schl 79:181-8.

Winkelman NW, Moore MT. (1944) Neurohistopathologicfindings in experimental electric shock treatment. JNeuropathol Exp Neural 3:199-209.

Winslade WJ, Liston EH, Ross JW, Weber KD. (1984)

P.313

Medical, judicial, and statutory regulation of ECT in theUnited States. Am J Psychiatry 141: 1349-55.

Wise MG, Ward SC, Townsend-Parchman W, Gilstrap LC,III, Hauth JC. (1984) Case report of ECT during high-risk pregnancy. Am J Psychiatry 141: 99-101.

Wolford JA. (1957) Electroshock therapy and aorticaneurysm. Am J Psychiatry 113:656.

Woodruff RA, Pitts FM, Jr, McClure JN, Jr. (1968) Thedrug modification of ECT. Arch Gen Psychiatry 78:605-11.

Wyant GM, MacDonald WB. (1980) The role of atropine inelectroconvulsive therapy. Anaesthesia Intensive Care

8:445-50.

Yalow RS, Varsano-Aharon N, Echemendia E, Berson SA.(1969) HGH and ACTH secretory responses to stress.Harm Metab Res 7:3-8.

Yassa R, Hoffman H, Canakis M. (1990) The effect ofelectroconvulsive therapy on tardive dyskinesia: Aprospective study. Convuls Ther 6:194-8.

Yatham LN, Clark CC, Zis AP. (2000) A preliminary studyof the effects of electroconvulsive therapy on regionalbrain glucose metabolism in patients with majordepression. J ECT 76:171-6.

Yoshinaga H, Kobayashi K, Sato M et al. (1996)Investigation of bilateral synchronous spike-wavedischarge by EEG topography. Brain Topogr 8:255-60.

Youmans CR, Jr, Bourianoff G, Allensworth DC, Martin

WL, Derrick JR. (1969) Electroshock therapy and cardiacpacemakers. Am J Surg 778:931-7.

Young RC, Alexopoulos GS, Shamoian CA. (1985)Dissociation of motor response from mood and cognitionin a parkinsonian patient treated with ECT. Biol


Zachrisson OC, Balldin J, Ekman R et al. (2000) Noevident neuronal damage after electroconvulsive therapy.Psychiatry Res 96:157-65.

Zamora EW, Kaelbling R. (1965) Memory andelectroconvulsive therapy. Am J Psychiatry 722:546-54.

Zamrini EY, Meador KJ, Loring DW, Nichols FT, Lee GP,Figueroa RE et al. (1990) Unilateral cerebral inactivationproduces differential left/right heart responses.Neurology 40:1408-11.

Zeidenberg P, Smith R, Greene L, Malitz S. (1976)Psychotic depression in a patient with progressivemuscular dystrophy: treatment with multiple monitoredelectroconvulsive therapy. Diseases of the Nervous

System 37:21-3.

Zhu W-X, Olson DE, Karon BL, Tajik AJ. (1992) Ann

Intern Med 777:914-5.

Zibrak JD, Jensen WA, Bloomingdale K. (1988) Aspirationpneumonitis following electroconvulsive therapy inpatients with gastroparesis. Biol Psychiatry 24: 812-4.

Zielinkski RJ, Roose SP, Devanand DP, Woodring S,Sackeim HA. (1993) Cardiovascular complications of ECTin depressed patients with cardiovascular disease. Am J


Zimmerman M, Coryell W, Stangl D, Pfohl B. (1986) AnAmerican validation study of the Newcastle scale. III.Course during index hospitalization and six -monthprospective follow-up. Acta Psychiatr Scand 73:412-5.

Zimmerman M, Pfohl B, Coryell W, Stangl D. (1987) Theprognostic validity of DSM-III Axis IV in depressedpatients. Am J Psychiatry 744:102-6.

Zinkin D, Birtchnell J. (1968) Unilateral electroconvulsivetherapy: Its effects on memory and its therapeuticefficacy. Br J Psychiatry 774:973-88.

Zis A, McGarvey KA, Clark CM, Lam RW, Patrick L,Adams SA. (1993) Effect of stimulus energy onelectroconvulsive therapy-induced prolactin release. Con

vuls Ther 9:23-7.

Zis AP, Goumeniok AD, Clark CM, Grant BEK, Remick RA,Lam RW, Garland EJ. (1991) ECT-induced prolactinrelease: effect of sex, electrode placement and serotoninuptake inhibition. Hum Psychopharmacology 6:155-60.

Zis AP, Manji HK, Remick RA, Grant BEK, Clark CM.(1989a) Effect of the 5HT2 antagonist ketanserin on theECT-induced prolactin release. Biol Psychiatry 26:102-6.

Zis AP, McGarvey KA, Clark CM, Lam RW, Adams SA.(1992) The role of dopamine in seizure-induced prolactinrelease in humans. Convuls Ther 8: 126-30.

Zis AP, Yatham LN, Lam RW et al. (1996) Effect ofstimulus intensity on prolactin and cortisol releaseinduced by unilateral electroconvulsive therapy. Neuro

psychopharm 75:263-70.

Zorumski CF, Burke WJ, Rutherford JL, Reich T. (1986)ECT: Clinical variables, seizure duration, and outcome.


Zung WW, Rogers J, Krugman A. (1968) Effect ofelectroconvulsive therapy in memory in depressivedisorders. Rec Adv Biol Psychiatry 70:160-78.

Zwil AS, Bowring MA, Price TRP, Goetz KL, Greenbarg JB,Kane-Wagner G. (1990) Prospective electroconvulsivetherapy in the presence of intracranial tumor. Convuls

Ther 6:299-307.

electroconvulsive therapy.9780195148206.29298

Documents