prescribing for elderly patients || clinical pharmacology of ageing

1 Clinical pharmacology of ageing

Arduino Mangoni1, Paul Jansen2 andStephen Jackson3

1Department of Clinical Pharmacology, Flinders Medical Centre, Adelaide, Australia2Department of Geriatric Medicine, University Medical Centre Utrecht,The Netherlands3Department of Clinical Gerontology, King’s College Hospital, London, UK

EPIDEMIOLOGY

Trends in population ageing

In 2000, the worldwide population of thoseaged ≥65 years was an estimated 420 mil-lion, a 9.5 million increase from 1999. Duringthe period 2000–2030, the worldwide popula-tion aged ≥65 years is projected to increase byapproximately 550 million to 973 million, in-creasing from 6.9 % to 12.0 % worldwide, from15.5 % to 24.3 % in Europe, from 12.6 % to20.3 % in North America, and from 6.0 % to12.0 % in Asia. The largest increases in abso-lute numbers of elderly people will occur indeveloping countries.

The ageing of the world’s population is theresult of two factors: declines in fertility andincreases in life expectancy. Life expectancyin developed countries now ranges from 76 to80 years and also has increased in developingcountries since 1950. A higher life expectancyat birth for females compared with males is al-most universal, approximately seven years inEurope and North America but less in develop-ing countries.

The world has experienced an epidemiolog-ical transition in the leading causes of death,from infectious disease and acute illness to

chronic disease and degenerative illness. Devel-oped countries in North America, Europe, andthe Western Pacific already have undergone thistransition, and other countries are at differentstages of progression. The epidemiologicaltransition, combined with the increasing numberof older people, represents a challenge for publichealth. In the United States, approximately 80 %of all people aged ≥65 years have at least onechronic condition, and 50 % have at least two.

The increased number of older peoplewill lead to increased healthcare costs. Thehealthcare cost per capita for those aged ≥65years in developed countries is three to fivetimes greater than the cost for those aged <65years, and the rapid growth in the number ofolder people, coupled with continued advancesin medical technology, is expected to createincreasing pressure on health and long-termcare spending.

Medication use in elderly patients

Safe and effective pharmacotherapy remainsone of the greatest challenges in geriatricmedicine. Elderly patients often suffer fromseveral chronic disorders and consequentlyuse more drugs than any other age group. Thediminished physiological reserve associated

Prescribing for Elderly Patients Edited by Stephen Jackson, Paul Jansen and Arduino Mangoni© 2009 John Wiley & Sons, Ltd. ISBN: 978-0-470-02428-7

2 PRESCRIBING FOR ELDERLY PATIENTS

with ageing can be further depleted by effectsof drugs and acute or chronic disease states.Ageing alters the pharmacokinetics andpharmacodynamics of many drugs. In addition,pharmacotherapy may be complicated bydifficulties with obtaining drugs or complyingwith drug regimens.

In most developed countries, about twothirds of the population ≥65 years takeprescription and nonprescription over thecounter (OTC) drugs. At any given time,an average elderly person uses four to fiveprescription drugs and 2 OTC drugs and fills12–17 prescriptions a year. The frail elderlypatient uses the most drugs. Drug use is greaterin hospitals and nursing homes than in thecommunity; typically, a nursing home residentreceives at least seven to eight drugs.

The type of drug used varies with the setting.Community patients use analgesics, diuretics,cardiovascular drugs, and sedatives most often;nursing home residents use antipsychotics andsedative-hypnotics most commonly, followedby diuretics, antihypertensives, analgesics, car-diac drugs, and antibiotics. Psychoactive drugsare prescribed for ∼65 % of nursing home pa-tients and for ∼55 % of residential care patients;∼7 % of patients in nursing homes receive ≥three psychoactive drugs concurrently.

Many drugs benefit elderly patients. Somecan be life-saving (i.e. antibiotics and throm-bolytic therapy). Oral hypoglycemic agents canimprove independence and quality of life whilecontrolling chronic disease. Antihypertensivedrugs and influenza vaccines can help preventor decrease morbidity. Analgesics and antide-pressants can control debilitating symptoms.Therefore, appropriateness, that is whether thepotential benefits outweigh the potential risks,should guide therapy.

AGE-RELATED CHANGESIN PHARMACOKINETICS

Drug absorption

Pharmacokinetic studies on the effect ofageing on drug absorption have providedconflicting results. Several studies have not

shown age-related differences in absorptionrates for different drugs (1). However, otherstudies have shown a reduced absorption ofvitamin B12, iron and calcium and an increasedabsorption of the drug levodopa. For drugsabsorbed by passive diffusion there is littleevidence for an age related decline.

First-pass metabolismand bioavailability

There is a reduction in first-pass metabolismwith advancing age. This is probably due toa reduction in liver mass and, for high clear-ance drugs, the consequential reduction in bloodflow. The bioavailability of drugs undergoingextensive first-pass metabolism can be signifi-cantly increased. By contrast, the first-pass ac-tivation of several pro-drugs, such as the ACEinhibitors enalapril and perindopril, might beslowed or reduced (2).

Drug distribution

Significant changes in body composition occurwith advancing age, such as a progressivereduction in the proportion of total body waterand lean body mass (3). This results in arelative increase in body fat. Polar drugs thatare mainly water-soluble tend to have smallervolumes of distribution (V) resulting in higherserum levels in older people (e.g. gentamicin,digoxin, lithium, and theophylline). Bycontrast, non-polar compounds (e.g. benzo-diazepines, morphine and amiodarone) tendto be lipid-soluble and so their V increaseswith age. The main effect of the increasedV is a prolongation of half-life. IncreasedV and elimination half life (t1/2z) havebeen observed for drugs such as diazepam,thiopental, lidocaine, and clomethiazole (4).The reduction in V for water-soluble drugstends to be balanced by a larger reduction inrenal clearance (CL), with a smaller effect ont1/2z, as shown in the following equation:

t1/2z= Ln(2) · V

CL

where t1/2z= elimination half-life, Ln(2) =

natural log of 2 (0.693), V = apparent volumeof distribution, and CL = clearance.

CLINICAL PHARMACOLOGY OF AGEING 3

Protein binding

Acidic compounds (e.g. diazepam, phenytoin,warfarin, salicylic acid) bind mainly to albu-min whereas basic drugs (e.g. lidocaine, pro-pranolol) bind to alpha-1 acid glycoprotein.Although no substantial age-related changesin the concentrations of both these proteinshave been observed, albumin is commonly re-duced in malnutrition or acute illness whereasalpha-1 acid glycoprotein is increased duringacute illness. The main factor determining drugeffect is the free concentration of the drug.Although plasma protein binding changes mighttheoretically contribute to drug interactions orphysiological effects for drugs that are highlyprotein-bound, its clinical relevance is probablylimited (5).

Drug clearance

Kidney

The age-related reduction in glomerular filtra-tion rate affects the clearance of many drugssuch as water-soluble antibiotics, diuretics,digoxin, water-soluble beta-blockers, lithium,and some non-steroidal anti-inflammatorydrugs. The clinical importance of such re-ductions of renal excretion is dependent onthe likely toxicity of the drug. Drugs with anarrow therapeutic index like aminoglycosideantibiotics, digoxin, and lithium are likely tohave serious adverse effects if they accumulateonly marginally more than intended. In elderlypatients serum creatinine may be within the ref-erence limits, while renal function is markedlydiminished. The Cockcroft and Gault (6) orthe Modification of Diet in Renal Disease (7)equations both use serum creatinine, age andgender and may be helpful for a better estima-tion of glomerular filtration rate in this situation.

Liver

Drug clearance by the liver depends on the ca-pacity of the liver to extract the drug from theblood passing through the organ (hepatic ex-traction ratio) and hepatic blood flow. Drugscan be classified into three groups according totheir extraction ratio (E): high (E > 0.7, such

as clomethiazole, dextropropoxyphene, glyc-eryl trinitrate, lidocaine, pethidine, and pro-pranolol), intermediate (E 0.3–0.7, such asaspirin, codeine, morphine, and triazolam), andlow extraction ratio (E < 0.3, such as carba-mazepine, diazepam, phenytoin, theophylline,and warfarin). When E is high, the clearanceis rate-limited by blood flow. When E is low,changes in blood flow produce little changesin clearance. Therefore, the reduction in liverblood flow with ageing affects mainly the clear-ance of drugs with a high extraction ratio. Ofmuch greater importance is the reduction inliver volume of as much as 30 % across theadult age range. This results in a reduction inclearance of a similar magnitude (8).

Several studies have shown significantage-related reductions in the clearance ofmany drugs metabolized by phase-1 pathwaysin the liver. These involve reactions such asoxidation and reduction. By contrast, phase-2pathways (e.g. glucuronidation) do not seem tobe significantly affected (8).

AGE-RELATED CHANGESIN PHARMACODYNAMICS

Studies of drug sensitivity require measurementof concentrations of drug in plasma, as well asmeasurement of drug effects. This is becausechanges in pharmacokinetics with increasingage may result in changes in response duepurely to the changes in drug concentrations.Some important pharmacodynamic age-relatedchanges are illustrated in Table 1.1.

Anticoagulants

There is evidence of a greater inhibition of syn-thesis of activated vitamin K-dependent clottingfactors at similar plasma concentrations of war-farin in elderly compared to young patients.However, the exact mechanisms responsible forthe increased sensitivity are unknown.

Cardiovascular and respiratory drugs

Although elderly subjects are less sensitive tothe effects of verapamil on cardiac conduction,


Table 1.1. Selected pharmacodynamic changes with ageing

Drug Pharmacodynamic effect Age-related change

Adenosine Heart-rate response ↔Diazepam Sedation, postural sway ↑Diltiazem, Acute and chronic

antihypertensive effect↑

Verapamil Acute PR interval prolongation ↓Diphenhydramine Postural sway ↔Enalapril ACE inhibition ↔Furosemide Peak diuretic response ↓Heparin Anticoagulant effect ↔Isoprenalin Chronotropic effect ↓Morphine Analgesic effect ↑

Respiratory depression ↔Phenylephrine α1-adrenergic agonism ↔Propranolol Antagonism of chronotropic

effects of isoprenalin↓

Scopolamine(hyoscine)

Cognitive function ↓

Temazepam Postural sway ↑Warfarin Anticoagulant effect ↑Legend: ↑= increase; ↓= decrease; ↔= no significant change; ACE = angiotensin-convertingenzyme.

the effect on blood pressure and heart rate tendsto be greater in older than in younger patients.This might be explained by an increased sen-sitivity to the negative inotropic and vasodila-tor effects of verapamil as well as diminishedbaroreceptor sensitivity. The acute intravenousadministration of diltiazem causes greater pro-longation of the PR interval (dromotropic ef-fect) in young than in elderly subjects.

Reduced β-adrenoreceptor function is ob-served with advancing age. Elderly patientsare less sensitive to the chronotropic effect ofisoprenaline. The impaired response, however,is due primarily to an age-related reduction inthe influence of reflex cardiovascular effectson heart rate rather than reduced β-adrenergicsensitivity. Both salbutamol (β-agonist) andpropranolol (β-antagonist) show reducedresponses with age. This is secondary toimpaired β-receptor function due to reducedsynthesis of cyclic AMP following receptorstimulation. The total number of receptorsseems to be maintained but the post-receptorevents are changed because of alterations ofthe intracellular environment. The responsive-ness of α-adrenoreceptors is preserved withadvancing age.

Psychotropic drugs

Elderly patients are particularly vulnerable toadverse effects of neuroleptics, such as ex-trapyramidal symptoms, arrhythmias, and pos-tural hypotension. Agents with anticholinergiceffects can also impair cognition and orienta-tion in patients with a cholinergic deficit suchas those with Alzheimer’s disease. Advancingage is also associated with increased sensitivityto the central nervous system effects of benzo-diazepines. The exact mechanisms responsiblefor the increased sensitivity to these drugs withageing are unknown, however.

ADVERSE DRUG REACTIONSAND DRUG INTERACTIONS

Adverse drug reactions (ADRs) are an impor-tant cause of morbidity and mortality in elderlypatients. It is not clear however whether ad-vancing age per se is a cause of increased riskof ADRs. Nursing home and frail elderly pa-tients appear to be at high risk of ADRs (9).

Although around 10 % of the general popu-lation take more than one prescribed medicine,


the incidence of combination therapy is greatestin the elderly, in females, and in those whohave had a recent hospital admission. Patientsaged >65 years use on average four pre-scribed medications. A list of common drug

interactions in elderly patients is illustrated inTable 1.2.

The risk of ADRs is exponentially ratherthan linearly related to the number ofmedicines taken. More than 80 % of ADRs

Table 1.2. The most common drug interactions in elderly patients

Drug Drug Impact Mechanism of interaction

Warfarin NSAIDs Potential for seriousgastrointestinalbleeding

NSAIDs increase gastric irritationand erosion of the protectivelining of the stomach anddecrease platelet function duringclot formation

Warfarin Sulfa drugs Increased effects ofwarfarin, withpotential forbleeding

Warfarin’s activity maybeprolonged due to a decreasedproduction of vitamin K byintestinal flora during sulfa drugadministration

Warfarin Macrolides Increased effects ofwarfarin, withpotential forbleeding

Macrolides inhibit the metabolismof warfarin. The activity ofwarfarin may also be prolongeddue to alterations in the intestinalflora and its production ofvitamin K for clotting factorproduction

Warfarin Quinolones Increased effects ofwarfarin, withpotential forbleeding

The exact mechanism for thewarfarin-quinolone druginteraction is unknown.Reduction of intestinal floraresponsible for vitamin Kproduction by antibiotics isprobable as well as decreasedmetabolism of Warfarin

Warfarin Phenytoin Increased effects ofwarfarin and/orphenytoin

Currently unknown, but one theorysuggests a genetic basisinvolving liver metabolism ofwarfarin and phenytoin

ACE inhibitors Potassiumsupplements

Elevated serumpotassium

Inhibition of ACE results indecreased aldosterone productionand potentially decreasedpotassium excretion.

ACE inhibitors Aldosteroneantagonists andpotassiumsparing diuretics

Elevated serumpotassium

Additive effects on reducedpotassium elimination

Digoxin Amiodarone Digoxin toxicity Multiple theories exist, but actualmechanism is unknown.Amiodarone may decrease theclearance of digoxin, resulting inprolonged digoxin half-life.There may also be an additiveeffect on the sinus node activity.

Digoxin Verapamil Bradycardia and heartblock

Synergistic effect on sinus nodeand atrioventricular node

Theophylline Quinolones Theophylline toxicity Inhibition of hepatic metabolism oftheophylline by the quinolones


causing admission or occurring in hospital aretype A, i.e. they are dose related, predictableand potentially avoidable. Antibiotics, anti-coagulants, digoxin, diuretics, hypoglycaemicagents, antineoplastic agents and nonsteroidalanti-inflammatory drugs are mainly responsiblefor type-A ADRs. Type-B ADRs (idiosyncraticreactions) are less common but can beassociated with serious toxicity (e.g. hepato-toxicity with flucloxacillin and co-amoxiclav;anaphylactic shock with penicillins).

ADHERENCE

Although the term compliance has gone out offashion, in practice the three terms compliance,concordance and adherence all refer to the ex-tent to which patients comply with the drugregimen they agreed with the prescriber. Wewill use the word adherence here.

The efficacy and safety of medicines islargely determined by adherence. Adherenceis defined as the extent to which a person’sbehaviour, taking medication, following a diet,and/or executing life-style changes, corre-sponds with recommendations agreed with ahealth-care provider (10). Poor adherence tothe treatment of chronic disease is an importantproblem. One of the first articles pointingat the lack of adherence was published in1957; in only 50 % of the patients who wereprescribed tuberculostatics was the drug wasfound in urine (11). More recently a Cochranereview found 50 % non-adherence in patientsusing medicines for chronic diseases (12).Adherence to antihypertensives and statintherapy is often even lower. Within one yearof the start of antihypertensives 50 % of thepatients have stopped using these drugs (13).The adherence of elderly patients, prescribedstatins, is 60 % after three months, 43 % aftersix months and 26 % after five years (14).

The consequences of non-adherence areconsiderable including hospital admissions(33–60 % of drug related hospital admissions)and higher mortality (15, 16). Even with useof placebo, high adherence had a 3.5 timesgreater effect on reducing mortality than the

overall active treatment with candesartan inchronic heart failure (17). This finding suggeststhat high adherence for taking medicines, isassociated with high adherence for life-styleadvice.

The identification of patient non-adherenceis important. Factors that contribute to poor ad-herence are summarized in Table 1.3 (18). Ingeneral practice non-adherence is often detectedby looking in the medicines cupboard at home.Another method makes use of pharmacy re-fill records comparing the number of dispenseddoses with the number of prescribed doses.A very helpful starting point is to ask the patientand family for the problems they encounteredwith the drug regimen. The patient should notbe blamed for poor adherence. The ability ofpatients to follow treatment plans is frequentlycompromised by several factors, including thecharacteristics of the disease (e.g. cognitive im-pairment), social system, heath care system,economic factors and patient-related factors.A tool for screening patient adherence is theBrief Medication Questionnaire (19). The Med-ication Event Monitoring System (MEMS) is anelectronic device which records the time anddate when a medication container was opened.These devices are used in controlled studies andnot in daily practice, because of the expense.Other methods for detecting non-adherence arephysiological markers, like low heart-rate withuse of beta-blockers, or biochemical measure-ments in blood or urine Such as plasma an-giotensin converting enzyme assays to monitorACEI adherence.

Several methods have been shown to im-prove adherence. The most effective approachis multi-level targeting at several factors withseveral interventions. However effective inter-ventions are often complex and not suitable fordaily practice. Education in self-management ofthe drug regimen has limited effects. A simpleand very effective method is the reduction ofdose frequency. Adherence is found the highestwith a dose frequency of once a day (79 %),decreasing to 69 % with b.i.d., 65 % with t.i.dand 51 % with q.i.d (20).

Integrating the patient’s perspective intotreatment plans is considered very important.


Table 1.3. Methods of measuring adherence (adapted from Ref (18))

Methods Advantages Disadvantages

Directly observed therapy Most accurate Patients can hide pills in mouthand then discard them;impractical for routine use

Biochemical measurement of themedicine or metabolite ormeasurement of a biologicalmarker

Objective Variations in metabolism and“white coat” adherence cangive a false impression;expensive

Patient questionnaires orself-reports

Simple, inexpensive, mostuseful in clinical practice

Susceptible to error anddistortion

Pill counts Objective, quantifiable and easyto perform

Data easily altered by thepatient (e.g. pill dumping)

Rates of prescription refills Objective, easy to obtain data A prescription refill is notequivalent to ingestion ofmedication; requires a closedpharmacy system

Assessment of the patient’s clinicalresponse

Simple; easy to perform Factors other than medicationadherence can affect clinicalresponse

Electronic medication monitors Precise; results are easilyquantified; tracks patterns oftaking medication

Expensive; requires return visits

Measurement of physiologicmarkers

Often easy to perform Marker may be absent for otherreasons

Patient diaries Help to correct for poor recall Easily altered by the patientQuestionnaire for caregiver, for

patients who are cognitivelyimpaired.

Help to correct for poor recall;simple; objective

Susceptible to error anddistortion

The behaviour of prescribers is changingfrom a paternalistic one-way style towardsconcordance to improve adherence (21).

POLYPHARMACY VERSUSAPPROPRIATE PRESCRIBING

Polypharmacy, often defined as the concur-rent use of five or more different drugs, iscommon among elderly patients. Inappropri-ate polypharmacy contributes to unwanted andoften preventable clinically relevant drug-drugand drug-disease interactions as well as adversedrug reactions (ADRs). Around 12 % of elderlypatients in hospitals are admitted because ofADRs (22). It is estimated that about 47–72 %of these ADRs are avoidable (23, 25). Polyphar-macy itself is not necessarily undesirable.

Polypharmacy, however, is a concept thataddresses only the inappropriate use of medica-tion. The term appropriate prescribing addresses

the problems of both inappropriate use of medi-cation as well as inappropriate non use of med-ication (or undertreatment).

OVER-THE-COUNTER MEDICINES

Elderly people are the largest consumers ofover-the-counter (OTC) medicines. The switchof prescription drugs to OTC medicines is en-couraged. Government policies make it possibleto obtain increasing numbers of former pre-scription drugs from pharmacies, health foodshops, supermarkets or by mail-order. The goalis to provide greater choice for individualsand to shift the responsibility for health careas well as the costs to individuals. The ap-proach to OTC medicines varies between coun-tries. In general it is considered that OTC canbe used for short-term self-limiting illnesseswith lower doses of commonly prescribed drugsand should have few important adverse effects.


Some consider herbal drugs, vitamins and min-erals also as OTC drugs.

Older people use OTC medicines to treatminor complaints such as pain, constipation,colds and gastro-intestinal symptoms (25). Themost commonly used OTCs are aspirin, parac-etamol, NSAIDs, antihistamines and histamineH2 receptor antagonists. Recently in severalcountries statins are also available OTC. Thereare concerns regarding the safety of OTCmedicines, especially in elderly patients. In par-ticular, NSAIDs may cause gastrointestinal tox-icity and sedatives, increase may the risk offalls. The use of multiple medications increasesthe risk of drug interactions and adverse effects.Cebollero-Santamaria et al. showed that bleed-ing from a peptic ulcer was associated withthe use of NSAIDs in 81 % of 84 patients andthat 95 % had purchased their NSAIDs OTC(26). The use of recommended doses of OTCNSAIDs has a relatively good safety profilecompared to presciption NSAIDs, however pa-tients may take higher doses for a longer periodwith serious gastrointestinal toxicity as a re-sult (27). Many older people use OTC drugs toimprove their sleep. The risks associated withthis use have not been examined (25). Also theproblems in older people with OTC H2 receptorantagonists, such as confusion, and OTC statins,such as liver and skeletal muscle toxicity, arenot clearly defined.

Documentation of OTC medicines is poor.A study showed that only 5 % of OTC drugs,used by patients prior to and during hospi-talization, were recorded on drug charts (28).Asking elderly patients, especially those admit-ted to hospitals, for their use of OTC drugsis important to prevent double-prescription andclinically relevant interactions. This applies alsoto herbal drugs as St. John’s wort. St. John’swort is used to treat depressive symptoms. Itinduces the metabolism of several drugs and di-minishes the absorption of digoxin (Table 1.4)(29). Other possible interactions are the dimin-ishing effect of warfarin caused by cranberryjuice and Ginseng (30, 31).

The increasing availability of OTC drugsclearly has benefits. Nevertheless, prescribers

Table 1.4. Interactions with St. John’s wort(adapted from Ref (29))

amitriptyline Steady-state concentrationdecreased by 22%

ciclosporine Steady-state concentrationdecreased by 52%

digoxin Steady-state concentrationdecreased by 25%

simvastatin AUC decreased by 50%tacrolimus Steady-state concentration

decreased by 80%theophylline Steady state concentration

decreased by 50%cumarin derivatives INR 50% lower

AUC = Area under plasma concentration time curve

must always pay close attention to concomi-tant OTC medication use in order to minimizeadverse drug reactions.

PRESCRIBING AUDIT

Clinical audit is fundamental to providing ahigh quality service. This is particularly true forprescribing. A variety of approaches have beenadvocated ranging from the use of purely de-scriptive prescribing indicators through applica-tion of consensus guidelines to strictly evidencebased approaches. The use of purely descriptiveapproaches alone achieves little but is a usefuladjunct to comparing observed prescribing to agold standard.

Oborne et al. (1997) have described threetypes of prescribing indicator for use in pre-scribing audit (45). Purely descriptive indica-tors include mean numbers of drugs prescribedand numbers of drugs classified in the BritishNational Formulary as “Black Triangle” drugs(7). These are recently introduced drugs forwhich all adverse events should be reported.Indicators of unnecessary or potentially harm-ful prescribing include duplications such as H2

receptor blockers and proton pump inhibitorsand potentially harmful drugs such as longacting hypoglycaemic agents which should nolonger be used. The third category of indica-tor is evidence based. These indicators mea-sure the extent to which research evidence is


put into practice. Examples include the use ofantithrombotic therapy in atrial fibrillation andthe use of aspirin in coronory artery disease.These approaches are superceding the Beers cri-teria (47) which are associated with a numberof problems.

MEDICATION REVIEW

Medication review is an essential processin the management of patients with chronicdisease. This process should be driven by fourquestions and should involve the patient asmuch as possible:

1. Which drugs are necessaryfor the patient?

The first step should be to look at all theproblems and diseases of the patient and todetermine which drugs are indicated (32) Itis important to identify indicated drugs thatare missing. In making decisions, prescribersshould consider the remaining life expectancy,goals of care and potential benefits of medica-tions (33).

2. Which drugs are not necessary?

The next step is to look at the medicines thepatient uses, including the OTC drugs, and todetermine for which drugs there is still an in-dication. Unnecessary duplications with otherdrugs should be looked for (e.g. H2 blockersand proton pump inhibitors), taking into ac-count drugs that will be added to the regimenas above.

3. Are there better alternativesfor the remaining drugs?

The proposed drugs should be prescribable to-gether. It is important to look at clinically sig-nificant drug-drug and drug-disease interactions(34). It is also important to ask the patient aboutadverse drug reactions and, if so, to look at al-ternatives. Where drugs have similar efficacy/safety profiles the least expensive option shouldbe prescribed.

4. Is the dose and the dose frequencyappropriate?

Consider if the prescribed dose is still correct.Have there been any changes in the clearanceof the drug, e.g. a change in renal function?

The patient should always be asked aboutproblems with drug adherence. Adherence canbe increased in several ways, but most evidenceexists for reduction of the number of dailydoses (20).

UNDERTREATMENT

Undertreatment is a common reason forinappropriate prescribing. It has been shownthat undertreatment is frequent in elderlypatients, despite the use of many medicines(35, 36). Undertreatment in elderly patients isreported in a high percentage of patients withmyocardial infarction, chronic heart failure,atrial fibrillation, hyperlipidemia, osteoporosis,COPD, depression, pain and cancer (35–43).Choudhry et al. concluded that a physician’sexperience with bleeding events associated withwarfarin in patients can cause underprescriptionof warfarin to other patients (39). Kuzuya et al.showed that the incidence of polypharmacyamong frail community-dwelling older peopleis lower in the oldest members (>85 years)because of underuse of medications forchronic diseases (43). Kuipers et al. founda clear relationship between polypharmacyand underprescription (32). The probability ofunderprescription increased significantly withthe number of medicines (Figure 1.1).

It appears that general practitioners and spe-cialists are not willing to prescribe more drugsto frail old patients with current polypharmacyfor reasons of complexity of drug regimens,fear of ADRs, interactions and poor adher-ence. Research has shown that for some medicalproblems a so-called treatment-risk paradox orrisk-treatment mismatch exists meaning that pa-tients who are at highest risk for complicationshave the lowest probability to receive the rec-ommended pharmacological treatment (38, 40).The application of clinical practice guidelines


1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0Est

imat

ed p

roba

bilit

y of

und

ertr

eatm

ent

0 2 4 6 8 12 14 16 1710

Number of drugs used

Figure 1.1. Estimated probability of underprescrip-tion related to the number of drugs (adapted fromRef (32))

(CPGs) to the care of older patients with severalcomorbid diseases may have undesirable effectsand there could be reasons not to treat all prob-lems. Moreover, the evidence of the benefit ofCPG application in elderly patients with comor-bid disease is lacking. Boyd et al. estimated thatif the relevant CPGs were followed a hypothet-ical patient would be prescribed 12 medications(44). However, undertreatment may be harmfulfor the patient. In optimizing polypharmacy, at-tention should be directed not only to overtreat-ment but also to possible undertreatment. Theaim is to enhance appropriate prescribing to pa-tients with comorbid diseases. In conclusion,undertreatment is considered an important prob-lem in elderly patients with comorbid diseasesalthough the evidence base is lacking.

REFERENCES

1. Gainsborough N, Maskrey VL, Nelson ML,Keating J, Sherwood RA, Jackson SH et al. Theassociation of age with gastric emptying. AgeAgeing 1993; 22(1): 37–40.

2. Wilkinson GR. The effects of diet, aging anddisease-states on presystemic elimination andoral drug bioavailability in humans. Adv DrugDeliv Rev 1997; 27(2-3): 129–59.

3. Fulop T, Jr., Worum I, Csongor J, Foris G,Leovey A. Body composition in elderly people.I. Determination of body composition by mul-tiisotope method and the elimination kinetics

of these isotopes in healthy elderly subjects.Gerontology 1985; 31(1): 6–14.

4. Klotz U, Avant GR, Hoyumpa A, Schenker S,Wilkinson GR. The effects of age and liverdisease on the disposition and elimination ofdiazepam in adult man. J Clin Invest 1975;55(2): 347–59.

5. Benet LZ, Hoener BA. Changes in plasma pro-tein binding have little clinical relevance. ClinPharmacol Ther 2002; 71(3): 115–21.

6. Levey AS, Bosch JP, Lewis JB, Graeme T,Rogers N, Roth D. A more accurate method toestimate glomerular filtration rate from serumcreatinine: a new prediction equation. Modifi-cation of Diet in Renal Disease Study Group.Arch Intern Med 1999; 130: 461–70.

7. British National Formulary 54; 2007. BritishMedical Association and Royal PharmaceuticalSociety of Great Britain.

8. Zeeh J, Platt D. The aging liver: structural andfunctional changes and their consequences fordrug treatment in old age. Gerontology 2002;48(3): 121–7.

9. Routledge PA, O’Mahony MS, WoodhouseKW. Adverse drug reactions in elderly patients.Br J Clin Pharmacol 2004; 57(2): 121–6.

10. De Geest S, Sabate E. Adherence to long-termtherapies: evidence for action. Eur J CardiovascNurs 2003; 2(4): 323.

11. Dixon WM, Stradling P, Wootton ID. Outpa-tient P.A.S. therapy. Lancet 1957; 273(7001):871–2.

12. Haynes RB, Yao X, Degani A, Kripalani S,Garg A, McDonald HP. Interventions to en-hance medication adherence. Cochrane Data-base Syst Rev 2005; (4):CD000011.

13. Bloom BS. Daily regimen and compliance withtreatment. BMJ 2001; 323(7314): 647.

14. Benner JS, Glynn RJ, Mogun H, Neumann PJ,Weinstein MC, Avorn J. Long-term persistencein use of statin therapy in elderly patients. JAMA2002; 288(4): 455–61.

15. McDonnell PJ, Jacobs MR. Hospital admissionsresulting from preventable adverse drug reac-tions. Ann Pharmacother 2002; 36(9): 1331–6.

16. Senst BL, Achusim LE, Genest RP, CosentinoLA, Ford CC, Little JA et al. Practical approachto determining costs and frequency of adversedrug events in a health care network. Am JHealth Syst Pharm 2001; 58(12): 1126–32.

17. Granger BB, Swedberg K, Ekman I, GrangerCB, Olofsson B, McMurray JJ et al. Ad-herence to candesartan and placebo and out-comes in chronic heart failure in the CHARM


programme: double-blind, randomised, con-trolled clinical trial. Lancet 2005; 366(9502):2005–11.

18. Osterberg L, Blaschke T. Adherence to medica-tion. N Engl J Med 2005; 353(5): 487–97.

19. Svarstad BL, Chewning BA, Sleath BL,Claesson C. The Brief Medication Question-naire: a tool for screening patient adherence andbarriers to adherence. Patient Educ Couns 1999;37(2): 113–24.

20. Claxton AJ, Cramer J, Pierce C. A systematicreview of the associations between dose reg-imens and medication compliance. Clin Ther2001; 23(8): 1296–1310.

21. Bissell P, May CR, Noyce PR. From compli-ance to concordance: barriers to accomplishinga re-framed model of health care interactions.Soc Sci Med 2004; 58(4): 851–62.

22. Beijer HJ, de Blaey CJ. Hospitalisations causedby adverse drug reactions (ADR): a meta-analysis of observational studies. Pharm WorldSci 2002; 24(2): 46–54.

23. Pirmohamed M, James S, Meakin S, Green C,Scott AK, Walley TJ et al. Adverse drug reac-tions as cause of admission to hospital: prospec-tive analysis of 18 820 patients. BMJ 2004;329(7456): 15–19.

24. Leendertse AJ, van den Bemt PMLA, Eg-berts ACG. Hospital admissions related to med-ication (HARM). Arch Int Med 2008; 168:1890–6.

25. Francis SA, Barnett N, Denham M. Switchingof prescription drugs to over-the-counter status:is it a good thing for the elderly? Drugs Aging2005; 22(5): 361–70.

26. Cebollero-Santamaria F, Smith J, Gioe S, VanFrank T, Mc CR, Airhart J et al. Selectiveoutpatient management of upper gastrointestinalbleeding in the elderly. Am J Gastroenterol1999; 94(5): 1242–7.

27. Lewis JD, Kimmel SE, Localio AR, MetzDC, Farrar JT, Nessel L et al. Risk ofserious upper gastrointestinal toxicity withover-the-counter nonaspirin nonsteroidal anti-inflammatory drugs. Gastroenterology 2005;129(6): 1865–74.

28. Oborne CA, Luzac ML. Over-the-countermedicine use prior to and during hospitalization.Ann Pharmacother 2005; 39(2): 268–73.

29. Huang SM, Hall SD, Watkins P, Love LA,Serabjit-Singh C, Betz JM et al. Drug interac-tions with herbal products and grapefruit juice:a conference report. Clin Pharmacol Ther 2004;75(1): 1–12.

30. Suvarna R, Pirmohamed M, Henderson L. Pos-sible interaction between warfarin and cranberryjuice. BMJ 2003; 327(7429): 1454.

31. Yuan CS, Wei G, Dey L, Karrison T, Nahlik L,Maleckar S et al. Brief communication: Ameri-can ginseng reduces warfarin’s effect in healthypatients: a randomized, controlled Trial. Ann In-tern Med 2004; 141(1): 23–7.

32. Kuipers MAJ, Marum van RJ, Egberts ACG,Jansen PAF. Relationship between polyphar-macy and underprescribing. Br J Clin Pharma-col 2008; 65: 130–3.

33. Holmes HM, Hayley DC, Alexander GC, SachsGA. Reconsidering medication appropriatenessfor patients late in life. Arch Intern Med 2006;166(6): 605–9.

34. Tamblyn RM, McLeod PJ, Abrahamowicz M,Monette J, Gayton DC, Berkson L et al. Ques-tionable prescribing for elderly patients in Que-bec. CMAJ 1994; 150(11): 1801–9.

35. Sloane PD, Gruber-Baldini AL, Zimmerman S,Roth M, Watson L, Boustani M et al. Medica-tion undertreatment in assisted living settings.Arch Intern Med 2004; 164(18): 2031–7.

36. Higashi T, Shekelle PG, Solomon DH, KnightEL, Roth C, Chang JT et al. The quality of phar-macologic care for vulnerable older patients.Ann Intern Med 2004; 140(9): 714–20.

37. Soumerai SB, McLaughlin TJ, Spiegelman D,Hertzmark E, Thibault G, Goldman L. Adverseoutcomes of underuse of beta-blockers in el-derly survivors of acute myocardial infarction.JAMA 1997; 277(2): 115–21.

38. Lee DS, Tu JV, Juurlink DN, Alter DA, Ko DT,Austin PC et al. Risk-treatment mismatch in thepharmacotherapy of heart failure. JAMA 2005;294(10): 1240–7.

39. Choudhry NK, Anderson GM, Laupacis A,Ross-Degnan D, Normand SL, Soumerai SB.Impact of adverse events on prescribing war-farin in patients with atrial fibrillation: matchedpair analysis. BMJ 2006; 332(7534): 141–5.

40. Ko DT, Mamdani M, Alter DA. Lipid-loweringtherapy with statins in high-risk elderly pa-tients: the treatment-risk paradox. JAMA 2004;291(15): 1864–70.

41. Heston LL, Garrard J, Makris L, Kane RL,Cooper S, Dunham T et al. Inadequate treatmentof depressed nursing home elderly. J Am GeriatrSoc 1992; 40(11): 1117–22.

42. Turner NJ, Haward RA, Mulley GP, SelbyPJ. Cancer in old age–is it inadequately in-vestigated and treated? BMJ 1999; 319(7205):309–12.


43. Kuzuya M, Masuda Y, Hirakawa Y, Iwata M,Enoki H, Hasegawa J et al. Underuse of med-ications for chronic diseases in the oldest ofcommunity-dwelling older frail Japanese. J AmGeriatr Soc 2006; 54(4): 598–605.

44. Boyd CM, Darer J, Boult C, Fried LP, Boult L,Wu AW. Clinical practice guidelines and qualityof care for older patients with multiple comorbiddiseases: implications for pay for performance.JAMA 2005; 294(6): 716–24.

45. Oborne CA, Batty GM, Maskrey V, SwiftCG, Jackson SHD, Development of prescribing

indicators for elderly medical patients. Br J ClinPharmacol 1997; 43: 91–7.

46. Cockroft TW, Gault MW, Prediction of creati-nine clearance from serum creatinine, Nephron1976; 16: 31–4.

47. Fick DM, Cooper JW, Wade WE, Waller JL,MacLean JR, Beers MH. Updating the Beerscriteria for potentially inappropriate medicationuse in older adults: results of a US consensuspanel of experts. Arch Intern Med . 2003; 163:2716–24.

prescribing for elderly patients || clinical pharmacology of ageing

Documents