case study: registered nurses and resident physicians

Chapter 8

Case Study: Registered Nursesand Resident Physicians

ContentsPage

SHIFT WORK IN NURSING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Registered Nurses: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Evidence of Dissatisfaction With Shift Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Consequences of Shift Work in Nursing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

EXTENDED DUTY HOURS AND GRADUATE MEDICALEDUCATION PROGRAMS . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Libby Zion Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Current Status of Graduate Medical Education Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Regulation of Residents’ Hours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Effects of Extended Duty Hours on Residents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .CHAPTER PREFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

BoxesBox

155155156157158159

166168170170176178179

Page

8-A. Impact of Shift Work on Nursing in Great Britain.... . . . . . . . . . . . . . . . . . . . . . . . ● .,..... 1588-B. Limitations on Residents’ Hours in Great Britain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1718-C. Reduction of Residents’ Hours in New Zealand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

Figure8-1. Prevalence of Hospitals With

Table

FigurePage

on Residents’ Hours . . . . . . . . . . . . . . . . . . . . . . . . . . 170

TablesPage

8-1. Studies of Health Consequences of Shift Work in Nursing . . . . . . . . . . . . . . . . . . . . . . . . . . .8-2. State Legislative Action Regarding Residents’ Work Hours . . . . . . . . . . . . . . . . . . . . . . . . . .8-3. Residents’ Working Hour Requirements in the Residency Review Committee’s

Updated Accreditation Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

163171

174

Chapter 8

Case Study: Registered Nurses and Resident Physicians

Health care is a 24-hour-a-day business. Peopleget sick and need care around the clock, andhospitals must be staffed with health care profes-sionals ready to meet those needs. This means thatsome health care employees, such as nurses andresidents, l must work at night, on weekends, and onholidays. Extended duty hours and circadian rhythmdisruption play a major role in the training ofresidents. Historically, that training has been markedby long, intensive hours over several years. Someresidents have worked more than 130 hours per weekin shifts of 12 to 60 hours, with every other night oncall (58). The resulting sleep deprivation is a majorsource of stress in residency (23). Likewise, manynurses must work nondaytime shifts and endure theresulting circadian rhythm disruption for the dura-tion of their professional lives.

Although the issues discussed in this case studyaffect some portion of virtually every group workingin hospitals (e.g., laboratory and x-ray technicians,pharmacists, transport teams, chaplains, social work-ers, runners, admitting clerks, housekeepers), thefocus here will be limited to nurses and residents.This case study examines the effects of shift work onthe health and well-being of nurses and the qualityof care they deliver and how the structure of residenttraining programs may result in sleep deprivationthat can affect the health and well-being of residentsand their patients.

SHIFT WORK IN NURSING

Jane Doe . . . worked as a registered nurse (R.N.)at a hospital on the West Coast. She had beenworking all night and was in the 12th hour of a12-hour shift. She was under time pressure tocomplete her work She was disconnecting a pa-tient’s I.V. and had to dispose of a used I.V. needle.In the process, she was accidentally stuck by theneedle, which turned out to contain HIV-infectedblood. Jane Doe seroconverted and is now HIV-positive. The fact that she was tired at the end of a12-hour shift, that the hospital was understaffed, andthat the hospital did not have proper containersaccessible to dispose of sharp needles all combined

to create a hazardous working condition, with tragicresults (88)0

A nurse at a Northeastern hospital was onrestricted duty due to a heart condition but eventuallywas required to work 18 hours straight. She was onmedication and increased the dosage to help coverthe demands of her work situation. She became ill,and her case, along with 14 others, was submitted tolabor arbitration (105). The case was settled justprior to the arbitration hearing (98).

A malpractice case was filed against a nurse in theNortheast because a patient died during the nurse’ssecond continuous 8-hour shift (105).

Registered Nurses: An Overview

Anecdotes such as these are easy to find, but thecontribution work hours may play in them is difficultto document. Several explanations for this seemplausible. First, statistics on the relationship be-tween hours of work and adverse events for nursesare not tallied by the U.S. Department of Labor, andthere is no statutory requirement that unions oremployers collect such data. As illustrated by thecase of Jane Doe, extended duty hours may be onlyone of several factors contributing to an incident,and a nurse may therefore not interpret an incidentas related to shift work. Even if nurses do believethat shift work is causing problems, they may find itdifficult to prevail when they file grievances overshift work issues unless these issues are covered intheir collective bargaining agreements. Nonunionnurses often work at facilities without formalgrievance procedures and, thus may have less oppor-tunity to pursue shift work concerns through astructured complaint process (89).

A number of unions representing nurses, includ-ing the American Nurses Association, Service Em-ployees International Union, American Federationof State, County and Municipal Employees, and theFederation of Nurses and Health Professionals (adivision of the American Federation of Teachers),bargain over shift assignments, length of timebetween shifts, shift of choice, split shifts, and shift

~For tie ~Wo~e~ of ~~ ~ha~ter, a resident is my post~duate medi~ &fiee, inclu~g interns, in ~fi rnedic~ ~(i swgkal specialties. Residentsare also referred to as house officers or house staff.

–155–297-935 - 91 - 6 QL 3

156 ● Biological Rhythms: Implications for the Worker

rotations for nurses. In the words of one unionemployee, however:

. . . our success at limiting shift work through thecollective bargaining process has been modest atbest. We attribute this in part to the lack of any legalor regulatory framework on shift work and extendedhours. It is very difficult to persuade employers toadopt measures that exceed existing legal require-ments. In addition, in many of these industries, shiftwork and night work are unavoidable. In these cases,we attempt to give the employees more of a choiceand more control over their hours of work (88).

Some nurses must work extended hours as acondition of their employment, while others wel-come the opportunity to mass their hours. Bothsituations may present challenges to the well-beingof nurses, their families, and their patients. Thissection reviews research on shift work in nursing andthe consequences of shift work for family and sociallife, health, and work performance of nurses.

R.N.s are the largest health care profession.2

According to a recent survey, there were 2,033,032R.N.s licensed to practice in the United States in1988, of whom 1,627,035 (80 percent) were em-ployed in nursing positions (67). Nursing is pre-dominantly a woman’s profession; in 1988, only3 percent of registered nurses were men. Inaddition, 71 percent of R.N.s were married, and54 percent had at least one child living at home.One-half of all R.N.s in 1988 were under age 40, and63 percent were under age 45. Thus the majority ofnurses have families or are in their childbearingyears, sociodemographic factors that bear on shiftwork.

Of those employed in nursing in 1988,68 percentwere working full-time and 32 percent were workingpart-time. Hospitals traditionally have been the prin-cipal employer of R.N.s. In 1988, 68 percent of em-ployed registered nurses worked in hospitals, and noother single type of employment setting accountedfor more than 8 percent of employed nurses. Nursinghomes and other long-term care facilities employedonly 7 percent of R.N.s in 1988. Ninety percent ofR.N.s employed in hospitals in 1988 worked in non-Federal, short-term care hospitals, 6 percent workedin Federal Government hospitals, and 4 percent

worked in non-Federal, long-term care hospitals.Among R.N.s working in hospitals in March 1988,2.7 percent (or 30,358 nurses) were working througha temporary employment service (67).

Prevalence

No national survey of R.N.s provides estimates ofthe prevalence of shift work or the number of yearsthe typical R.N. spends in shift work during her orhis career. However, since the organizations thatemploy the vast majority of nurses—hos-pitals, nursing homes, and other extended carefacilities—require 24-hour operations, it is likelythat a sizable proportion of R.N.s are engaged inshift work. Moreover, many nurses also workovertime, since they draw time-and-a-half pay forovertime (62). Data from one survey show thatamong 764 responding community hospitals, anaverage of71 percent of R.N.s were reported to worksome nondaytime hours, and an average of 33 per-cent of R.N.s worked some combination of day,evening, and night shifts, including rotating all threeshifts; furthermore, only 7 percent of R.N.s hadevery weekend off (l).

Another survey reported that 46 percent of R.N.semployed in nursing homes in 1985 worked non-standard hours, and 9 percent worked rotating shifts(107). The smaller proportion of R.N.s on rotatingshifts in nursing homes compared with hospitalscould be attributed to the fact that nursing homeR.N.s are more likely to be in supervisory positionsthan in staff positions.

Estimates for the total U.S. labor force suggestthat shift work is more prevalent among nursesthan women workers in general (see ch. 4). Datacollected in 1985 on wage and salary workers age16 and over revealed that 13 percent of full-timewomen workers and 43 percent of part-time womenworkers were involved in some type of shift work.Among full-time workers in health assessment andtreatment occupations (the category that includesR.N.s), 31.3 percent were in some type of shift work.Rotating shifts were the most frequent kind of shiftwork in these occupations. (Ch. 4 gives a morethorough discussion of the extent of shift work in thelabor force.)

2 ~~ ~epofi pefi~ to @~te~d ~~es. N~sing perso~el in the us. he~th c~e fidustry include @s&d n~ses, licensd pmCtiCd-VOCdiOlld

nurses, and unlicensed ancillary personnel, including nursing aides. It is estimated that these three occupations included over 3 million persons employedin health care in 1984 (108).

Chapter 8-Case Study: Registered Nurses and Resident Physicians ● 157

Photo credit: Joint Commission on the Accreditation of Healtheare Organizations

Nurses tend to patients within view of the nursing station.

Patterns

Traditionally, hospitals and nursing homes haveoperated on three 8-hour shifts (day, evening, andnight shifts), with nurses typically working five8-hour shifts per week on either a freed or rotatingbasis, including weekend work.3 Shifts typicallyoverlap for one-half hour, in order that nursesleaving the unit can communicate with those arriv-ing regarding patient status or anticipated admis-sions. The pattern of activity in hospitals varies bytime of day, so the tasks and workloads of nurses canvary by shift. Night shifts are generally consideredto be slower and to require a smaller complement ofnurses, since most patients are asleep and fewerprocedures take place at night. A possible exceptionmight be the emergency room, where activity can befrenetic during the night shift. On the other hand,night nurses are expected to be able to manageemergencies when other services are not readilyavailable.

Shift patterns (e.g., timing and length of shifts,patterns of rotation), the method of shift assignment(the degree to which nurses can control theirassignments by bidding on preferred shifts or tradingshifts), and compensation for shift work (e.g., higherhourly wages for undesirable shifts) are determinedby the employing organization and vary widely. Insome decentralized organizations, these policies aredetermined at the unit level. Some aspects of shiftwork, such as requirements regarding rotation,mandatory overtime, or pay differentials, becomeissues in collective bargaining by nurses, althoughonly 20 percent of hospitals report having organizedcollective bargaining for R.N.s (predominantlythrough State nurses’ associations) (l).

Typically, nursing coordinators or nursing man-agers have authority over shift assignments for staffnurses and resolve scheduling conflicts. Assign-ments to undesirable shifts (e.g., night or weekendwork or rotating shifts as opposed to fixed shifts) areoften an inverse function of seniority. Some nursing

3 In nursing, the day shift typic-ally refers to the 8-hour shill occnrring during daytime hours (e.g., 7 a.m. to 3:30 p.m.), as contrasted with evening(e.g., 3 p.m. to 11:30 p.m.) and night shifts (e.g., 11 p.m. to 7:30 a.m.). In this chapter, nurses who workfmed day shifts are not regarded as shift workers.


administrators reward seniority with more desirableshifts as a retention strategy (61).

Six distinct hospital scheduling alternatives ap-pear to exist, although some are used in very fewhospitals. The models are:

●

●

●

●

●

●

the traditional five 8-hour shifts per week,including rotation within a l-week cycle;the “Baylor Plan,” consisting of two or three12-hour shifts for separate weekend staff, withthe regular staff working traditional 8-hourshifts;‘‘4 to 40,’ consisting of four 10-hour shifts perweek;“7-on, 7-off,” consisting of seven 10-hourshifts on alternative weeks;“ 12-hour shifts,” consisting of three 12-hourshifts one week, four 12-hour shifts the next;and‘‘customized schedules," involving choice ofvarious shif t lengths (47). -

Traditional 8-hour shifts continue to predominate innursing. Community hospitals report that nearly 79percent of R.N.s, on average, work 8-hour shifts, 17percent work 12-hour shifts, and 4 percent workother types of shifts, including partial shifts of 4 to6 hours (l). Although a number of experiments withhospital shift lengths and patterns have been re-ported, these appear to affect a relatively smallproportion of R.N.s. For example, one hospital inNorth Carolina has reported a 12-hour shift systemin which nurses work seven 12-hour days in suc-cession, followed by 7 days off (40). Urban hospitalsappear to be more innovative with regard to shiftissues because they face greater competition fornursing staff (61). At Johns Hopkins Hospital, aprofessional practice model, in which the nursesagree to provide 24-hour coverage of their unit inexchange for self-management and annual salaries,has been in place on some nursing units since 1981.The model permits nurses to design their own shiftsystems and to make and monitor their own shiftassignments (27).

Differentials in pay for shift work are used widelyby U.S. hospitals. In a 1988 survey, over 97 percentof hospitals reported paying more to full-time andpart-time R.N.s for evening and night shifts (l).However, only 41 percent of hospitals reported pay-ing more for weekend shifts (l). In 1987, theaverage hourly differential paid for the eveningshift by U.S. hospitals was $0.79, for the night

Box 8-A—Impact of Shift Work on Nursingin Great Britain

A recent study for the British Department ofHealth entailed a survey of the shift schedules ineffect in large (400+ beds) general hospitals inEngland and Wales. Over 50 percent of all suchhospitals supplied details of their shift schedules.Only a small minority of hospitals (less than 20percent) had a regular shift schedule, The majorityof hospitals (65 percent) claimed to have a flexiblesystem in which the individual nurses’ requestswere taken into account when drawing up the dutyrosters. The remainder had an irregular system inwhich no account was taken of nurses’ requests.

Despite these differences, timing and duration ofshifts were remarkably consistent among hospitals.Approximately 80 percent of the early morningshifts started between 7:30 a.m. and 8 a.m. andlasted for 8 to 8.5 hours. The starting time, but notthe duration, of the late afternoon shift was slightlymore variable. Ninety percent started between12:30 p.m. and 2 p.m. Similarly, most night shiftsstarted between 8:30 p.m. and 9:30 p.m. and lastedbetween 10 and 12 hours. Perhaps more important,while new schedules involved more than two orthree successive early or late shifts, spans of sevenor eight successive night shifts were not uncom-mon, These data contrast with U.S. patterns,presumably reflecting national differences.

SOURCES: OffIce of Technology Assessment, 1991; based onS. Folkard, Department of psychology, Universityof Shefileld, Sheffkld, United Kingdom, personalcommunicatio~ January 1991.

shift $1.02 (5). Nonfinancial compensation for shiftwork (e.g., extra days off or child-care) is rare. Theeffectiveness of pay differentials in attracting nursesto undesirable shifts is unknown. Box 8-A discussesshift work schedules for nurses in Great Britain.

Evidence of Dissatisfaction With Shift Work

A number of polls and research studies conductedduring the last decade provided evidence that shiftwork, or the nurse’s perceived lack of control overscheduling, was associated with job dissatisfactionand turnover. However, few prospective studies ofjob satisfaction have distinguished between theeffects of shift work and other job, organizational, orpersonal attributes. Results of a 1981 Texas surveyof 3,500 nurses suggested that work overload andshift work issues were intermingled and that, ingeneral, nurses desired greater control over workassignments and scheduling (1 13). (The job condi-

Chapter 8--Case Study: Registered Nurses and Resident Physicians ● 159

tion eliciting the most dissatisfaction, however, wassalaries.)

In one 1987 poll, 1,643 nurses were asked whatcaused them the greatest dissatisfaction with theirjobs:

●

●

●

76 percent of nurses responding said increasedwork hours;51 percent of nurses responding said increasedshift rotation; and43 percent of nurses responding said increasedweekend work due to the current nursingshortage.

In addition, 65 percent of nurses responding said thata minimum of every other weekend off was mostimportant in evaluating their present jobs, and 53percent reported that a permanent shift assignmentwas most important. (Salary and benefits wereranked equal in importance with a permanent shiftassignment) (30).

A l-year prospective study of nursing job satisfac-tion and turnover asked R.N.s in two large university-affiliated hospitals to report their satisfaction withscheduling of work hours. Overall, nurses weredissatisfied with their scheduling (114). In onehospital, working rotating shifts, as opposed to friedshifts, was found to lower nurses’ perceptions oftheir control over the work environment. Lowerperceived control, in turn, was the strongest predic-tor of job dissatisfaction. In the second hospital, shiftwork had no significant impact on perceived control(115).

A study of job satisfaction of R.N.s in fiveshort-term acute care hospitals measured satisfac-tion at two points, 8 months apart. Assignment to theday shift was associated with higher job satisfaction(12). A study of 146 nursing units in 17 hospitalsexamined unit-level determinants of turnover rates.Greater shift rotation among R.N.s in a unit was seenas an indicator of less staff cohesiveness, which waspredicted to produce higher rates of turnover. In fact,extent of shift rotation was found to be a marginallysignificant predictor of turnover rates (2).

Only one randomized trial of different shiftpatterns in nursing has been reported. In a 788-bedtertiary-care hospital (i.e., a hospital that provideshighly intensive, sophisticated care for medicalconditions that are difficult to manage in a commu-nity hospital), 12 randomly selected medical-surgical units staffed by both R.N.s and licensed

practical nurses (L.P.N.s) were assigned at randomto one of four schedules:

●

●

●

●

fixed shifts;computer-assisted scheduling in which nurses’shift preferences were considered;select-a-plan, in which unit nurses designedtheir own scheduling system, typically combin-ing 8- and 12-hour shifts; anda control group.

Descriptions of job attitudes were obtained from 98percent of nurses before and after the shift assign-ments, but the length of the followup period isunspecified. The results did not indicate substantialvariations by schedule in R.N.s’ job attitudes orturnover at followup (20).

A number of commissions and task forces inrecent years have recommended that healthservice organizations address the issue of shiftwork. A 1981 study of nursing and nursing educa-tion noted flexible scheduling as a strategy toincrease nursing labor supply to hospitals andnursing homes (48). A 1983 report recommendedthat flexible scheduling be developed as a compo-nent of models for organizational change in nursing(68). A study of magnet hospitals found that flexiblescheduling and elimination of rotating shifts wereamong the four most important factors promotingrecruitment and retention of nursing staff (3).

A consensus appears to have developed regardingthe need to provide shift work alternatives for nursesand to increase nurses’ involvement in schedulingdecisions. A 1988 report cited work scheduling as amajor source of stress among hospital R.N.s (108).Most recently, the authors of a study of 421 hospitalsand over 15,000 nurses in six urban areas recom-mended that hospitals offer varied and flexibleschedules as a means of responding to the workpreferences of nurses and resolving the currentnursing shortage (91).

Consequences of Shift Work in Nursing

While there is a body of research that ad-dresses the consequences of shift work on thehealth, well-being, and performance of workersin a variety of industries, these industries employmostly men; relatively little research has beenconducted specifically on women shift workers(24). As a result, it is difficult to generalize aboutnurses from existing studies. (The consequences of


Photo credit: Harvati Community Health Plan

Patient getting his blood pressure checked by a nurse.

shift work are discussed more fully in ch. 5.) Innursing, type of shift is often confounded by age,because younger, less experienced nurses tend to beassigned to rotating shifts. This discussion ofconsequences focuses on studies in which nurseshave been subjects, but it will allude occasionally tostudies of other women shift workers or to theabsence of information about women shift workers.

Work Performance

Studies of the work performance of shift workershave generally focused on lost productive time as aconsequence of health problems (e.g., absenteeismdue to illness or use of health services) and on errorsand injuries as a consequence of fatigue or disruptedcircadian rhythms. A few studies of nurses havereported results for sick days, use of health services,self- or supervisor-reported performance levels, andvarious indicators of quality of patient care.

Sick Days and Use of Health Services—A 1977study of the health consequences of shift workconducted by the Stanford Research Institute for theNational Institute for Occupational Safety andHealth (NIOSH) used employee records to studysick days and clinic visits by nurses on four shifts.(The number of sick days alone is not a reliableindicator of health problems, since sick days arelimited by personnel policies.) Nurses on a rotatingschedule tended to take more sick days thannurses on fixed shifts, and rotating nurses tended

to have more serious reasons for taking sick days(e.g., acute respiratory infection, upper gastro-intestinal tract distress). With regard to clinicvisits, rotating nurses attended workplace clinicsduring work hours more frequently than fixed shiftnurses and for a wider array of complaints (103),indicating that rotating nurses lose more time fromwork (due to both sick days and clinic visits) andcould be less productive on the job.

Job Performance Ratings-A few studies of theeffects of shift work on the job performance ofindividual nurses have used self-reports of perform-ance or the reports of supervisors. The NIOSH studyfound that nurses on fixed afternoon-evening androtating shifts reported lower levels of satisfactionwith their work performance than did freed night andday shift nurses (103). Another study found thatnurses on rotating shifts received lower job perform-ance ratings by their supervisors, compared to nurseson freed shifts (49).

In a recent study of job performance and job-related stress among 482 R.N.s in five hospitals inthe Southeastern United States, self-reported jobperformance was examined in relation to type ofshift worked. On overall job performance, nurses onrotating shifts reported the lowest performancelevel, followed by freed afternoon nurses, freednight nurses, and freed day nurses. The investigatorsfound that only one of a number of dimensionsstudied was significantly associated with shift: forperformance with regard to professional develop-ment, rotating and freed afternoon nurses scoredthemselves lower than fixed day and freed nightnurses. There were no significant differences by shiftfor leadership, teaching-collaboration, planning-evaluation, or interpersonal relations-communica-tions. Rotating nurses also reported the highestlevels of job-related stress. These relationships heldwhen anticipated turnover, position level, and lengthof time on the present shift schedule were controlled(21).

Quality of Patient Care—Very little researchhas been conducted on the relationship betweenshift work and quality of patient care by nurses.The few studies that have addressed the effects ofshift work on quality have been small-scale evalua-tions of shift work demonstration projects.

In nursing studies in general, quality of care istypically measured in terms of the nursing careprocess rather than patient outcomes. Process meas-

Chapter 8-Case Study: Registered Nurses and Resident Physicians . 161

ures of quality usually take place at the unit level andinvolve chart audits, in which raters examine nursingcare plans to determine whether specific procedureshave been recorded. An alternative method is directobservation of nursing procedures, in which a raterassesses performance along a number of dimen-sions. Other indicators of quality include incidentreports (medication errors, accidents, or injuries)and patient satisfaction. One should remember,however, that quality of care as defined by providerand quality of care as defined by patient may not bethe same (15).

Research to date implies two mechanisms bywhich shift patterns might affect quality of nursingcare. First, shift work patterns that are more compat-ible with circadian rhythms would be expected toresult in less fatigue and increased alertness on thejob. Second, shift work patterns that are moresatisfying to nurses would be expected to result ingreater nurse retention, unit cohesiveness, and conti-nuity of care across shifts.

A 4-month trial of the 4 to 40 workweek (four10-hour days per week) in a pediatrics unit employ-ing both R.N.s and L.P.N.s was conducted in Seattle,Washington. The new scheduling system, comparedto the traditional 8-hour, 5-day shift system, wasexpected to improve quality of care, defined in termsof intershift continuity of care. Process measures ofnursing care quality revealed little change from thetraditional system. Staff reports of quality, however,did show improvement (50).

In another program, a 12-hour shift system wasintroduced on an intensive care unit and evaluatedover a 6-month period. The impact of the 12-hourshift, compared to the traditional 8-hour shiftpattern, was measured with regard to nurses’ jobsatisfaction, interpersonal relations in the workenvironment, nurses’ health status, nurses’ fatigueand alertness, and quality of patient care. Nosignificant differences were found between types ofshift for alertness (reaction times) or fatigue, al-though in the 12-hour shift system reaction time wasfaster for day shift than for night shift nurses. Usingretrospective analysis, the investigators found nosignificant difference between the quality of nursingcare at the time studied and that during a correspond-ing period of the previous year. Data from incidentreports showed no change in the rate of incidentswith the switch to 12-hour shifts (29).

In a pilot study of 12-hour shifts in a surgicalintensive care unit, quality of patient care prior toimplementation of 12-hour shifts was compared toquality of care 1 year later. Some increase insubjective feelings of fatigue and decreased accu-racy on performance tests was reported, but evidencefrom chart audits revealed no significant changes,and nurses perceived that their performance im-proved (63).

One small study compared 10 R.N.s working8-hour shifts with 10 R.N.s working 12-hour shiftsin the intensive care units of two Midwesternhospitals. The Quality Patient Care Scale (QUAL-PACS) was administered by an observer to provideconcurrent ratings of patient care provided by eachnurse. Only one of five behavior categories rated bythe QUALPACS demonstrated a statistically signifi-cant difference between 8- and 12-hour shift work-ers, although nurses on 12-hour shifts scored lowerthan nurses on 8-hour shifts on all of the dimensions.Since most of the 12-hour nurses volunteeredcomments about the fatigue they experienced, theinvestigators interpreted these results as indicatingpotential problems with quality of care on 12-hourshifts (69).

It is not known how extended workdays andcompressed workweeks affect patients’ satisfac-tion with nursing care. Conceivably, patients couldreact differently to various shift schedules, depend-ing on the availability of their primary care nurse.One study reports anecdotal evidence of patientsatisfaction with a 12-hour shift system in whichnurses worked 7 consecutive 12-hour days followedby 7 days off (40). However, patients who do not seetheir primary nurse for days at a time may experiencethis as discontinuity in their care.

Health

As discussed in chapter 5, shift work can affecthealth. Disruption of circadian rhythms because ofshift work could have differing effects on men andwomen due to hormonal differences between thesexes. Differences between men’s and women’scircadian rhythms and adjustment of these rhythmsto shift work have been studied, but no definitiveconclusions have been drawn (8). Congressionalinterest in research on women’s health, as exempli-fied by the Women’s Health Equity Act of 1991,supports the theory that the results of research onmale subjects cannot be assumed to apply to women.

162 . Biological Rhythms: Implications for the Worker

Table 8-1 lists recent studies of health outcomesof shift work in nursing. The key findings arediscussed below. Research has demonstrated thatnurses working night and rotating shifts suffermore sleep disturbances than other nurses andmay be at higher risk for various other healthproblems (89).

Sleep Disturbances—Deficits in the quantity orquality of sleep are associated with physical oremotional disorders and with problems in alertnessand performance that can produce injuries on thejob. All but one of the studies summarized i n t a b l e8-1 measured quantity or quality of sleep as afunction of nurses’ shift work. Outcomes studiedincluded amount of sleep, sleep stages and rapid eyemovement (REM) sleep, interruptions of sleep, andsubjective fatigue. A frequent focus has been theadaptation of sleep patterns to different amounts ofnight work. In general, nurses working rotatingshifts and night shifts involving only a few nights onduty had more sleep disturbances than other nurses,although few studies compared nurses on a varietyof shifts or controlled for such variables as age, shiftwork history, and family circumstances. None of thestudies examined the effects of chronic sleep disrup-tion on nurses’ health or work performance.

Several of the studies made a distinction betweenpart-time and full-time night nurses (e.g., thoseworking only 2 nights a week or 2 nights insuccession compared to those working more than 2nights a week or in succession) in an effort to studyshort-term compared to long-term sleep adjustmentto shift work. These studies found that full-timenight nurses were better adjusted than part-timenight nurses in terms of quantity and quality of sleep(34,64,65). Such adjustment was attributed in part togreater commitment to night work among full-timenight nurses, as well as to greater compatibility offamily situations with night work and day sleepamong these nurses (e.g., no children at home).

The dual issues of a selection effect (i.e., nursesselecting shift work based on compatibility withlifestyle) and family roles as mediating variables inadjustment to night work were noted in most of thestudies. For women night shift workers more thanfor men, family responsibilities (e.g., child care,carpooling, housework) during nonworking hourstake precedence over the need to compensate forlost sleep (39). Duration of sleep was found to begreater among unmarried subjects than among

married subjects with children (39). Nurses' life-styles, not just their shift patterns, determinesleep patterns (39). Another study found thatfatigue in nurses working nights (in a sample ofnurses who worked various shifts at irregularintervals) appeared to be due to social factors(number of children, age, and being married or livingwith a partner) (45). Greater number of children alsodecreased the amount of sleep achieved by nursesafter work on the night shift.

The NIOSH study mentioned earlier (103) foundthat nurses on fixed night shifts and on rotatingshifts generally reported more problems withsleep, as compared with nurses on fixed day orfixed afternoon or evening shifts. These includedsubjective reports on the overall adequacy of sleep,trouble falling asleep, awakening during sleep andfalling asleep after awakening, waking up feelingtired, and feeling tired during work. Althoughnurses on all four types of shifts reported desiringthe same amount of sleep, nurses on fixed nightshifts reported getting the least sleep. Further,rotating shift nurses, compared with those onfixed shifts, showed significantly higher inci-dence of fatigue and inadequate sleep patternswhen the effects of age and marital status werecontrolled (103).

Digestive Disorders—The NIOSH study reporteddata on nutritional intake, appetite, and digestiveproblems among nurses by type of shift. Nurses onfixed afternoon or evening shifts reported betterappetites than nurses on other shifts. Night shiftnurses reported needing fewer meals in a 24-hourperiod than other nurses, and day shift nursesreported needing the most. Rotating shift nursesreported more snacking. Day nurses reported themost meals eaten with family or fiends.

Rotating and day nurses reported significantlymore bloating or feeling full; rotating and afternoonnurses reported more gastritis; and rotating nurseshad higher incidence of trouble digesting food.Rotating shift nurses showed a significantlyhigher incidence of digestive trouble than othernurses, when age and marital status were con-trolled (103).

In another study, nurses working three or moreconsecutive nights exhibited greater adjustment intheir meal times during a period of 12 consecutivedays than nurses working single nights or only 2nights in succession. Nurses working more nights

Table 8-l-Studies of Health Consequences of Shift Work in Nursing

study Sample Shifts compared Major findings

Bryden and Holdstock, 12 nursing students in 1 hospital in1973 South Africa (all female)

Felton, 1976 39 nurses in Hawaii (all female)

Folkhard et al., 1978

Gadbois, 1981

Harma et al., 1988

30 nurses in 1 hospital in England(all female)

898 nurses and nursing auxiliariesin 61 hospitals in France (allfemale)

128 nurses and nursing aides in 1hospital in Finland (all female)

Hildebrandt and Stratmann, 6 nurses in Germany1979

lnfante-Rivard et al.,1989

418 nurses and nursing aides in 1hospital in Canada (all female)

Day shift, night shift in samenurses

Night shift, postnight duty in samenurses

Early and late day shifts,permanent nightshirt (full-time,4 nights/week; part-time,2 nights/week)

Fixed night work only (variednumber of nights on duty)

Irregular rotating shifts only(combinations of day, evening,night shifts in 3-week cycles)

7-to 18-day period of night workcompared with 10-dayrecovery period in samenurses

Fixed day, fixed evening (withinboth groups, comparisonsbetween those with andwithout prior night work)

Daytime sleep was shorter than night sleep and had moreinterruptions. REMa sleep occurred sooner in day sleepthan in night sleep.

Peak body temperature; excretion of sodium, potassium, andcreatinine in the urine; and osmolality (concentrations ofthese substances in the urine) occurred later in the day innurses on night duty. After returning to day shift, nurses’urinary sodium, creatinine, and osmolality cycles returnedto baseline pattern, but temperature and potassium did notafter 10 days of followup. Fewer hours of sleep and poorerquality of sleep reported while on night duty.

Circadian rhythms (oral temperature, subjective alertness andwell-being) of full-time night nurses showed greateradjustment than those of part-time nurses, both on firstnight shift and in adjustment from first to second nights.Part-timers reported less sleep and less “calmness” onwaking.

Self-reported sleep duration was shorter, and sleepinterruptions more frequent, for married women withchildren than for unmarried women. Mothers with youngchildren went to bed later in the day.

Neuroticism and morningness were found to increase shift-cycle fatigue, and greater maximal oxygen consumptiondecreased it. Morningness, older age, and having childrendecreased sleep duration after night shift. Shift workexperience and morningness decreased sleep quality afternight work, and oxygen consumption increased it.Gastrointestinal symptoms were increased by neuroticism,marriage, and older age. Neurovegetative symptoms wereincreased by marriage and neuroticism. Musculoskeletalsymptoms were increased,by marriage, neuroticism, andphysical activity and were decreased by oxygenconsumption and muscle strength.

Three “evening type” nurses reacted to night work withflattening of circadian amplitude (temperature and heartrate) and greater tolerance. Three “morning type” nursesdeveloped increased amplitude, higher amounts ofsubjective complaints (e.g., headache, nervousness,irritation), greater compensation for lost sleep, and lowersubjective vigilance.

Prevalence of 9 sleep disorder symptoms ranged from 6percent to 53 percent. Evening workers were at higher riskof not being alert and receptive at rising and at lower risk ofearly morning awakening. Prior night work was associatedwith day tiredness and quantity of sleep.

Table 8-l-Studies of Health Consequences of Shift Work in Nursing-Continued

Study Sample Shifts compared Major findings

Kuchinski, 1989 146 R.N.sb in 1 hospital in Cincin-nati (all female)

LeClerc et al., 1988 824 nurses in 10 hospitals inFrance

Matsumoto, 1978 5 nurses in 1 hospital in Japan

Minors and Waterhouse,1983

Minors and Waterhouse,1985

Tasto et al., 1978;Smith et al., 1979

Verhagen et al., 1987

Webb, 1983

26 nurses

14volunteernurses in 1 hospital inEngland (11 female)

1,219 full-time nonsupervisoryR.N.s and L. P.N.sC in 12 U.S.hospitals (1,195 female)

167 nurses in 6 hospitals in Bel-gium (all female)

19 nurses age 50 to 60 in 3hospitals in Florida comparedwith forty 50- to 60-year-oldfemale university employeeswith no history of shift work

Fixed shifts, rotating shifts

Permanent day, permanent night,rotating without night, rotatingwith <5 nights/month, rotatingwith 5+ nights/month

Day shift, night shift in same nurses

Single nights, paired nights, nightsin blocks of 3+

Short-term night shift (1 to 3 suc-cessive nights), long-term nightshift (3 successive nights)

Fixed day, fixed afternoon/eve-ning, fixed night, rotating

Permanent night (full-time and part-time), 2-shift rotation with somenights

Current night shift with 10 years ofrotating shift experience; nursesnot on current shift work butextensive night or rotating shiftexperience; nurses not on cur-rent night shift and limited priorshift experience

No differences by shift were found in menstrual interval,duration of flow, amount of flow, dysmenorrhea, intermen-strual bleeding, or secondary amenorrhea.

For 38 health variables, including symptoms and healthservices utilization, type of shift had poor explanatorypower.

Time in bed and total sleeping time were lower in daytime thannighttime sleep. Number of awakenings was greater in daysleep. Daytime sleep differed qualitatively from night sleepin appearance of REM and non-REM sleep.

Nurses working single nights maintained a conventionaldiurnal routine. Nurses working a number of consecutivenights showed changes in routine, including afternoonnapping and different meal times.

Nurses working 3 successive nights were better adjusted tonocturnal activity and diurnal sleep, as reflected in amountof sleep, afternoon naps, and fewer subjective complaints.

Nurses who rotate shifts took more sick days than nurses onfixed shifts, suffered more injuries, and reported moredigestive trouble, leg and foot cramps, colds, nervousnessand shaky feelings, fatigue and exhaustion, use of stimu-lants, use of beer and liquor, tension or anxiety, depressionor dejection, confusion or bewilderment, neuroticism, sleepdisturbances, and problems adjusting work schedules andfamily life.

Rotating nurses reported more health complaints. Permanentnurses slept less than rotating nurses.

Little evidence of enduring effect of night shift experience onsleep patterns. Some evidence of moderately changedREM sleep.

aREM, rapid eye movement.bR.N., registered nurse.%.P.N., licensed practical nurse.

SOURCE: Office of Technology Assessment, 1991.


tended to have their evening meal around midnight,indicating a commitment and long-term adaptationto night shift work (64).

Being married increased the reported incidence ofgastrointestinal symptoms (poor appetite, gaseousdistention, heartburn, constipation, and diarrhea),nervous and psychological symptoms (nervousnessand tautness, headache, nightmares, giddiness andnausea, palpitation or irregular heartbeat, handsweating, pricking or numbness of body parts,decrease in libido), and musculoskeletal symptoms(back symptoms, neck, shoulder or hip symptoms,knee or ankle symptoms) (45).

Psychological and Nervous Symptoms—Psy-chological variables may be interpreted as outcomesof shift work or as outcomes of mediating variables.The NIOSH study used the Profile of Mood States(POMS) to measure psychological state and theEysenck Personality Inventory (EPI) to assess per-sonality traits. Psychological state changes overtime, while personality trait is more or less constant(25,54). On the POMS subscales:

●

●

●

●

nurses on a rotating shift reported significantlyhigher tension and anxiety than afternoon-evening and night workers;

rotating nurses reported significantly moredepression and dejection than day or nightworkers;

rotating nurses reported significantly less vigorand more fatigue than nurses on all other shifts;and

rotating nurses reported significantly moreconfusion and bewilderment than nurses on allother shifts.

On the EPI, nurses on a rotating shift were signifi-cantly more anxious and irritable than nurses onother shifts. Interestingly, day shift nurses scoredlower on sociability than afternoon-evening andnight nurses (103).

Anxiety, irritability, and nervousness increaseshift-cycle fatigue in nurses and contribute to ahigher reported incidence of gastrointestinal, nerv-ous, psychological, and musculoskeletal symptoms,regardless of shift (45). A measure of morningness(preferring daytime activity) as opposed to evening-ness (preferring nighttime activity) was also predic-tive:

●

●

The

morningness decreased reported fatigue in morn-ing shift nurses and increased it in evening shiftnurses; andmorningness also decreased the duration andquality of sleep after night shift work (45) (seech. 5).

relationship between specific psychologicalsymptoms and degree of social interaction, whichmay be reduced by shift work, is not known.

Menstrual Dysfunction-Few studies have ex-amined the relationship between shift work andreproductive health in nurses, although environ-mental and occupational conditions can contrib-ute to menstrual dysfunction (51). The NIOSHstudy included some indicators of menstrual dys-function and found that nurses on fixed day shiftsreported more irregular menstrual periods than othernurses and that nurses on all types of freed shiftsreported more menstrual cramps than did rotatingnurses. Night and rotating nurses, however, reportedspending more time lying down due to menstrualcramps, and rotating nurses reported more tension,nervousness, weakness, and sickness at menstrua-tion, as well as longer periods, than did nurses onother types of shifts (103).

In a recent study of the menstrual characteristicsof 146 R.N.s age 20 to 40 in a university hospital, 43percent of the nurses studied worked fixed shifts and57 percent worked rotating shifts. No significantdifferences between these two groups of nurses werefound for self-reported length of the menstrualperiod, duration of flow, amount of flow, dysmen-orrhea, intermenstrual bleeding, or secondary amen-orrhea (55).

Substance Use and Abuse-Only the NIOSHstudy has reported differences in nurses’ substanceuse by type of shift:

●

●

●

●

●

The

rotating nurses, compared with nurses on freedshifts, were more likely to use stimulants;rotating and afternoon shift nurses drank morebeer;rotating nurses drank more alcohol of all kinds;rotating and day shift nurses were more likelyto drink on workdays; andnight nurses were more likely to drink beforegoing to work.

information obtained on quantity of alcoholconsumed was limited, and no conclusions can bedrawn regarding alcohol dependency. Similarly, it is


not possible to determine from the reported datawhether “drinking on workdays occurs during oroutside of work hours or whether drinking onworkdays causes problems in performance (103). Italso is not possible to determine whether people whodrink more prefer different shifts (25,54).

Injuries—Fatigue and impaired cognitive andpsychomotor function during shift work can result ininjuries. As mentioned in chapter 5, laboratorystudies have shown that speed, reaction time, andaccuracy decline after evening hours, suggestingthat more accidents and injuries are likely to occurat night (33). Only the NIOSH study and a reanalysisof those data (97) have reported injuries by type ofshift worked by nurses. Rotating shift workerssuffered more injuries than fixed shift workers.Injury rate was not influenced by age, length ofemployment, or marital status. Furthermore, cuts,bruises, and punctures (mainly to fingers) were themost frequent category of injury on all shifts, androtating nurses had significantly more such injuriesthan nurses on freed shifts (97).

Family and Social Life

The NIOSH study provides some informationabout the effects of shift work on the family andsocial life of nurses. Questionnaires were adminis-tered to 1,219 fill-time, nonsupervisory R.N.s andL. P.N.s in 12 U.S. hospitals; 98 percent of the nurseswere female, the average age was 32.9 years, and 45percent were married. The sample was fairly evenlydistributed across four work schedules: fried dayshift, freed afternoon or evening shift, fixed nightshift, and rotating shift (generally, rotation occurredbetween two shifts only) (103).

Results showed that the greatest disruption offamily and social life occurred for nurses onrotating shifts. Regarding degree of satisfactionwith the amount of time spent with their spouses,rotating and night shift nurses were most dissatis-fied. Night shift nurses reported the most interfer-ence with sexual activities, whereas nurses onrotating shifts reported the most complaints fromspouses regarding their work schedules. As foramount of time spent with children, nurses onrotating shifts were the least satisfied, althoughafternoon and evening shift nurses reported that theyassumed the least responsibility for child rearing anddisciplining. Finally, rotating and afternoon shiftnurses were least satisfied with the time available for

personal activities, sports, and social activities(103).

EXTENDED DUTY HOURS ANDGRADUATE MEDICAL

EDUCATION PROGRAMS“Shifts” implies that there is a preset beginning

and end to your work hours, however long they maybe. We do not work shifts. We just work, When thework is done we can go. If that’s 36 hours later, sobe it. During my internship year, I routinely spent120 hours a week in the hospital. These hours werelargely devoted to menial, nonmedical tasks calledscut work, a catch-all term for patient-care-relatedwork that does not require a medical degree (i.e.,inserting I.V.s, doing blood cultures, transportingpatients, or scheduling tests) (31).

I entered medical school with lofty aspirations ofselfless dedication to the sick, a life of joyfulaltruism. Like many of my colleagues, I saw theseworthy aspirations transformed into cynicism underthe pressure of sleepless call nights in a pediatricsinternship. My love for children became a loathing.As an intern in 1984, I worked over 135 hours aweek taking short call (until 10 p.m.) one night, longcall (40 hours) the next night, and 8 hours off the nextnight. I felt that sleep deprivation dangerouslyimpaired my judgment, giving me the sensation of asleep-walking nightmare. I remembered the story ofthe resident in our program a few years before I camewho had been so impaired by her sleep deprivationthat she fell asleep at the wheel after returning froma 40- to 45-hour shift. She is now quadriplegic andwill never be able to practice medicine again (62).

I was on call for the hospital trauma and codeteams eight times during a 14-week chaplaincyinternship. Like the medical staff, I worked 28- to31-hour shifts, with the most intense work duringnights of little or no sleep. Like them, I alternatedbetween adrenalin-charged highs when an ambu-lance or helicopter arrived and crushingly wearylows when patients died or were transferred. Icounted on the calm, alert time in between to do mybest work. Yet that period shrank as the shiftprogressed. My mind slowed after 2 a.m., and myempathy tumbled. Medical staff seemed to react thesame way. Sometimes tempers frayed as the nightwore on, and we all cut corners. No one was a verycheery friend for a couple of days after a night on call(117).

Graduate medical education, or residency, isformal medical education beyond the M.D. degree.This training incorporates substantial clinical expe-


rience and prepares the residents for practice inrecognized medical or surgical specialties. Resi-dents who complete this training become eligible totake the appropriate board certification examination.Resident training programs began at the JohnsHopkins School of Medicine roughly 100 years ago.In those early days, residents, as the title implies,were expected to live at the hospital, were onpermanent call, and were discouraged from marry-ing (83,100). The first-year trainee was called anintern; now, that year is often referred to aspostgraduate year one, PGY1. The training periodwas l e n g t h e n e d t o include addition-al years-junior resident (PGY2), senior resident(PGY3), and chief resident (PGY4).4 Some special-ties have a fifth year of residency (PGY5), andothers, such as neurosurgery, require 7 years. Thetypical resident has changed since the early days ofresident training. Today the resident is older, morelikely to be a woman, and more likely to have afamily or other commitments outside the hospital(59).

Although extended hours for resident trainingdate back to the early 1900s, when residents lived onhospital grounds and were on permanent call, inmore recent times some residents have rotated thenight shift every other night in addition to their dailyduties (100). In 1975, when the Committee of Internsand Residents, a union, struck New York City’shospitals, many programs revised this requirementand allowed an on-call schedule of every third night(100). According to the Association of AmericanMedical Colleges (AAMC):

. . . the resident benefits by being exposed to patientsthroughout the course of their illnesses. This allowsobservation of both the natural history of the illnessand the impact of the medical intervention. Toexperience all of the learning opportunities, theresident would have to be on duty 7 days a week, 24hours a day. Clearly, such a schedule is unrealisticand does not recognize the possible adverse impactsof fatigue or the resident’s commitments to otheractivities and interests. Therefore, assignment sched-ules for residents must be balanced between compet-ing objectives and constraints (82).

Long hours are a necessary part of residenttraining. They enable residents to follow the pro-

gression of a disease and to provide continuity ofcare. This has been the custom and practice inpostgraduate medical education for nearly 100years. However, questions and concerns have beenraised about the work schedules that have tradition-ally been used in resident training. It has been arguedthat the progression of a disease can no longer belearned in a hospital setting, since patients remain inhospitals for much shorter periods of time than theydid 20 or 30 years ago (10). In addition, somemaintain that long hours are a bow to tradition andan initiation into the profession-a rite of passage—that all would-be doctors must endure (6,100,1 12).Supporters of extended hours for residents argue thatlabels such as “rite of passage” are irresponsibleand destructive (95). Four additional points havebeen identified to support the current system ofextended duty hours with night call hours:

●

●

●

●

sleep-depriving night call is a valid learningexperience;individual idiosyncrasies, not night hours andsleep deprivation, make night call a negativeexperience;night call does not cause permanent distortionof personal and professional sensitivities; andquality of care is not compromised by sleep-deprived physicians (7).

It has been noted that due to the many recentadvances in medical knowledge and technology, thepractice of medicine is not what it was 100 years ago,when the concept of residency was created. Patientstoday come into the hospital sicker, and there aremany more drugs and technologies available to treatthem. In fact, the medical problems of patients aremore severe today than they were even a few yearsago (106). Prior to modern advances in medicaltechnology and pharmacology, physicians were verylimited in what they could do for patients. Physi-cians did their best to make patients comfortable andrelieve suffering, but few diseases could be cured.There were no intensive care units, only a handful ofantibiotics, and patients were admitted for observa-tion (a practice not routinely performed todaybecause of insurance regulations and reimbursementschedules) (16). Today, both the intensity of medicaltreatment and the pressure from patients for apositive outcome are higher.

qN~t ~ pGY4~ ~ chief msiden~. ~s tm refers t. the resident who devises the schedule and resigns cM. ody hI medicine ad pe&&iCS, ~dperhaps psychiatry, is the PGY4 automatically chief. In other specialties, such as pathology and surgery, usually one of the PGY5s is made chief. Theother PGY5s are called senior residents.


According to one chief resident:

In reality, a resident does not sit at the bedside andwatch the progression of a disease process. Afteradmitting a patient, the resident scurries about,drawing blood on another patient, checking labvalues, or admitting the next patient. Therefore, longhours for residents provide a cheap source of laborand less real teaching benefit, with very littlecontinuity. Furthermore, once residency finishes,rarely does a physician spend the whole night at thebedside of a patient, ready to pounce on the nextdevelopment of disease. In fact, many physiciansalternate call schedules with other physicians ormembers of their group. Very often the person who‘‘covers’ for a physician has never seen the patientbefore and only has a short description of the patientand the current medical problems. What happens tocontinuity of care in this situation? (17).

In addition, in some diseases, such as cancer, aperiod of weeks, months, or years is necessary tofollow disease progression in patients (117). Thereis also the position that it is not the resident but theR.N. who is constantly at the patient’s bedside; aresident may check on a patient frequently during ashift but usually does not spend more than a fewminutes with the patient (101).

One legal expert rejects the notion that injuriesmay result because doctors are fatigued by longshifts of duty and maintains that, despite being onduty for long stretches at a time, doctors are able tosleep during slack periods (71). A chief residentargues that, although there may be times when thereis a lull in the admissions process, it is during suchlulls that the resident tries to catchup on all his or herother work (I.V.s, drawing blood, assessing othersick patients on the floor, etc.) (17). A resident oncall at night is often responsible for the care of 40 to60 patients, and in some busy inner-city hospitals,first-year residents may have as many as 15 newadmissions each night (17). One senior residentpoints out that in many programs if a resident iscaught sleeping during slack hours, he or she isreprimanded (62). Further, sleeping during slackperiods does not take into account the likelihood thatsuch sleep may be short, unpredictable, and repeat-edly interrupted. Prior to the implementation of thenew regulations of hours of work in New York, it hadbeen reported that first-year residents averaged 2.6hours of sleep per day (44).

Libby Zion Case

An incident that recently caused the public tofocus on the issue of residents’ hours was the deathof a patient, Libby Zion, in a New York hospital. In1984, Libby Zion, an 18-year-old woman, wasadmitted to New York Hospital-Cornell UniversityMedical Center through the emergency room anddied a few hours later. Her father, an attorney and aformer reporter for the New York Times, charged thathis daughter had received substandard care at thehands of residents. Zion was treated by an intern anda junior resident. Both of them had been on duty forsome 18 hours prior to her admission.

During the 8 hours between her admission and herdeath, Zion was not examined by an emergencyroom attending physician (a physician who hascompleted all phases of the medical educationprocess and whose job is teaching interns andresidents), because none was on duty. Telephonediscussions were held with her private attendingphysician, although he did not examine her inperson. Further, it is argued that the intern andresident who were on duty failed to provide adequatemonitoring of the patient and prescribed medica-tions that were contraindicated in light of Zion’shistory of drug and medication use.

Grand Jury Findings

A grand jury convened to examine the circum-stances surrounding Libby Zion’s death determinedthat there were five contributing factors. Four factorsinvolved supervision and treatment (85). In addi-tion, the grand jury concluded that the number ofhours that interns and residents are required towork is counterproductive to providing qualitymedical care (85). While this practice may becost-effective for hospitals, providing them with acheap source of labor, the corresponding cost is adimini shed quality of health care (85). Physiciansmust be in full command of their mental faculties inorder to provide proper care for patients and tocontinue the learning process (85). The grand juryinvestigation found that major medical decisionsmade by inexperienced physicians acting alone andunsupervised can result in medical mistakes, some-times with fatal consequences. The likelihood thatsuch mistakes will occur is increased by the longhours that interns and residents must work (85).Moreover, the wee hours of the morning are whenthe attending physician coverage is routinely sparse.

Chapter 8--Case Study: Registered Nurses and Resident Physicians . 169

Hence, not only are residents tired and sleep-deprived, but they often have no senior personavailable for a second opinion or backup (10,31). Inaddition, physicians who are sleep-deprived andoverworked cannot make a commitment to continu-ity of care (9). Accordingly, the grand jury proposedthat regulations be promulgated to limit consecutiveworking hours for interns and residents. The grandjury failed to hand down any criminal indictmentsconcerning Libby Zion’s treatment and subsequentdeath, but it severely criticized the graduate medicaleducation system in New York and made fiverecommendations to improve the system. Concern-ing resident work hours, the grand jury recom-mended that the State department of healthpromulgate regulations to limit consecutive work-ing hours for interns and junior residents inteaching hospitals.

In response to the grand jury and other public andpolitical pressures, the New York State Departmentof Health convened an Ad Hoc Advisory Committeeon Emergency Services, which issued 19 recommen-dations to the health department (73). Three of theserelated to residents’ hours:

●

●

●

a 12-hour limit was to be placed on emergencyroom shifts;residents in acute care specialties not onemergency room services should work no morethan 80 hours per week averaged over a 4-weekperiod and should not be scheduled to work asa matter of course for more than 24 consecutivehours, with one 24-hour period of nonworkingtime per week; andno moonlighting would be allowed.

The commissioner of health promulgated regula-tions in 1988 modifying the overall organization andoperations of hospitals and staffs, including supervi-sion of residents and their hours:

●

●

as of October 1, 1988, the emergency room shiftfor both house staff and attending physiciansshould not exceed 12 hours (there are certainexceptions if the hospital meets certain cri-teria); andas of July 1, 1989, the scheduled workweek forother house staff in designated acute carespecialties should not exceed an average of 80hours over a 4-week period. No scheduled shiftshould exceed 24 consecutive hours. On-duty

Photo credit: George Washington University Media lCenter

Surgery is the specialty with the most demanding hours.

assignments should be separated by at least8 hours, with one 24-hour nonworking periodper week.5

The cost of implementing these new regula-tions was determined to be $226 million by theNew York Health Department, an increase ofabout 3 percent in its annual budget. The hospitalsasked for additional funding. Approximately $65million is needed for additional ancillary support—phlebotomists (technicians who withdraw bloodfrom patients’ veins), housekeepers, intravenousteams, messengers, and transport teams. It has beenreported that this ancillary support work can take upas much as 50 percent of a resident’s work hours(31). In some teaching hospitals, 70 to 80 percent ofa resident’s working hours at night have been foundto be devoted to such work (4). To replace workhours lost from resident physicians, $80 million is

s~me were special guidelines for smgew.


needed for additional residents, nurses, attendingphysicians, and physician assistants. The balance isto cover the costs of new supervision requirements,which are also part of the new regulations. It isestimated that the cost will be $3.1 billion over thenext 10 years and $5.7 billion over the next 15years. This cost was to be distributed between theState and other third-party payers. Medicare (theFederal program for the aged and disabled) was notobligated to pay its proportional share of the cost(99). The health commissioner allocated Medicarecosts to all other payers controlled by the State—Medicaid, Blue Cross, and the commercial payers(99). Blue Cross subsequently filed a successful suitto avoid payment of the Medicare costs apportionedto it (72). The commercial payers, and the State’sMedicaid program, also refused to pay the Medicareshare apportioned to them, leaving the system$50 to $60 million short (99).

Although the intent of the regulations was to limitresidents’ work hours, the wording of the regulationsis that no resident should be scheduled to work morethan 12 hours in the emergency room or scheduledto work more than 80 hours in a 4-week period (10).It has been reported that some hospitals in New Yorkhave not fully implemented the new regulations, andsome residents are still working hours in violation ofthem (38). Not enough time has elapsed to determinewhether the regulations will have any effect on thenumber of malpractice incidents in New York, sincea malpractice action may be filed several years afterthe alleged incident, and the regulations have beenin effect for only 2 years. No broad-based analysis ofthe effect of the preexisting working hours onsuccessful malpractice claims or quality of care wasever done (99).

While there have been some legislative andregulatory efforts involving the work hours ofresidents throughout the country, there have been nomandated or involuntarily enforced changes in anyState other than New York (99). Massachusetts’teaching hospitals have undertaken a voluntaryeffort to change the working conditions of residentsin that State (99). Table 8-2 discusses the currentstatus of residents’ work hours in some States. Box8-B discusses limitations on residents’ hours inGreat Britain. Box 8-C explains reductions inresidents’ hours in New Zealand. And figure 8-1shows the percentage of American hospitals, byownership, that have implemented policies whichset a maximum number of resident work hours.

Figure 8-l—Prevalence of Hospitals With Policieson Residents’ H-ours

Percent100

8 0 - r-1

All COTH V A Church Other Municipal Staten ■ 307 n ■ 55 n . 30 non- n ■ 41

profitn D 57

= 1988-89 ~ 1989-90 m 1 9 9 0 - 9 1

Percent of Council on Teaching Hospitals (COTH) with hours forhouse staff, by ownership, 1988-91.SOURCES: “Council on Teaching Hospitals Survey Shows Increase in

Stipends, Prevalence in Housestaff Hours Policies,” Associa-tion of American Medical Colleges W-k/y Report, Oct. 19,1990, p. 1; S. Zimmerman, “VA Leads COTH Hospitals inResident Hours Policies,” COTHReport24:5, 1990.

Current Status of Graduate MedicalEducation Programs

As of January 1,1990, there were 6,591 accreditedresident training programs in the United States in24 major specialties and 38 subspecialties (94). Ofthe approximately 85,330 residents in training pro-grams, nearly 40 percent were in training for internalmedicine, family practice, or pediatrics (94).

Regulation of Residents' Hours

Regulation of residents’ hours has traditionallybeen under the control of the Accreditation Councilfor Graduate Medical Education (ACGME), whichhas five member organizations:

American Board of Medical Specialties;American Hospital Association;American Medical Association;Association of American Medical Colleges;andCouncil of Medical Specialty Societies.

Each member organization appoints four representa-tives to ACGME, and the Council also has a residentrepresentative, a public representative, and a nonvot-ing representative of the Federal Government. ACGMEsets standards and procedures for accreditation that


Table 8-2—State Legislative Action Regarding Residents’ Work Hours

California: The legislature is vigorously debating whether to Michigan: There are plans to reintroduce a bill that died in 1990.control hours by legislation or through voluntary restrictions The State’s hospital association contends that such legislationenforced by the medical schools themselves. An hours bill, is not necessary, since all but three teaching hospitals alreadycalling for an 80-hour workweek with 2 days off every 2 have policies on this issue.weeks and no more than 1 day in 3 on call, was defeated inthe Senate in 1990 but will be considered by the Assembly

Minnesota: The legislature has no active bill, but the topic is being

Ways and Means Committee in January 1992.overseen by two senators, one of whom may reintroduce anearlier bill if not satisfied with residents’ working conditions.

Connecticut: A bill was introduced in March 1988 but it did not Missouri: The legislature is not considering a bill during the 1991get out of committee. A new version was introduced in 1991. term. A 1988 bill regarding residents’ hours and working

Hawaii: There has been no active bill in 3 years. A bill was renditions died in committee.introduced in 1988 but will not be reintroduced since the Nevada: The University of Nevada School of Medicine at Reno,sponsor now believes that ACGME guidelines will be the State’s only academic medical center, has executed itssufficient to control residents’ hours. own hours limitations on residents.

Illinois: A bill was introduced in 1988 that would have requiredthe Department of Public Health to examine residents’

New Jersey: The legislature has never considered a bill regarding

working conditions, but the Illinois Hospital Associationresidents’ hours. The State health department discussed

lobbied strongly against it. The sponsor of the bill ispossible guidelines several years ago but never implementedchanges.

considering reintroducing it in a future session.Iowa: There are no plans to introduce legislation in this area, but

New York: It is still the only State that has implemented rules

in 1988 a preliminary request for a draft bill on residents’regarding residents’ hours and supervision. New York passed

working conditions was made to the legislative serviceits law administratively in 1989, while the legislature merely

bureau.passed a bill for its funding. During 1990 the State began toenforce the rules.

Massachusetts: For the first time in 3 years, the legislaturedoes not plan to consider a bill in this area. Legislators say

Pennsylvania: For the first time in 3 years, no legislator has plans

they are optimistic about proposed regulations by the State’sto introduce a bill on this issue. No reason for this change wasgiven.

medical centers themselves.SOURCES: Office of Technology Assessment, 1991; based on L. Page, “Resident Hours Legislation Has Slowed Down,” American Medical News, Mar. 11,

1991, pp. 49-52.

BoX 8-B—Limitations on Residents’ Hours in Great Britain

In an effort to end the potential dangers posed by overworked resident physicians, the English minister forhealth recently announced that the Ministerial Group on Junior Doctors’ Hours has developed an agreement toreduce the maximum workweek for resident physicians to 72 hours. The Ministerial Group includes representativesof the United Kingdom’s health departments, National Health Service management, consultant physicians, residentphysicians, and medical colleges.

The agreement calls for a short-term goal of 83 hours of work per week and a long-term goal of 72 hours ofwork per week for all residents, especially those in the most overworked positions. In the first step of the agreement’senactment, the government will fund the creation of an additional 200 consultant and 50 staff grade positions during1991 and 1992. Next, the agreement will be translated into practical measures to be carried out on the local level.It calls for the immediate formation of regional committees to analyze local problems and suggest solutions.

The agreement strongly supports various types of shift work, especially in busy areas, and lists several methodsfor change. These include the sharing of duties by residents, additional coverage in the busiest posts, and areorganization of work schedules to allow for more part-time doctors and time off after night duty. Examining howto maximize the use of other staff, such as nurses and midwives, will also be important.

This announcement came 2 days before the English Court of Appeal refused to dismiss a claim by a doctorwho sued the Bloomsburg Health Authority for damages caused by the health authority’s requirement that he workat least 88 hours a week. The doctor claimed that these hours were unbearable and that they deprived him ofnecessary sleep to the point of posing a danger to his own health and that of his patients.

SOURCES: officttof Technology Assessrneng 1991; J. BurkhsrL “Englaud Imposes 72-HouI Limit on Residents,’* Jourwdof theA~rica~Medicd Association 265:2742, 1991. ,.


Box 8-C—Reduction of Residents' Hours in New Zealand

In 1985, the New Zealand Resident Medical Officers’ Association was successful in reducing residents’ hours-from 100 per week to a maximum of 72 hours per week. In April 1988, three doctors who aided in implementingthese changes spoke at a convention of the National Federation of Housestaff Organizations in Boston. Theyexplained the reasons for this reduction, how the new hours system is structured, and the results of the decreasedhours:Rationale Behind the Reduced Hours:

The system as it existed before 1985 produced sleep-deprived doctors who were unfairly paid and open toabuse by hospital administrations.Doctors who were the most sleep-deprived tended to work the longest hours and care for the sickest patients.Studies showed that two-thirds of residents were working over 72 hours per week and a large number wereworking over 90 hours per week.The ability to learn and the motivation to work were reduced by insufficient sleep,Continuity of care is already broken by changes in shifts; therefore, the question becomes one of when itis best to break the continuity, not whether to break it.Other professions that take responsibility for human lives and safety (e.g., airline pilots) have regulated theirworkers through shortened shifts.Generally, the night shift is the quietest, and residents’ work between midnight and 8 a.m. is focused on newpatients and organizing work rather than on continuing work on old patients.Quantity of residents’ work does not equal quality.The benefit of shortened hours to patients greatly outweighs administrative difficulties in implementingdifferent schedules.

Limits on Hours:A resident’s shift may be no longer than 16 hours, and there must be an 8-hour break before the nextscheduled shift can begin.Residents may work a maximum of 72 hours per week averaged over no more than 4 weeks.Emergency rooms have been targeted as high-pressure areas; thus, a resident’s shift there can last no longerthan 10 hours, and there must be a 9-hour break before the next shift begins.Emergency room residents may work a maximum of 60 hours per week, with a 50-hour-per-week averageover 3 months.Under normal circumstances, only one shift may be worked per day.Adequate handover time must be provided between shifts.No resident shall be on duty or on call for more than 12 days without a scheduled rest period of at least48 hours.Exceptions may be made on a case-by-case basis, but in considering extensions of hours, priority must begiven to adequacy of rest and sleep for the resident involved.

Results:*

●

☛

☛

●

●

☛

●

☛

Residents’ hours have decreased significantly: a study conducted in Auckland showed approximately1 percent of residents working 72 hours per week; the average number of hours worked now is 54 per week,a 19 percent decrease.Contrary to predictions before the regulations went into effect, few problems with scheduling have arisen.Doctors are more awake, alert, and enthusiastic about their jobs.Patients have noted improved care, since they are no longer treated by sleep-deprived doctors.There is nothing to stop a resident from following a particular patient, but he or she does so without theongoing responsibility for that patient’s care.Productivity has increased, especially on the night shift.Residents on the night shift are expected to do more, such as keep up patient records, than they were whenworking longer hours.Nurses find the residents easier to work with, especially on the night shift.Residents have more time with their families and have time to pursue outside interests.

SOURCES: 0.ff3ce of Technology Assessment, 1991; J. Cooper, J. Mawsom and B. SWi.nbq past presidents, New Zealand Rmidtmt MedicalOftlcers’ Associatkxl exeerptsfrorapresentations given at the National Federation of HousestaffOrgardzations, Bosto% MA, April1988; J. Stoke, Acting Chief Medical Officer, Workforce Development, Department of Healm Wellington New Zealand, personalcommunication June 12, 1991.

Chapter 8--Case Study: Registered Nurses and Resident Physicians . 173

apply to all training programs for residents. Thesestandards have the force of law, since residenttraining programs cannot operate without ACGMEaccreditation (76). Under ACGME rules, residenttraining programs must be in “substantial compli-ance’ with standards in order for their accreditationto continue.

The American Medical Association (AMA) Houseof Delegates passed two resolutions in early 1988regarding conditions in resident training programs,which are monitored by a Residency Review Com-mittee (RRC) in each of the 24 specialties:

one asking each RRC to develop guidelines forresidents’ work hours and conditions, takinginto account the intensity of the particularservice environment; andone directing each RRC to develop guidelinesregarding standards for achieving clinical com-petence, with a specific focus on weekly orcontinuous work hours, adequate time off, anaverage of no more than every third night oncall, and one 24-hour period off every 7 days.(106).

Of the 24 specialty boards, 16 have adopted specificlanguage addressing residents’ work hours. Theother eight are in areas where there is little nightwork (e.g., nuclear medicine, physical medicine andrehabilitation, radiation oncology, preventive medi-cine), thus the issues of extended hours and night callare not as great.

Accreditation Council for Graduate MedicalEducation

As described above, within the past few years, theACGME has moved to revise its requirements forprograms regarding supervision and hours and hasdirected the RRCs to introduce new standards onresidents’ work hours. Responses of the RRCsranged from specific (emergency medicine mandat-ing no more than 60 hours per week, with 12-hourshifts) to general (thoracic surgery giving theprogram director responsibility for ensuring reason-able in-house duty hours) to no response (generalsurgery and psychiatry) (see table 8-3). According tothe Association of American Medical Colleges,guidelines vary because of differences in the de-mands and patient care requirements among special-ties (102). For example, the surgical RRCs areunwilling to accept the one 24-hour period free ofpatient care responsibilities because of a strong

belief that surgical residents should see their pre-and postoperative patients every day. On the otherhand, residents in pathology have little patientcontact and do not require extended hours (41).

The RRC for internal medicine, under the aus-pices of the ACGME, implemented new rules thattook effect in October 1989 (76). Internal medicineresidents must observe a maximum workweek of80 hours averaged over 4 weeks. Residents are alsoto spend, on average, 1 day a week away from thehospital, and their continuous duty in the emergencyroom should not exceed 12 hours (76). The rules callfor adequate opportunity to rest and sleep when oncall for 24 hours or more, and residents should be oncall no more than once every third night. In addition,limits were placed on the number of patientsassigned to residents.

The American Board of Medical Specialtiesrecently vetoed a proposal to let residents have 1 dayoff per week and to be on call no more than onceevery third night (77). Continuity of care was citedas the reason for such action. As a result of this veto,the Council of Medical Specialty Societies (CMSS),one of the other ACGME member organizations,recently failed to take a position on proposed workhours limits for residents. In a panel discussionbefore the CMSS deadlock vote, an AMA executivevice president argued that universal limits on hourswould show Government regulators that the house ofmedicine was taking care of the problem of residenthours and overwork (78). According to this sameAMA executive vice president, the ACGME hopesto adopt language on hours revisions that applies toall specialties, but this probably will not happen inthe near future (106).

As a means of informing medical students aboutthe work environment, including the hours require-ment, at various residencies, the AMA has instituteda computerized information system called Fellow-ship in Residency Electronic Interactive DatabaseAccess System (FREIDA). Through FREIDA, med-ical students who are selecting a graduate programcan determine the program’s:

average number of hours per week on dutyduring the first year;average number of 24-hour periods per monthoff duty;maximum number of consecutive hours permit-ted on duty; andwhether moonlighting is allowed (93).

Table 8-3-Residents’ Working Hour Requirements in the Residency Review Committee’s Updated Accreditation Guidelines

Specialty Specific hourly requirementsHours not

General guidelines addressed

Allergy and immunology Adequate backup should be provided if workload maycreate resident fatigue sufficient to jeopardize pa-tient care.

Emergency medicine

Family practice

Internal medicine

Anesthesiology 1 full day out of 7 away from hospital; on call no morethan every third night; off-duty time should besufficient to avoid fatigue.

Dermatology On average, 1 day/week should be free from hospitalduties; no more than once every third night shouldbe spent on call; rest and sleep should be providedwhen on call for at least 24 hours.

On average, no more than 60 hours/week; at least 1day/week away from hospital; no more than 12hours/shift; at least an equivalent amount of time offbetween duty periods.

Should be adequate resident staff to prevent excessiveworkloads; suggest the following: on average, 1day/week away from residency program; on call nomore than every third night; ensuring backup avail-able if there is a sudden increase in workload.

In no case should resident go off duty until proper careof patient is ensured. No more than 80 hours/weekaveraged over 4 weeks; on average, 1 day/week outof hospital; on call no more than once every thirdnight and have adequate backup; should be ade-quate opportunity and facilities to rest when on for24 hours or more.

Neurological surgery

Neurology

Child neurology

Nuclear medicine

Obstetrics-gynecology

Ophthalmology

Orthopedic surgery

Continuity of care is most important; should haveadequate backup; scheduling should avoid pro-longed and excessive duties on a regular basis.

Should be allowed 1 day/week away from hospital; oncall not more than every third night; should beadequate physician coverage.

xx

1 day/week away from hospital; on call no more thanevery third night; adequate backup must be pro-vided.

80 hours/week limit averaged over 4 weeks; 1 day/week away from hospital; on call no more than everythird night.

On average, an 80-hour/week limit for direct patientcare.

Otolaryngology

Pathology

Pediatrics

Physical medicine andrehabilitation

Plastic surgery

Preventive medicine

Psychiatry

Radiology (diagnostic,pediatric, neuroradiology)

Residents should not be required to perform exces-sively difficult or prolonged duties regularly; theyshould have the opportunity to spend an average of1 day/week free of hospital duties; on call no morethan every third night, except to maintain continuityof care; adequate opportunity and facilities for restwhen on call for 24 hours or more.

On average, 1 day/week free of hospital duties; on callno more than every third night; backup should beprovided when patient care responsibilities areheavy.

Resident should be on call on an average of every thirdor fourth night in any year; schedule should have amonthly average of 1 day/week without assignedduties in the program.

No hour limits, but should have 1 day/week free fromhospital duties; should be on call no more than onceevery third night; must recognize patient duties arenot discharged at any given time.

1 day/week off; on average, every other night off duty;annual vacation time is additional time off; adequatebackup if excessive patient care could lead toresident fatigue or jeopardize patient care; supportsystems, including counseling to minimize stressand fatigue, should be provided.

1 day/week off; on average, on call no more than onceevery third night; must have adequate backup.

Radiology (nuclear, oncology) Residents should have 1 day/week off; on call no morethan once every third night.

General surgeryThoracic surgery

Resident hours on duty should be scheduled to avoidexcessive stress and fatigue.

Residents must not be regularly required to performexcessively prolonged duties; responsibility of pro-gram director to ensure reasonable in-hospital dutyhours.

x

Responsibility of program director to ensure reasonablein-house duty hours so residents do not have toperform excessively prolonged periods of dutyregularly.

Urology On average 1 day/week free of hospital duties; on callno more than every third night; adequate opportu-nity to rest when on call for 24 hours or more;adequate backup if work volume jeopardizes qualityof patient care.

SOURCE: Office of Technology Assessment, 1991.


It has been pointed out by one resident that somespecialties, such as pediatrics, dermatology, radiol-ogy, and obstetrics-gynecology, are attracting morewomen. These women, as part of the old guard andas new recruits, are insisting on changes to make thecurrent system more manageable (31). Some pro-grams are beginning to implement a system knownas night float to assist residents with hours coverage(74,104). Night float can take several forms, butbasically the house staff on call will admit patientsuntil 10 p.m., for example, and then leave thehospital when their workups are completed. At 10p.m. the night float team comes on duty (they havebeen off duty all day) and covers admissions andcalls until the next morning, when duties will againbe turned over to the house staff. This arrangementcan also be used for a day float system. It has beenreported that not all programs have the funds toimplement this type of system (74).

Effects of Extended Duty Hours on Residents

Research on sleep deprivation began as early as1896 (79). In 1961 a researcher concluded thatinterns had little time for sleep and spent only a smallproportion of their time on direct patient care (81).Ten years later, another study reported similarfindings regarding the daytime activities of internsin California (42). A more recent study observedinternal medicine interns and residents for 5 nightsat three teaching hospitals. It determined that:

●

●

●

●

●

●

●

●

●

the interns and residents admitted betweenthree and eight patients each night;each house officer received 16 to 25 calls pernight:up to 12 percent of their time was spent doingprocedures such as inserting I.V. catheters ordrawing blood specimens, most of which couldhave been done by nonphysicians;from 87 to 175 minutes of on-call time werespent in direct patient evaluation;the mean time spent on each new patientevaluation ranged from 17 to 31 minutes;the mean time before the evaluation wasinterrupted ranged from 7 to 11 minutes;66 to 187 minutes per night were spent docu-menting new patient evaluations in the hospitalrecord;the average sleep time ranged from 122 to 273minutes; andthe mean time before sleep was interruptedranged from 40 to 86 minutes.

While no finding was made as to whether thesehouse officers were sleep-deprived because of theirwork schedules, the researchers concluded thathouse officers:

. spend relatively little time in direct patientcontact;

. spend considerable time charting; and● are frequently interrupted while working and

trying to sleep (59).

The results of studies examining the effects ofextended hours on residents’ performance are equiv-ocal. One study noted deficits in grammaticalreasoning in a group of five physicians after sleepdeprivation (84). These researchers found that inanother 15 subjects deprived of sleep, the ability tocompensate for fatigue led to improved performanceon a more complicated reading test; however,additional time was needed to maintain accuracy andcomplete the test. They concluded that physicianscan compensate for the effects of sleep loss in theperformance of both simple and complex psychomo-tor and cognitive functions (84). In another study,33 surgical residents were given a comprehensivepsychometric test battery (87). Results showed nodifferences in performance between sleep-deprivedand rested residents.

A recent study of the effects of sleep deprivationon the performance of residents assigned to threeservices (trauma, vascular, and cardiothoracic) of amajor teaching hospital failed to support the ideathat sleep deprivation impairs cognitive and motorperformance of residents (26). Results of this studyindicated that sleep deprivation did not affectperformance, as measured by a series of fivepsychomotor tests of visual and auditory attention,reasoning ability, spatial visualization ability, andfine motor coordination. Moreover, analysis of thecorrelation between certain sleep parameters (totalsleep and longest period of uninterrupted sleep) andperformance on each component of the psychomet-ric test battery identified changes in performance onsome tests but only insignificant effects due to sleeploss (26). The results of this study are controversialand have been challenged on the basis of themeasures and experimental design used (13,14,18,19,32,43,60,66,75,80,90,96,1 10).

Other studies have found decrements in somemeasures of performance. For example, a study inEngland found decreased mathematical abilitiesamong sleep-deprived residents (57). Other studies


Photo credit: George Washington University Medical Center

Being able to confer with other residents and attending physicians is a critical part of graduate medical education.

involving sleep-deprived interns have found in-creased errors in reading electrocardiograms (EKGs)and increased time required to complete the task(36,37,11 1). In one of these studies, a mood scalewas administered simultaneously, and results indi-cated that the rested interns felt more elation, socialaffection, egotism, and vigor and less fatigue andsadness than sleep-deprived interns (11 1). In anotherstudy, rested and fatigued anesthesiology residentsshowed a difference in response time to abnormali-ties appearing on a simulated monitor (28), while astudy of sleep-deprived and rested interns assessingspeed of information processing, decisionmaking,recent memory, and mood showed that the sleep-deprived interns had less ability to retain informa-tion, longer response latencies, and greater mooddisturbances, including hostility, than the restedinterns (46).

Family and Personal Consequences

There are a great number of stresses associatedwith graduate medical education, and workinglong duty hours is one of them. Cultivating andmaintaining a healthy relationship with a spouse orsignificant other can be difficult when a resident isconstantly tired. Recognizing that sleep deprivationcan be a stressor, one study detected no relationshipbetween gender and stress among internal medicineresidents (56), while a separate study found thatwomen residents reported more stress than menresidents but were more likely to mobilize externalsupport to cope with it (11).

Divorce and broken relationships often resultfrom the stresses of residency (23). One survey offamily practice residents determined that 41 percentbelieved their performance was impaired because of


marital problems (70). Another survey found that 17percent of residents felt that their current relation-ships were in jeopardy because of job stresses (53).

The outcome of pregnancy during residency for1,293 residents and 1,494 wives of residents whoserved as controls was the subject of a recent study(52). It was determined that there were no significantdifferences in the proportion of pregnancies endingin miscarriage, ectopic gestations, and stillbirths.The rate of voluntary termination of pregnancyamong women residents was three times higher thanamong the wives. The women residents reportedworking twice as many hours per week as the wives(in some residents’ cases more than 100 hours perweek), with pregnant residents averaging 6 to 7on-call nights per month. Premature labor requiringbed rest or hospitalization was nearly twice ascommon among the residents as among the wives, aswas preeclampsia or eclampsia. However, placentalabruption was less likely to occur in residents. Thestudy concluded that there were no significantdifferences in pregnancy outcome between womenresidents and wives of male residents.

It has been alleged that sleep deprivation is thecause of numerous automobile accidents amonghouse staff (22,92). Anxiety and depression arecommon among residents (23), and several studieshave documented incidents of significant depression(53,86,109). Women house staff may be at greaterrisk of depression than men (35), in part becauseresidents who are mothers may have primary respon-sibility for child rearing when they are not on dutyat the hospital. Drug dependency and alcoholdependency are also prevalent in the medical profes-sion (116). Many hospitals are implementing pro-grams to help ‘‘impaired physicians’ cope withstress.

While many stressors may combine to affect aresident’s home life, it is unclear that extendedduty hours in themselves are responsible fornegative consequences. No studies have focusedsolely on the impact of extended duty hours onhome life.

SUMMARY AND CONCLUSIONSBoth nursing and resident training are character-

ized by shift work. Information about shift work innursing is limited by the absence of national data onnurses’ shift prevalence and career experience withshift work; few studies of the short- or long-term

consequences of shift work for nurses’ family andsocial life, health, and work performance; and fewstudies of the impact of various schedules on thequality of patient care. These limitations are due inpart to a general lack of research attention to shiftwork. However, available data may also reflectprevailing assumptions about the effectiveness ofpay differentials in solving shift-specific staffingproblems and about the ease with which resigningnurses can be replaced, at least until recently, by asteady stream of new graduates who will be willingto work shifts early in their careers. Increaseddemand for R.N.s could prompt more health careadministrators to examine shift work policies and toexperiment with alternatives as a means of improv-ing recruitment and retention of nurses.

While more research is needed on the short- andlong-term effects of shift work on a range of nurses’health outcomes and on work performance, someadverse effects of shift work have been established.The findings of the NIOSH study associate rotatingshift work with more digestive problems, moretension and stress, and a higher rate of injury. Whilethe NIOSH study was an important examination ofa number of health consequences of shift work in alarge sample of R.N.s, it was cross-sectional, reliedprimarily on self-reported symptoms, and containedfew measures of other job attributes that could beconfounded with type of shift worked (e.g., work-load, degree of control over work-related decisions,method of shift assignment).

Furthermore, working conditions of hospital nurseshave changed since the NIOSH study was conductedin 1977. Hospital nurses today are coping withdecreased lengths of patient stay and increasedseverity of illness as a result of prospective reim-bursement of hospitals. These factors may exacer-bate work stress and health problems in general,independent of shift work, or they may interact withshift work to produce extremely high stress levelsamong those nurses working undesirable shifts andconfronting patients who require more intensivenursing care. These factors may also offset anypositive effects of alternative shift schedules as arecruitment or retention strategy. More research isneeded on various shift work issues, including theimpact of changes in the health care system onworkload, shift work, job stress, and related issues.

Graduate medical education as a training processsubjects young doctors-in-training to many stresses,


one of which is long hours. This has been the case formany generations of doctors. Some have begun toquestion the wisdom of this system, arguing thatwith advances in medical technology doctors havemany more treatment decisions to make, and thesleep deprivation and fatigue that result from longhours can reduce the resident’s ability to maketimely and appropriate decisions. The Libby Zioncase demonstrated how long hours may contribute toerrors in patient care. This case focused attention onthe issue of extended duty hours for house staff.New York has responded with limitations on sched-uling of house staff hours. The ACGME has alsoestablished guidelines for reduction of residents’hours. The problem that has yet to be resolved is howthese new changes will be funded in a health careeconomy that is financially strapped.

There have been a number of studies on extendedduty hours and sleep deprivation in medical resi-dency, but results are inconclusive. Proponents ofrevision argue that, despite study conclusions, it isdifficult to function at maximurn proficiency whenone is sleep-deprived and fatigued. Almost noresearch has been done to determine the effect thatextended duty hours have on the personal lives ofresidents.

1.

2.

3.

4.

5.

6.

7.

CHAPTER 8 REFERENCESAdams, C., 1988 Report of the Hospital NursingPersonnel Surveys (Chicago, IL: American HospitalAssociation, 1989).Alexander, J.A., “The Effects of Patient Care UnitOrganization on Nursing lhrnover, ” HeaZth CareManagement Review 13:61-72, 1988.American Academy of Nursing, Task Force onNursing Practice in Hospitals, Magnet Hospitals:Attraction and Retention of Professional Nurses(Kansas City, MO: American Nurses Association,1983). ‘American College of Surgeons, “FundamentalCharacteristics of Surgical Residency Programs,”American College of Surgeons Bulletin 73(8):22-23,1988.American Hospital Association, Report of theHospital Nursing Personnel Survey, 1987 (Chi-cago, IL: American Hospital Association, 1987).Asch, D.A., and Parker, R.M., “The Libby ZionCase: One Step Forward or Two Steps Backward?”New England Journal of Medicine 318:771-775,1988.Asken, M.J., and Raham, D.C., ‘Resident Perform-ance and Sleep Deprivation: A Review,’ Jourrud ofMedical Education 58:382-388, 1983.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Asso, D., “Cyclical Variations,” Sex Differences inHurnanPeformance, M.A. Baker (ed.) (Chichester:John Wiley & Sons, 1987).Bell, B., “The New Hospital Code and the Supervi-sion of Residents,” New York State Journal ofMedicine 88:617-619, 1988.Bell, B., Chairman, Bell Committee, New YorkNY, personal communication, April 1991.BlackWell, B., “Prevention of Impairment AmongResidents in Training,” Journal of the AmericanMedical Association 255:1177-1178, 1986.Blegen, M. A., and Mueller, C.W., “Nurses’ JobSatisfaction: A Ixmgitudinal Analysis,” Researchin Nursing and Health 10:227-237, 1987.Blocker, D., letter, Journal of the American MedicalAssociation 261:862, 1988.Bonnet, M.H., Veterans’ Administration Hospital,Long Beach, CA, personal communication, Decem-ber 1990.Brown, T., Association of American Railroads,Washington, DC, personal communication, March1991.Burke, K., “Rite of Passage for the Wrong Rea-sons,” Baltimore Sun, Sept. 25, 1990.Burke, K., Chief Resident, Department of InternalMedicine, Prince George’s Hospital Center, Chev-erly, MD, personal communication, April 1991.Burlingarne, M.W., letter, Journal of the AmericanMedical Association 261:860, 1988.Chelmowski, M.K., letter, Journal of the AmericanMedical Association 261:860, 1988.Choi, T, Jameson, H., Brekke, M., et al., “Effectson Nurse Retention: An Experiment With Schedul-ing,” Medical Care 24:1029-1043, 1986.Coffey, L. C., Skipper, J.K., Jr., and Jung, F.D.,“Nurses and Shift Work: Effects on Job Performa-nce and Job-Related Stress,’ Journal of AdvancedNursing 13:245-254, 1988.Cohen, B.D., “Must Residents Work Exhausting36-hour Shifts?” Baltimore Sun, Sept. 4, 1990.Colford, J.M., and McPhee, S.J., “The RaveledSleeve of Care: Managing the Stresses of ResidencyTraining,’ Journul of the America nMedicalAssoci-ation 261:889-893, 1989.Colligan, M.J., “Methodological and Practical Is-sues Related to Shift Work Research,” Journal ofOccupational Medicine 22:163-166, 1980.Comperatore, C.A., U.S. Army Aeromedical Re-search Laboratory, Fort Rucker, AL, personal com-munication, February 1991.Deaconson, TF., O’Hair, D.P., I_Jwy, M.F., et al.,“Sleep Deprivation and Resident Performance,”Journal of the American Medical Association 260:1721-1727, 1988.Dear, M.R., Weisman, C. S., and O’Keefe, S.,‘‘Evaluation of a ‘Contract’ Model for Professional


28.

29.

30.

31.

32.

53.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

Nursing Practice,” Health Care Management Re-view 10:65-77, 1985.Denisco, R. A., Drummond, J.N., and Gravenstein,J. S., “The Effect of Fatigue on the Performance ofa Simulated Anesthetic Monitoring Task,’ Journalof Clinical Monitoring 3:22-24, 1987.Eaton, P., and Gottselig, S., “Effects of ImngerHours, Shorter Week for Intensive Care Nurses,”Dimensions in Health Service 57(8):25-27, 1980.Editors, “Nursing Shortage Poll Report,” Nursing88 18:3341, 1988.Eghrari, J., Resident Physician, Childrens’ NationalMedical Center, Washington, DC, personal commu-nication, May 1991.Feldman, J. M., letter, Journal of the AmericanMedical Association 261:862, 1988.Finn, P., “The Effects of Shift Work on the Lives ofEmployees,” Monthly Labor Review 104:31-35,1981.Folkard, S., Mo* T.H., and Lmbban, M. C., “Short-and Imng-Terrn Adjustment of Circadian Rhythmsin ‘Permanent’ Night Nurses,’ Ergonomics 21 :785-799, 1978.Ford, C. V., and Wentz, D. K., “The Internship Year:A Study of Sleep, Mood States, and Psychophysiol-ogic Parameters, 99 SouthernMedicalJourna177: 1435-1442, 1984.Friedman, R. C., Bigger, TJ., and Kornfeld, D. S.,“The Intern and Sleep Imss,” New England Jour-nal of Medicine 285:201-203, 1971.Friedman, R. C., Kornfeld, D. S., and Bigger, TJ.,“Psychological Problems Associated With SleepDeprivation in Interns,’ Journal of Medical Educa -tion 48:436-441, 1973.Friedman, T.H., attorney, private practice, NewYork, NY, April 1991.Gadbois, C., “Women on Night Shift: Interdepen-dence of Sleep and Off-the-Job Activities,” Nightand Shljl Work: Biological and Social Aspects.Proceedings of the 5th International Symposium onNight and Shljl Work, Permanent Commission andInternational Association on Occupational Health(Oxford: Pergamon Press, 1981).Ganong, W.L., Ganong, J.M., and Harrison, E.T,“The 12-Hour Shift: Better Quality, Imwer Cost,”Journal of Nursing Administration 6~2):17-29,1976.Gienapp, J. C., Executive Secretary, AccreditationCouncil for Graduate Medical Education, Chicago,IL, personal communication, January 1991.Gillanders, W., and Heiman, M., “Time StudyComparisons of 3 Intern Programs,” Jourrud ofMedical Education 46:142-149, 1971.Goodman, S., letter, Journal of the AmericanMedical Association 261:861-862, 1988.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.59.

60.

Greater New York Hospital Association, Recom-mendations of the New York State Department ofHealth’s Ad Hoc Advisory Committee on Emer-gency Services: A Preliminary Analysis Preparedby the Greater New York Hospital Association(New York NY: August 1987).Harma, M., Ilmarinen, J., and Knauth, P., “PhysicalFitness and Other Individual Factors Relating to theShift Work Tolerance of Women,” ChronobiologyInternational 5:417-424, 1988.Hart, R.P., Buchsbaum, D.G., Wade, J. B., et al.,“Effect of Sleep Deprivation on First-Year Resi-dents’ Response Times, Memory and Mood,”Journal of Medical Education 62:940-942, 1987.Health Care Advisory Board, Nurse Recruitmentand Retention (Washington, DC: November 1987).Institute of Medicine, Nursing and Nursing Educa-tion: Public Policies and Private Actions (Washing-ton, DC: National Academy Press, 1983).Jamal, M., and Jamal, S.M., “Work and NonworkExperiences of Employees on Fixed and RotatingShifts: An Empirical Assessment,” Journal ofVocational Behavior 20:282-293, 1982.Kent, L.A., “The 440 Workweek on Trial,”American Journal of Nursing 72:683-686, 1972.Keye, W.R., “Environmental Exposure and AlteredMenstrual Function,” Reproduction: The New Fron-tier in Occupational and Environmental HealthResearch (New York, NY: Alan R. Liss, 1984).Klebanoff, M.A., Shiono, P.H., and Rhoads, G. C.,“Outcomes of Pregnancy in a National Sample ofResident Physicians,” New England Journal ofMedicine 323:1040-1045, 1990.Koran, L., and Litt, I., ‘‘House Staff Well-Being,”Western Journal of Medicine 148:97-101, 1988.Krueger, G.P., and Lieberman, H.R., U.S. ArmyResearch Institute of Environmental Medicine,Natic~ MA, personal communication, February1991.Kuchinski, B. B., “The Effects of Shift Work on theMenstrual Characteristics of Nurses,” doctoraldissertation, Johns Hopkins University, 1989.Landau, C., Hall, S., Wartman S.A., et al., “Stressesand Supports During Residency Training,” Divi-sion of Educational Measurement and Research:Research in Medical Education (Chicago, IL:American Medical Association, 1984).Leighton, K., and Livingston, M., “Fatigue inDoctors,” Lizncet 1:1280, 1983.Litwin, M. S., letter, New York Times, Jan. 10,1990.Lurie, N., Rank, B., and Parenti, C., “How DoHouse Officers Spend Their Nights?’ New EnglandJournal of Medicine 320:1673-1677, 1989.McCall, T. B., letter, Journal of the AmericanMedical Association 261:859, 1989.


61.

62.

63.

64.

65.

66.

67,

68.

69.

70.

’71.

72.

73.

74.

75.

76.

McLaughlin, B., and Mohr, P., “Approaches toShift Work in Non-Nursing Occupations and Nurs-ing,” Final Report, vol. 2: Support Studies andBackground Information, U.S. Department ofHealth and Human Services, Secretary’s Commis-sion on Nursing (Washington, DC: U.S. Govern-ment Printing Office, December 1988).Martin, A.A., attorney and Senior Resident (PGY-3), Department of Pathology, 7Mane UniversityMedical Center, New Orleans, LA, personal com-munication, March 1991.Mills, M.E., Arnold, B., and Wood, C.M., “Core-12: A Controlled Study of the Impact of 12-HourScheduling,” Nursing Research 32:356-361, 1983.Minors, D. S., and Waterhouse, J.M., “CircadianRhythm Amplitude-Is It Related to Rhythm Adjust-ment and/or Worker Motivation?” Ergonomics26:229-241, 1983.Minors, D. S., and Waterhouse, J.M., “CircadianRhythms in Deep Body Temperature, UrinaryExcretion and Alertness in Nurses on Night Work,”Ergonomics 28:1523-1530, 1985.Monk, T.H., and Buysse, D. J., letter, Journal of theAmerican Medical Association 261:860-861, 1988.Moses, E., Selected Findings From the 1988 SampleSurvey of Registered Nurses, U.S. Department ofHealth and Human Services, Health Resources andServices Administration, 1989, and unpublisheddata from the survey.National Commission on Nursing, Summary Reportand Recommendiztions (Chicago, IL: Hospital Re-search and Educational Trust, 1983).Nelson, B.J., and Blasdell, A.L., “ComparingQuality on Eight-and Twelve-Hour Shifts,” Nurs-ing Management 19(11 ):64A-64H, 1988.Nelson, F., and Henry, W., “Psychosocial FactorsSeen as Problems by Family Practice Residents andtheir Spouses,’ Journal of Family Practice 6:581-589,1978.New York Law Journal, “Report Finds BetterPractice May Have Saved Woman’s Life,” NewYork Law Journal, Jan. 27, 1987, p. 1.New York State Department of Health, FinuZReportof the New York State Ad Hoc Advisory Committeeon Emergency Services (Albany, NY: 1987).New York State Conference of Blue Cross and BlueShield Plans v. Axelrod, 145 Misc. 2d 336 (1989).Nolan, J., “New Regulations and Their CostImplications,’ paper presented at the Regulation ofPhysician Training Programs Conference, CoralGables, FL, March 1989.Nowack, D.H., and Reuben, D.B., letter, JournaZ ofthe American Medical Association 261:861, 1988.Page, L., “ACGME Review Panel Sets Up HourLimits for Internist Residents,’ American MedicalNews, Nov. 18, 1988, p. 2.

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

Page, L., “ABMS Nixes University Residents’Work Reforms,” American Medical News, Sept.28, 1990, p.1.Page, L., “Surgeons’ Votes Deadlock CMSS,”American Medical News, Apr. 1, 1991, p. 4.Patrick, G., and Gilbert, J., “Studies From thePsychological Laboratory of the University ofIowa,” Psychology Review 3:475-483, 1896.Pawel, B.R., letter, Journal of the American Medi-cal Association 261:860, 1988.Payson, H.E., Gaenslen, E. C., Jr., and Stargardter,F.L., “Time Study of an Internship on a UniversityMedical Service,’ New England Journal of Medi-cine 264:439-443, 1961.Petersdorf, R.G., Memorandum No. 88-12, Associa-tion of American Medical Colleges, Washington,DC, Mar. 8, 1988.Petersdorf, R.G., and Bentley, J., “Residents’Hours and Supervision,’ AcademicMedicine 64(4):175-181, 1989.Poulton, E. C., Hunt, G. M., Carpenter, A., et al.,“The Performance of Junior Hospital DoctorsFollowing Reduced Sleep and bng Hours ofWorlq” Ergonomics 21:279-295, 1978.Report of the Fourth Grand Jury for the April/MayTerm of 1986 Concerning the Care and Treatmentof a Patient and the Supervision of Interns andJunior Residents at a Hospital in New York County,New York Supreme Court, New York County, Dec.31,1986.Reuben, D., “Depressive Symptoms in MedicalHouse Officers: Effects of I.xwel of Training andWork Rotation,” Archives of Internal Medicine145:286-288, 1985.Reznick, R.K., and Folse, J.R., “Effect of SleepDeprivation on the Performance of Surgical Resi-dents,” American Journal of Surgery 154:520-525,1987.Rhinehart, L., writer-researcher, Health and SafetyDepartment, Service Employees International Union,letter to W. Pachter, Office of Technology Assess-ment, July 20, 1990.Rhinehart, L., writer-researcher, Health and SafetyDepartment, Service Employees International Union,personal communication, January 1991.Ritchie, K., letter, Journal of the American MedicalAssociation 261:859-860, 1988.Roberts, M., Minniclq A., Ginzberg, E., et al., WhatTo Do About the Nursing Shortage (New York M:Commonwealth Fund, 1989).Rovner, S., “The Battle Over Limiting Residents’Hours,” Washington Post, Oct. 16,1990, Health, p.42.Rowley, B.D., Director, Department of Data Sys-tems, American Medical Association, Chicago, IL,personal communication, Nov. 19, 1990.


94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

Rowley, B.D., Baldwin, D. C., McGuire, M. B., etal., “Graduate Medical Education in the UnitedStates,’ Journal of the American Medical Associa-tion 264:822-832, 1990.Scherr, L., “Residency Under Scrutiny-A CaseReport,’ American College of Physicians Observer7(7):3, 1987.Schwartz, C. E., letter, Journal of the AmericanMedical Association 261:859, 1989.Smith, M.J., Colligan, M.J., Frockt, I.J., et al.,“Occupational Injury Rates Among Nurses as aFunction of Shift Schedule,” Journal of SafetyResearch 11:181-187, 1979.Smith, R., Assistant Director, Professional, &o-nomic and General Welfare, Massachusetts NursesAssociation, letter to W. Pachter, Office of Technol-ogy Assessment, June 8, 1990.Sommers, E., Senior Vice President, Greater NewYork Hospital Association, testimony before theCommittee on Health, City Council of New YorlqNY, January 1991.Stone, M., “New York Hospital on the Spot: ThreeBaffling Deaths Jolt a Proud Institution,’ New YorkTimes, June 22, 1987, pp. 4047.Sundt, F., Registered Nurse, Georgetown Univer-sity Hospital, Washington, DC, personal communi-cation, Apr. 30, 1991.Swanson, A.G., Vice President for Graduate Medi-cal Education and Vice President for InternationalPrograms, Association of American Medical Col-leges, Washington, DC, personal communication,Apr. 25, 1991.Taste, D.L., Colligan, M. J., Skjei, E.W., et al.,Health Consequences of Shifi Work (Cincinnati,OH: U.S Department of Health and Human Serv-ices, National Institute for Occupational Safety andHealth, March 1978).Teich, D. L., Assistant Professor of Medicine,George Washington University, Washington, DC,personal communication, April 1991.Tepas, D., Professor and Director, Division ofIndustrial and Organizational Psychology, Depart-ment of Psychology, University of Connecticut,personal communication, Apr. 2, 1990.

106.

107.

108.

109.

110.

111.

1120

113.

114.

115.

116.

117.

Told, J. S., Executive Vice President, AmericanMedical Association, Chicago, IL, personal com-munication, May 1991.U.S. Department of Health and Human Services,Public Health Service, National Center for HealthStatistics, “Characteristics of Registered Nurses inNursing Homes: Preliminary Data From the 1985National Nursing Home Survey,” Advance DataFrom Vital and Health Statistics, No. 152, DHHSPub. No. (PHS)88-1250, G. Strahan (cd.) (Hyatts-ville, MD: Public Health Service, Apr. 14, 1988).U.S. Department of Health and Human Services,Secretary’s Commission on Nursing, Find Report,VOIS. 1 and 2 (Washington, DC: December 1988).Valko, R., and Clayton, P., “Depression in theInternship,” Diseases of the Nervous System 36:26-29, 1975.Walbroehl, G. S., letter, Journal of the AmericanMedical Association 261:861, 1988.Walerstein, S.J., Rosner, F., and Wallace, E.Z.,“House Staff Stress,” New York State Journal ofMedicine 89:454-457, 1989.Wallis, C., “Re-Examining the 36-Hour Day: NewYork State hads a Movement to Change the WayU.S. Doctors Are Trained,” Time, Aug. 31, 1987,p. 54.Wandelt, M. A., Pierce, P.M., and Widdowson,R. R., “Why Nurses ~ave Nursing and What CanBe Done About It,” American Journal of Nursing81:72-77, 1981.Weisman, C. S., “Recruit From Within: HospitalNurse Retention in the 1980s,” Journal of NursingAdministration 12:24-31, 1982.Weisman, C. S., Alexander, C. S., and Chase, G.A.,‘‘Determinants of Hospital Staff Nurse hover,”Medical Care 19:431443, 1981.Western Medical Journal, “Medical Staff Confer-ence: Stress and Medical Training,” Western Medi-cal Journal 142:814-819, 1985.Windle, P., Senior Associate, U.S. Congress, Officeof Technology Assessment, Washington, DC, per-sonal communication, May 1991.

case study: registered nurses and resident physicians

Documents