2land.oa

The new england journal of medicine Effect of Reducing Interns’ Work Hours Christopher P. Landrigan, M.D., M.P.H., Jeffrey M. Rothschild, M.D., M.P.H., John W. Cronin, M.D., Rainu Kaushal, M.D., M.P.H., Elisabeth Burdick, M.S., Joel T. Katz, M.D., Craig M. Lilly, M.D., Peter H. Stone, M.D., Steven W. Lockley, Ph.D., David W. Bates, M.D., and Charles A. Czeisler, Ph.D., M.D., for the Harvard Work Hours, Health and Safety Group b a c k g r o u n d
Although sleep deprivation has been shown to impair neurobehavioral performance, (C.P.L., J.W.C., S.W.L., C.A.C.), General In- few studies have measured its effects on medical errors.
ternal Medicine (J.M.R., R.K., E.B., D.W.B.),Infectious Disease (J.T.K.), Pulmonary andCritical Care Medicine (J.W.C., C.M.L.), and Cardiology (P.H.S.) and the Internal Med- We conducted a prospective, randomized study comparing the rates of serious medical icine Residency Program (J.T.K.), Depart- errors made by interns while they were working according to a traditional schedule ment of Medicine, Brigham and Women’sHospital; the Division of Sleep Medicine, with extended (24 hours or more) work shifts every other shift (an “every third night” call schedule) and while they were working according to an intervention schedule that S.W.L., C.A.C.); and the Division of General eliminated extended work shifts and reduced the number of hours worked per week.
Pediatrics, Department of Medicine, Chil-dren’s Hospital Boston and Harvard Med- Incidents were identified by means of a multidisciplinary, four-pronged approach that ical School (C.P.L.) — all in Boston. Ad- included direct, continuous observation. Two physicians who were unaware of the in- dress reprint requests to Dr. Landrigan at terns’ schedule assignments independently rated each incident.
the Division of Sleep Medicine, Depart-ment of Medicine, Brigham and Women’sHospital, 221 Longwood Ave., Boston, MA 02115, or at [email protected].
During a total of 2203 patient-days involving 634 admissions, interns made 35.9 per- cent more serious medical errors during the traditional schedule than during the in- tervention schedule (136.0 vs. 100.1 per 1000 patient-days, P<0.001), including 56.6 Copyright 2004 Massachusetts Medical Society. percent more nonintercepted serious errors (P<0.001). The total rate of serious errorson the critical care units was 22.0 percent higher during the traditional schedule thanduring the intervention schedule (193.2 vs. 158.4 per 1000 patient-days, P<0.001).
Interns made 20.8 percent more serious medication errors during the traditionalschedule than during the intervention schedule (99.7 vs. 82.5 per 1000 patient-days,P=0.03). Interns also made 5.6 times as many serious diagnostic errors during thetraditional schedule as during the intervention schedule (18.6 vs. 3.3 per 1000 patient-days, P<0.001).
c o n c l u s i o n s
Interns made substantially more serious medical errors when they worked frequentshifts of 24 hours or more than when they worked shorter shifts. Eliminating extendedwork shifts and reducing the number of hours interns work per week can reduce seri-ous medical errors in the intensive care unit.
Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. e x t e n d e d w o r k h o u r s a n d s e r i o u s m e d i c a l e r r o r s o f i n t e r n s and Safety Study from July 2002 to June 2003 in the the Journal 33 years ago, Friedman and col- medical intensive care unit (MICU) and coronary ileagues reported that interns made almost care unit (CCU) of Brigham and Women’s Hospi- twice as many errors reading electrocardiograms tal, a large academic hospital in Boston, after ap-
after an extended (24 hours or more) work shift proval by the institutional review board. The MICU
than after a night of sleep.1 More recent studies and CCU were selected for study because they are
have similarly found that surgical residents made the rotations of this internal-medicine training pro-
up to twice the number of technical errors in the gram with the longest work hours and because
performance of simulated laparoscopic surgical medical errors have been detected at higher rates
skills after working overnight than after a night of in critical care settings than in other settings.15,16
sleep.2,3 Although many prior studies have been Both units have 10 adult critical care beds. Data
methodologically limited by the use of nonvali- were not collected on patients admitted for fewer
dated self-reports on the timing of sleep and in- than four hours, patients undergoing elective aller-
adequate accounting for circadian phase and chron- gy desensitization, or the rare patients who board-
ic sleep loss, as reviewed elsewhere,4-6 the literature ed on the units but who were not cared for by the
as a whole suggests that sleep deprivation causes MICU or CCU team.
substantial decrements in physicians’ performance
of discrete neurocognitive and simulated clinical design of intervention trial
tasks.4-8 The clinical importance of sleep curtail- In collaboration with the leadership of the resi-
ment has remained unclear, however,4-6 owing to a dency program and unit directors, we designed an
lack of studies conducted in clinical care environ- intervention work schedule for interns that elimi-
ments4,9 and the possibility that scheduling inter- nated extended (24 hours or more) work shifts and
ventions designed to mitigate sleep deprivation reduced the number of scheduled hours of work
may simultaneously introduce discontinuities in to 63 per week (Fig. 1). The traditional MICU house-
care.10,11
staff team consisted of three interns and three third- Within hospitals, of all trainees, interns (post- year residents, whereas the CCU team consisted of graduate year 1) typically work the greatest num- three interns and two second-year residents. Eachber of hours per week.12,13 The extended (24 hours intern and resident on these teams worked over-or more) work shifts and long workweeks of in- night in the hospital every third night. A residentterns may make them especially prone to fatigue- from another hospital service assumed patient careinduced errors. In a survey of house officers, 41 responsibilities in the CCU on nights when neitherpercent reported fatigue as a cause of their most se- of the daytime CCU residents was working. Underrious mistake. Most of these events occurred while this rotation, interns’ scheduled workweeks aver-they were interns, and 31 percent reportedly re- aged 77 to 81 hours, depending on the clinic assign-sulted in fatalities.14 ment, with up to 34 continuous hours of scheduled To understand the effects of interns’ sleep dep- work when clinic occurred after they were on call rivation on serious medical errors, we conducted (Fig. 1A).
a comprehensive comparison of errors while interns During the intervention schedule, interns’ work followed a traditional work schedule and errors hours and overnight work schedules were changed.
while they followed an intervention work schedule Interns’ traditional extended work shifts were di-that was designed to reduce sleep deprivation. Our vided in two: a “day-call” intern worked the firstgoals were to compare the rates of serious errors half of a traditional call (from 7 a.m. to 10 p.m.);directly involving interns on the two schedules, since a “night-call” intern worked the second half (frominterns were the focus of our scheduling interven- 9 p.m. to 1 p.m. the following day). To effect thistion, and to compare the overall rates of serious schedule, four interns shared patient care respon-medical errors in order to track the effects of in- sibilities during the rotation. The maximal sched-terns’ schedules on the system as a whole.
uled hours of work were 60 to 63 per week, withconsecutive hours of work limited to approximate-ly 16 hours (Fig. 1B). The intervention did not alter the schedules or staffing of second- or third-year The Intern Sleep and Patient Safety Study was con- residents or other clinical personnel. ducted as part of the Harvard Work Hours, Health Our goal was to improve interns’ opportunities Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. The new england journal of medicine to sleep while minimizing errors due to handoffs After providing written informed consent, in- of patient care and cross-coverage.10 To minimize terns were randomly assigned to work either thecross-coverage errors, we developed a sign-out intervention schedule in the CCU and the tradi-template for interns to use in all critical care rota- tional schedule in the MICU or the converse; thesetions (both intervention and traditional sched- rotations were distributed throughout the year.
ules)17 and incorporated an hour of overlap in the Data collected during a pilot intervention sched-evenings for interns on the intervention schedule ule involving four interns that was discontinuedto sign out formally (see Figure A of the Supple- after the first ICU rotation were not included. Asmentary Appendix, available with the full text of detailed in the article by Lockley et al. in this issuethis article at www.nejm.org) under the supervi- of the Journal,18 although actual work hours oftension of the senior resident.
exceeded those scheduled during both the tradi- A Traditional Schedule
Wednesday
Thursday
Saturday
Clock Time
B Intervention Schedule
Wednesday
Thursday
Saturday
Clock Time
Figure 1. Representative Work Hours during a Single Week for the Whole Team of Interns during the Traditional Schedule
(Panel A) and the Intervention Schedule (Panel B).
Scheduled work hours are indicated by the bars. Panel A shows the traditional rotation in which a team of three interns provided continuous coverage on a repeated three-day schedule, consisting of a daytime “swing” shift on day 1 (7 a.m. to 3 p.m.) (e.g., Wednesday for Intern A), followed by an extended on-call shift from 7 a.m. on day 2 to noon on day 3 (e.g., Thursday through Friday for Intern A). Interns had the day off when a swing shift was to occur on a Saturday, Sunday, or Monday (e.g., Saturday for Intern A). Interns staffed weekly ambulatory clinics when they coincided with a swing shift or the latter half of an extended on-call shift. Panel B shows the intervention rotation in which a team of four interns provided continuous coverage on a repeated four-day schedule. Day 1 is the standard swing shift (e.g., Wednes-day for Intern 1); day 2 is “day call” from 7 a.m. to 10 p.m. (representing the first half of the traditional call) (e.g., Thurs-day for Intern 1); days 3 and 4 are “night call” from 9 p.m. on day 3 to 1 p.m. on day 4 (representing the second half of a traditional call) (e.g., Friday through Saturday for Intern 1). There is a one-hour scheduled overlap between the outgoing day-call intern and the incoming night-call intern (e.g., Wednesday from 9 p.m. to 10 p.m. for Intern 4 and Intern 3, re-spectively); this overlap was often extended as clinically required. Interns had the day off when a swing shift was to occur on a Saturday, Sunday, or Monday (e.g., Sunday for Intern 1). Interns only attended clinics when they coincided with the swing day. Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. e x t e n d e d w o r k h o u r s a n d s e r i o u s m e d i c a l e r r o r s o f i n t e r n s tional and intervention schedules, the intervention review, voluntary reports, and computerized event-successfully eliminated shifts of 24 hours or more, detection monitoring. Other methods of data col-reduced the number of hours worked by interns by lection, though less comprehensive, were designednearly 20 per week, increased the average daily du- to identify all serious medical errors — both those inration of sleep by nearly an hour, and reduced at- which interns were involved and those in which theytentional failures.
were not involved. Before beginning data collec-tion, all staff received intensive training in the con- d a t a c o l l e c t i o n a n d c l a s s i f i c a t i o n
sistent, objective collection of data using standard- To measure patients’ safety during the two sched- ized forms. Because it was not possible to blind dataules, we developed an intensive system of data col- collectors to the study schedule, determinationslection and evaluation that expanded on methods of the preventability and classification of eventspreviously used in the study of medication er- were not made by the primary data collectors. In-rors16,19 and also included continuous observation stead, each suspected error or adverse event iden-of interns by physicians. In this study, we focused tified was independently rated by two physicianon procedural and diagnostic errors in addition to investigators who were unaware of the identity ofmedication errors. The definitions used to classify those involved or whether the incident occurredincidents are provided in Table 1.
during the traditional or intervention schedule.
A team of two nurse chart reviewers and six phy- In the vast majority of cases, the serious errors sician observers collected data, supplemented by identified by observers were promptly addressedvoluntary reports from clinical staff and a comput- by medical staff with no need for action on the parterized event-detection monitor. Direct observation of the observers. Nonintercepted serious errorswas the principal means of detecting serious errors were generally detected by observers when theyin which interns were directly involved; physician were discussed by clinical staff. In the handful ofobservers followed study interns continuously, day cases in which observers identified possible er-and night in the hospital. In the afternoons after rors in the making with substantial potential towork rounds, when more than one intern was work- cause harm, they immediately alerted clinical staffing simultaneously, only one intern was observed to prevent harm to the patient.
at a time owing to staffing limitations. Residents Blinded reviewers categorized each incident as and other personnel on the units were not directly an adverse event, nonintercepted serious error, in-observed. Data collection for personnel other than tercepted serious error, or error with little potentialinterns was less comprehensive and relied on chart for harm (a category that was excluded from the Table 1. Definitions Used in the Study.
Definition
Any error in the delivery of medical care, whether harmful or trivial A medical error that causes harm or has substantial potential to cause harm, in- cluding preventable adverse events, nonintercepted serious errors, and inter-cepted serious errors, but not including errors with little or no potential for harm or unpreventable adverse events A serious medical error that is intercepted before reaching the patient A serious medical error that is not intercepted and therefore reaches the patient Unavoidable injury resulting from appropriate medical care Injury due to a nonintercepted serious error in medical management A serious medical error related to the ordering or administration of pharmaceu- tical agents, blood products, or intravenous fluids A serious medical error related to the performance of an invasive procedure, such as placement of a central venous or arterial catheter A serious medical error related to history taking, the performance of a physical examination, or the ordering or interpretation of a diagnostic test Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. The new england journal of medicine analysis) and rated the preventability of adverse 80 percent power to detect a 16 percent differenceevents using a Likert scale (was prevented, was def- in the rate of serious errors between groups.
initely preventable, was probably preventable, was We evaluated the reliability of the primary data- probably not preventable, or was definitely not pre- collection process by conducting dual direct ob-ventable); the preventability scale was dichoto- servation for a total of 10 patient-days; there wasmized to include only “preventable events” and 82 percent agreement between independent ob-“nonpreventable events” before analysis. Events servers with respect to the occurrence of a seriousdeemed more likely to be due to patients’ underly- medical error. At the review stage conducted by theing illness than to medical therapy were excluded. blinded investigators, we performed comprehen-Disagreements were resolved by discussion; the sive reliability testing of all incidents rated using theinterrater reliability was calculated before such kappa statistic. For reviewers’ judgments aboutdiscussion by means of the kappa statistic, as de- whether an incident was an adverse event, an inter-scribed below.
cepted serious error, a nonintercepted serious error,or an excluded event, the k was 0.90; the k was s t a t i s t i c a l a n a l y s i s
0.80 for the preventability of adverse events.
Patients’ characteristics and the mean daily cen-sus of the units during the intervention and tradi- tional rotations were compared by means of Fish-
er’s exact test; Wilcoxon’s nonparametric test for patient population
dichotomous, nonnormally distributed continu- The study involved 2203 patient-days (1294 during
ous variables; and a t-test for normally distributed the traditional schedule and 909 during the inter-
continuous variables. All statistical tests were two- vention schedule), representing 634 admissions to
tailed. The rates of diagnostic tests and procedures the units (385 during the traditional schedule and
per patient-day were compared between the two 249 during the intervention schedule) and 5888
schedules, and the distribution was assumed to be hours of direct observation of interns. The patients’
binomial. We compared the rates of medication characteristics and the units’ characteristics were
orders per patient-day between the two schedules, very similar during the traditional and interven-
assuming a Poisson distribution, since the pres- tion schedules (Table 2). The number of days in-
ence of rates of more than one order per patient- cluded in the traditional schedule exceeded that of
day precluded the use of the binomial distribution. the intervention schedule primarily because four in-
We compared the rates of intern-associated se- terns were required for the intervention schedule as rious medical errors per patient-day (for all interns compared with only three for the traditional sched-
combined) and of total serious medical errors per ule. Since all interns rotated through both sched-
patient-day between the intervention and tradition- ules, more traditional than intervention rotations
al schedules, assuming a binomial distribution. The were required to allow each intern to spend three
rates of all serious medical errors include all intern- weeks on each schedule. The patients’ length of stay
associated serious errors (those detected by direct and mortality rate did not differ significantly be-
observation and other methods) plus non–intern- tween the two schedules.
associated errors (identified by chart review, staff
reports, and the computerized monitor). We also serious medical errors by interns
compared the rates of type-specific errors (medi- Interns made 35.9 percent more serious medical
cation, procedural, and diagnostic) per patient-day, errors during the traditional schedule than dur-
assuming a binomial distribution. For all tests, two- ing the intervention schedule (136.0 vs. 100.1 per
tailed P values of less than 0.05 were considered 1000 patient-days, P<0.001) (Table 3). Interns made
to indicate statistical significance.
27.8 percent more serious errors that were inter- The study was powered to determine differenc- cepted during the traditional schedule than dur- es in rates of serious medical errors. Analyses of ing the intervention schedule (70.3 vs. 55.0 per 1000the rates of adverse events were performed, but the patient-days, P=0.02) and 56.6 percent more non-results were considered exploratory since we had intercepted serious errors that reached the patientsonly 11 percent power to detect a 25 percent dif- (44.8 vs. 28.6 per 1000 patient-days, P<0.001).
ference in intern-associated preventable adverse The rates of preventable adverse events did notevents. By contrast, the study was designed to have differ significantly between the two schedules.
Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. e x t e n d e d w o r k h o u r s a n d s e r i o u s m e d i c a l e r r o r s o f i n t e r n s a l l s e r i o u s m e d i c a l e r r o r s
Table 2. Characteristics of the Patients and the System.*
a n d a d v e r s e e v e n t s
The rate of all serious medical errors was 22.0 per- Traditional
Intervention
cent higher during the traditional schedule than Characteristic
Schedule
Schedule
during the intervention schedule (193.2 vs. 158.4 Patients
per 1000 patient-days, P<0.001) (Table 3). Intercept- ed serious errors occurred 37.2 percent more fre- quently during the traditional schedule than during the intervention schedule (95.1 vs. 69.3 per 1000 pa- tient-days, P<0.001). The overall rates of nonin- tercepted serious errors did not differ significantly between the two schedules, nor did the rates of pre- ventable adverse events. There was no significant difference in the rates of total adverse events (pre- ventable plus nonpreventable) between the tradi- tional and intervention schedules (85.0 vs. 93.5 per 1000 patient-days, P=0.31). Secondary analysis of the rates of serious medical errors in which interns were not involved revealed no significant differ- CCU and MICU
ences between the traditional schedule and the in- tervention schedule (40.2 vs. 38.5 per 1000 patient- t y p e s o f s e r i o u s m e d i c a l e r r o r s
Interns made 20.8 percent more serious medica- tion errors during the traditional schedule than during the intervention schedule (99.7 vs. 82.5 per * Plus–minus values are means ±SE.
1000 patient-days, P=0.03). Interns made 5.6 times † Scores for the Charlson comorbidity index can range from 0, indicating no se-as many serious diagnostic errors during the tra- rious coexisting conditions, to 6, indicating the presence of metastatic cancer ditional schedule as during the intervention sched- or infection with the human immunodeficiency virus.
‡ Acute Physiology and Chronic Health Evaluation (APACHE) scores can range ule (18.6 vs. 3.3 per 1000 patient-days, P<0.001).
from 0 to 71, with higher scores indicating an increased likelihood of death.
The rates of serious procedural errors among in- § Procedures performed by interns included placement (or rethreading) of cen-terns did not differ significantly between the two tral venous catheters, placement of arterial catheters, drawing of arterial blood, intubation, thoracentesis, placement of nasogastric and orogastric tubes, lumbar puncture, and removal of central catheters or tubes.
Analysis of the types of all errors (errors made ¶P<0.001 for the comparison with the traditional schedule.
by interns plus errors in which interns were not ¿ Interpretations of diagnostic tests by interns included interpretation of chest radiographs, other radiographs, electrocardiograms, and arterial blood gas involved) showed similar patterns (Table 3). Seri- ous medication errors occurred 17.1 percent morefrequently during the traditional schedule than dur-ing the intervention schedule (135.2 vs. 115.5 per1000 patient-days, P=0.03). The rates of serious procedural errors did not differ significantly be-tween the two schedules. Serious diagnostic errors Interns made 36 percent more serious medical er-were nearly twice as common during the tradition- rors during a traditional work schedule than dur-al schedule as during the intervention schedule ing an intervention schedule that eliminated extend-(21.6 vs. 11.0 per 1000 patient-days, P<0.001).
ed work shifts. These included significantly more Examples of each type of serious medical error serious medication errors and 5.6 times as many and adverse event observed in the study are pro- serious diagnostic errors. As a consequence, thevided in Table 4. Subcategories of medication and overall rates of serious medical errors were signifi-nonmedication errors are available in Table A of cantly higher during the traditional schedule thanthe Supplementary Appendix.
during the intervention schedule. Fortunately, most Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. The new england journal of medicine of serious medical errors by interns will translate Table 3. Incidence of Serious Medical Errors.
into a reduction in the rate of adverse events. Traditional
Intervention
The prospective, randomized nature of this Variable
Schedule
Schedule
study allowed for a rigorous evaluation of the ef- fects on patients’ safety of an intervention designed to improve interns’ sleep and thus decrease medi- Serious medical errors made
cal errors. Prior studies using before-and-after co- by interns
hort designs to assess the effects of scheduling in- terventions have provided limited and conflicting data. A before-and-after analysis of a scheduling in- tervention in one hospital that reduced residents’ work hours and decreased cross-coverage of unfa- miliar patients by senior residents found that the efficiency of care increased and the rates of errors among residents decreased.20 In contrast, an un- blinded, retrospective study of a New York State reg- ulation that decreased the number of hours worked by house staff but increased cross-coverage found All serious medical errors, unit-wide
that the efficiency of care declined and rates of med- ical complications increased.11 Each was limited by a before-and-after design, which precluded the exclusion of secular trends, increasing experience of house staff, cohort effects, or other external con-founders as possible explanations for the changes.
Because of concurrent changes in work hours, cross- coverage, and other aspects of care in these stud- ies, it was not possible to identify the elements that may have been responsible for the findings.
The overall incidence of serious errors and ad- verse events we detected is similar to that reportedin other studies of patients’ safety in the ICU. For ex-ample, Giraud et al.21 and Rubins and Moskowitz22 serious medical errors were either intercepted or documented the occurrence of 13 to 40 prevent-did not result in clinically detectable harm to the able adverse events per 1000 patient-days. The Har-patient. Although the study was not designed to vard Medical Practice Study23 reported lower rateshave sufficient statistical power to detect a differ- but used a less comprehensive method of data col-ence in preventable adverse events, the incidence of lection and a more restrictive definition of harm,intern-associated preventable adverse events was since it sought to detect injuries due to negligence.
27 percent higher during the traditional schedule Donchin et al. reported a higher rate of 1.7 errorsthan during the intervention schedule, a difference per patient-day but included errors with little poten-that was not statistically significant (20.9 vs. 16.5 tial for harm.15 The rates detected by Donchin et al.
per 1000 patient-days, P=0.21). The overall rates may also be higher because they focused on errorsof preventable adverse events (intern-associated in the unit as a whole, whereas we directly observedand non–intern-associated) were not significantly only interns. Moreover, during daytime hours, whendifferent during the traditional and intervention two or more interns were working simultaneouslyschedules (38.6 and 38.5 per 1000 patient-days, re- in different parts of the units, our staffing limita-spectively; P=0.91), although our intervention and tions allowed us to observe only one intern at a time.
observations were focused on the interns. This study Consequently, the true rate of serious errors in thewas not designed or powered to assess comprehen- units as a whole may have been higher.
sively the effect of the intervention on adverse event The article by Lockley et al.18 demonstrates that rates in the units as a whole. Therefore, it remains eliminating extended work shifts and reducing theto be determined whether the decrease in the rate number of hours worked by interns led to signifi- Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. e x t e n d e d w o r k h o u r s a n d s e r i o u s m e d i c a l e r r o r s o f i n t e r n s Table 4. Examples of Serious Medical Errors and Nonpreventable Adverse Events.
Category and Type
Description
As intern is preparing to perform a thoracentesis on the left side of the patient’s chest, the senior resident enters the room and informs the intern that the pleu-ral effusion is on right side of the patient’s chest. Several days after a patient with a history of flash pulmonary edema is admitted for congestive heart failure, intern reports that patient is in clinically stable condi-tion, having miscalculated that 24-hour input and output volumes are well matched (positive by 20 ml). The nurse is concerned that patient seemed over-loaded with fluid and in mild respiratory distress and requests a reevaluation. A recalculation by the senior resident reveals an error by a factor of 100: the pa-tient’s input and output volume has, in fact, been positive by 2000 ml for the prior 24 hours. Furosemide is promptly administered and the patient’s symp-toms improve.
Intern orders an intravenous vasopressin drip at rate of 0.2 U/min (overdose by a factor of 10 ). Nurse intercepts the order, and the rate is changed to 0.02 U/min. Patient with defibrillator implanted on left side urgently needs central access for inotropic support. Intern inserts a central venous catheter in the left subcla-vian vein. Not recognizing that the vein contains the wire from the defibrillator, the intern is having repeated difficulty advancing the introducer. In the middle of the placement, the cardiology fellow enters and asks the intern to abort the procedure immediately. The catheter is removed before it can interfere with or dislodge the defibrillator wire. A middle-aged patient with a complete heart block is admitted to the CCU. The in- tern fails to examine the patient’s back. The following day, the patient is noted to have a well-developed erythema migrans rash on the back, consistent with the presence of Lyme disease, which is later confirmed by serologic testing. Ini-tiation of Lyme therapy is delayed. Intern orders an antibiotic for a patient with a listed allergy to the medication. One dose is given before the error is detected, but the patient does not have an al-lergic reaction.
A right-sided tension pneumothorax develops after a technical error during place- ment of a subclavian venous catheter leads to pleural-space puncture. The attending physician devised a plan to transfuse a patient for a hematocrit of <30. Despite these instructions, the intern fails to check laboratory results for 36 hours. When the laboratory results are finally checked, hematocrit is found to have been 26 in the interim. The patient has tachycardia for a protracted time as a consequence. Bradycardia and hypotension develop owing to an inadvertent overdose of a ben- Transfusion is required for severe bleeding resulting from placement of a medical- ly indicated nasogastric tube in a patient with coagulopathy. There is no error in placement or technique.
A rash related to nafcillin develops in a patient with no known drug allergies.
cant improvements in interns’ sleep and reductions dependently account for the observed differencesin attentional failures. Although causality cannot in the rates of medical errors. Our randomizedbe established, it was our a priori hypothesis that study design greatly diminished the likelihood ofincreases in sleep resulting from the elimination of hidden confounding owing to secular trends, sea-extended work shifts and reduction of work hours sonal effects, learning over the course of the year,would lead to a decrease in serious medical errors.24 or other external factors unrelated to our study.
There were no significant differences between the Before we initiated the intervention schedule, two schedules in the patients’ severity of illness or concern was expressed that decreasing the num-other individual or systemic variables that could in- ber of hours interns worked might diminish their Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. The new england journal of medicine role in the units, thereby shifting the burden of cross-coverage can improve patients’ safety. Theseorder writing and procedures and, hence, the risk gains might not be realized in systems that use ex-of errors to more senior staff. Our results did not tensive cross-coverage.
bear out this concern: the number of medications Although our intervention decreased the rate ordered and tests interpreted by interns per patient- of serious errors overall, our efforts to optimize theday did not differ significantly between the two sign-out process were only partially successful. Theschedules, and interns performed significantly computerized template was never fully adopted, andmore procedures per patient-day during the inter- the effectiveness of the planned evening sign-outvention schedule. Moreover, the error rates among was frequently suboptimal. Although some groupssenior residents and other staff members were not of interns worked successfully as teams and effec-increased during the intervention schedule. Thus, tively signed out every evening, even in the absencethe substantially lower rates of errors by interns of formal training in team management, others didduring the intervention schedule cannot be due to not. In the latter case, the night-call intern was of-shifting of errors to more senior staff.
ten unaware of historical details regarding patients The Institute of Medicine’s report “To Err Is admitted by the day-call intern and sometimes Human”25 was notably silent regarding the issue performed poorly when describing these patientsof sleep deprivation, largely because data directly on morning rounds. This led to a widespread im-linking sleep deprivation and medical error have pression that communication on the interventionbeen lacking. Our study helps to fill this knowledge schedule was problematic, making the improve-gap and provides data suggesting that the sleep ments in patients’ safety we observed all the moredeprivation associated with the traditional extend- remarkable. We suggest that future scheduling in-ed shifts of 24 hours or more worked by interns terventions address this issue by adding formalmay contribute to the high risk of medical errors in evening rounds for the entire team. Such improve-critical care units.
ments, coupled with the elimination of extended It is important to emphasize that not all inter- work shifts, could further improve patients’ safety.
ventions that reduce interns’ work hours will in- Our study has several limitations. The interven- crease interns’ sleep26 or improve patients’ safety. tion schedule improved work hours but still involvedSchedule design is a critical factor in determining shifts that were long enough to induce a numberthe extent to which around-the-clock work sched- of attentional failures that was greater than wouldules disrupt wake–sleep cycles, even when the be expected among fully rested people.18 We stud-number of weekly work hours remains the same.27 ied two ICUs in a single hospital, and our resultsFurthermore, any systemic intervention that reduc- may not be generalizable to other settings. In ad-es work hours necessarily increases either provid- dition, although our study was very large as com-ers’ workload (i.e., the number of patients covered pared with prior observational safety studies,15by a provider at any time) or the number of hand- the study was not powered to detect differences inoffs in care between medical personnel on shorter the rates of preventable adverse events. Larger-work shifts. Either can lead to increased rates of scale, multicenter trials are needed to investigateerrors and adverse events.10 “Night-float” systems, this aspect.
which use residents on night shifts to allow physi- Another important limitation was our inability cians working extended work shifts protected time to blind the medical observers to the schedule offor sleep, have their own set of risks. Night-float the interns, an issue commonly encountered in in-residents often know patients less well than do vestigations of systemic interventions to maximizeother team members (particularly if multiple resi- patients’ safety. We addressed this in two ways:dents share responsibilities as night floats over the first, we instructed observers — none of whom werecourse of a week, or if night floats are responsible study investigators — in the importance of con-for an increased number of patients), and may sistent, objective detection of serious errors, re-themselves be sleep-deprived and error-prone.28 gardless of study schedule. Second, all initial ob-For these reasons, we ultimately decided not to im- servations were also reviewed by two independentplement a night-float system as a means of reduc- investigators who were blinded to the study’s con-ing interns’ work hours, as originally planned.24 ditions and who classified incidents with extremelyOur data support the hypothesis that elimination high reliability. Nonetheless, we cannot excludeof extended work shifts in a system that minimizes the possibility that some bias may have resulted Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. e x t e n d e d w o r k h o u r s a n d s e r i o u s m e d i c a l e r r o r s o f i n t e r n s from the inability to blind the primary detection tunity to sleep in the afternoon before workingprocess, though our reliability data suggest that overnight, the schedule thereby muted the effectthis bias was probably minimal.
of circadian misalignment on performance. Medi- Notably, our data on the high incidence of in- cal or surgical simulators could help isolate the ef- tercepted serious errors in ICU settings indicate fects of these interacting factors, since the rela-that the ability of other personnel to act as interns’ tive importance of these variables remains unclear.
safety net — nurses, pharmacists, and senior med- Strategic use of a novel regimen of caffeine32 or am-ical staff — is very important in preventing injury bient light of specific intensity and wavelengths37,38to patients as a result of interns’ errors. Therefore, may further mitigate the deterioration in perfor-future studies should seek to improve and measure mance resulting from circadian misalignment.
objectively the sleep and performance of all clinical In conclusion, the rates of serious medical er- unit personnel, since team performance may criti- rors in two ICUs were lowered by eliminating ex-cally affect patients’ safety.29 Having interns on a tended work shifts and reducing the number ofdifferent schedule than supervising residents may hours interns worked each week. Our results mayhave introduced discontinuities in education and have important implications for health policy, sinceinterfered with the traditional resident–intern men- more than 100,000 physicians are currently in train-torship bond. We would recommend that future ing in the United States.39 Most of these residentsstudies investigate the effects of eliminating the are regularly scheduled to work 30-hour shifts, sinceextended work shifts of interns and senior resi- extended work shifts and long workweeks con-dents, both to avoid this problem and because it is tinue to be permitted, even under the schedulingunlikely that interns are uniquely susceptible to the reforms instituted last year by the Accreditationadverse effects of sleep deprivation.
Council for Graduate Medical Education. Further Prior interventions that have proved successful modifications of these standards, particularly with in reducing serious medical errors in ICU settings respect to the duration of work shifts, may be need-have included the use of computerized provider ed to improve patients’ safety in teaching hospitalsorder entry (CPOE)30 and on-site monitoring of nationwide.
orders by clinical pharmacists.31 The higher intern- Supported by a grant (RO1 HS12032) from the Agency for associated rate of serious medical errors during Healthcare Research and Quality (AHRQ), affording data-confiden- tiality protection by federal statute (Public Health Service Act 42 the traditional schedule, even in the presence of U.S.C.); by a grant (RO1 OH07567) from the National Institute forCPOE, clinical pharmacists, unrestricted use of caf- Occupational Safety and Health, which provided a Certificate offeine by interns,32 and a perceived increase in the Confidentiality for data protection; by the Department of Medicine, Brigham and Women’s Hospital; by the Division of Sleep Medi- risk of handoff errors, indicates the extent of im- cine, Harvard Medical School; by the Brigham and Women’s Hos-pairment associated with extended work shifts. pital; and by a General Clinical Research Center grant (M01This observation corroborates the prior experimen- RR02635) from the National Center for Research Resources. Dr.
Landrigan is the recipient of an AHRQ career development award tal finding that a single night of continuous sleep (K08 HS13333); Dr. Cronin is the recipient of an AHRQ Nationaldeprivation causes decrements in performance sim- Research Service Award (F32 HS14130) and a National Heart, Lung,ilar to those induced by a blood alcohol level of and Blood Institute fellowship in the program of training in Sleep, Circadian, and Respiratory Neurobiology at Brigham and Women’s Hospital (T32 HL079010); Dr. Lockley is the recipient of a traveling By reducing consecutive and weekly work hours, fellowship from the Wellcome Trust, United Kingdom (060018/B/ our scheduling intervention attempted to address 99/Z); and Dr. Czeisler is supported in part by the National Space Biomedical Research Institute through NASA (NCC 9-58).
both acute sleep deprivation and chronic partial We are indebted to our data-collection team, without whom sleep deprivation. By reducing interns’ sleep depri- this project could not have occurred; to Patricia Aboagye-Kumi,vation and, hence, depth of subsequent sleep, we M.D., Megan Callahan, M.D., Vilma L. Castenada, M.D., Rajneesh S.
Hazarika, M.B., B.S., Kristina Martell-Waldrop, R.N., Tamara Sik- also indirectly addressed the adverse effects of sleep harulidze, M.D., Noma Rehman, M.D., and Victor Tsveybel, R.N., forinertia (i.e., an increased tendency to err on awak- their hard work and dedication; to the house staff, attending physi-ening) on performance, since such impairment is cians, nurses, and clinical pharmacists of the CCU and MICU for their ongoing support; to Victor J. Dzau, M.D., Anthony D. Whitte- a function of sleep depth.34 The schedule was also more, M.D., Jeffrey Otten, Matthew Van Vranken, Gary L. Gottlieb,designed to attenuate the circadian performance M.D., M.B.A., and Joseph Martin, M.D., Ph.D., for their support andnadir by taking advantage of the blunting of this leadership in fostering this study; and to Laura K. Barger, Ph.D., M.P.H., Erin E. Evans, B.S., R.P.S.G.T., Heather L. Gornik, M.D., nadir that occurs when the homeostatic sleep drive and DeWitt C. Baldwin, Jr., M.D., for assistance in designing and co-is lower.35,36 By providing interns with the oppor- ordinating the many facets of this effort.
Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved. e x t e n d e d w o r k h o u r s a n d s e r i o u s m e d i c a l e r r o r s o f i n t e r n s r e f e r e n c e s
Friedman RC, Bigger JT, Kornfeld DS.
errors in the intensive care unit. Crit Care 29. Leape LL, Berwick DM, Bates DW. What
The intern and sleep loss. N Engl J Med 1971; 16. Kaushal R, Bates DW, Landrigan C, et al.
dence-based medicine meets patient safety.
Medication errors and adverse drug events Jensen P, Rosenberg J. Laparoscopic perfor- in pediatric inpatients. JAMA 2001;285:2114- 30. Bates DW, Leape LL, Cullen DJ, et al. Ef-
mance after one night on call in a surgical fect of computerized physician order entry 17. Petersen LA, Orav EJ, Teich JM, O’Neil
serious medication errors. JAMA 1998;280: sign-out program to improve continuity of et al. Effect of sleep deprivation on the per- inpatient care and prevent adverse events.
31. Leape LL, Cullen DJ, Clapp MD, et al.
formance of simulated laparoscopic surgi- Jt Comm J Qual Improv 1998;24:77-87.
cal skill. Am J Surg 2003;186:169-74.
18. Lockley SW, Cronin JW, Evans EE, et al.
rounds and adverse drug events in the inten- Effect of reducing interns’ weekly work sive care unit. JAMA 1999;282:267-70. [Erra- clinicians and the safety of patients. N Engl hours on sleep and attentional failures.
32. Wyatt JK, Cajochen C, Ritz-De Cecco A,
19. Bates DW, Boyle DL, Vander Vliet MB,
Czeisler CA, Dijk DJ. Low-dose repeated caf- I, Owens J. Sleep loss and fatigue in residency Schneider J, Leape LL. Relationship between feine administration for circadian-phase- medication errors and adverse drug events.
20. Gottlieb DJ, Parenti CM, Peterson CA,
Lofgren RP. Effect of a change in house staff 33. Dawson D, Reid K. Fatigue, alcohol and
work schedule on resource utilization and studies concerning effects of sleep depriva- 34. Dinges DF. Are you awake? Cognitive
tion and fatigue on residents’ performance.
21. Giraud T, Dhainaut JF, Vaxelaire JF, et al.
performance and reverie during the hypno- Iatrogenic complications in adult intensive pompic state. In: Bootzin R, Kihlstrom J, Pilcher JJ, Huffcutt AI. Effects of sleep care units: a prospective two-center study.
Schacter DL, eds. Sleep and cognition. Wash- deprivation on performance: a meta-analy- ington, D.C.: American Psychological Asso- 22. Rubins HB, Moskowitz MA. Complica-
Buysse DJ, Barzansky B, Dinges D, et al.
tions of care in a medical intensive care unit.
35. Jewett ME, Kronauer RE. Interactive
Sleep, fatigue, and medical training: setting mathematical models of subjective alertness an agenda for optimal learning and patient 23. Brennan T, Leape LL, Laird NM, et al. In-
and cognitive throughput in humans. J Biol cidence of adverse events and negligence in 10. Petersen LA, Brennan TA, O’Neil AC,
hospitalized patients: results of the Harvard 36. Wyatt JK, Ritz-De Cecco A, Czeisler CA,
Cook EF, Lee TH. Does housestaff disconti- Medical Practice Study I. N Engl J Med 1991; nuity of care increase the risk for prevent- tonin rhythms, sleep, and neurobehavioral able adverse events? Ann Intern Med 1994; 24. Lamberg L. Long hours, little sleep: bad
function in humans living on a 20-h day. Am medicine for physicians-in-training? JAMA 11. Laine C, Goldman L, Soukup JR, Hayes
37. Czeisler CA, Johnson MP, Duffy JF, Brown
JG. The impact of a regulation restricting 25. Kohn LT, Corrigan JM, Donaldson MS,
eds. To err is human: building a safer health bright light and darkness to treat physiologic system. Washington, D.C.: National Acade- maladaptation to night work. N Engl J Med 12. Schwartz RJ, Dubrow TJ, Rosso RF, Wil-
26. Richardson GS, Wyatt JK, Sullivan JP, et
38. Lockley SW, Brainard GC, Czeisler CA.
liams RA, Butler JA, Wilson SE. Guidelines al. Objective assessment of sleep and alert- High sensitivity of the human circadian mel- for surgical residents’ working hours: intent ness in medical house staff and the impact atonin rhythm to resetting by short wave- vs reality. Arch Surg 1992;127:778-82.
of protected time for sleep. Sleep 1996;19: length light. J Clin Endocrinol Metab 2003; 13. Baldwin DC Jr, Daugherty SR, Tsai R,
Scotti MJ Jr. A national survey of residents’ 27. Czeisler CA, Moore-Ede MC, Coleman
39. Accreditation Council for Graduate
self-reported work hours: thinking beyond RH. Rotating shift work schedules that dis- specialty. Acad Med 2003;78:1154-63.
rupt sleep are improved by applying circadi- and filled positions by specialty for the 14. Wu AW, Folkman S, McPhee SJ, Lo B. Do
an principles. Science 1982;217:460-3.
house officers learn from their mistakes? 28. Cavallo A, Ris MD, Succop P. The night
float paradigm to decrease sleep deprivation: 15. Donchin Y, Gopher D, Olin M, et al.
A look into the nature and causes of human Copyright 2004 Massachusetts Medical Society. Downloaded from www.nejm.org at ASAF HAROFE MEDICAL CENTER on September 25, 2006 . Copyright 2004 Massachusetts Medical Society. All rights reserved.

Source: http://tpeople.co.il/mirsham/he/files/ReducingWorkHoursICUs.pdf