Sleep deprivation is common among resident physicians and clinical fellows. Current evidence about sleep science, performance, shift work, and medical errors consistently demonstrates positive impact from reduction of excessive duty hours, particularly when shift length is shortened. This article provides an overview of this literature, highlighting research on diminished physician cognitive performance due to sleep deprivation and the increase in the number of medical errors that is seen under these conditions. Accreditation Council on Graduate Medical Education trainee duty hour guidelines are reviewed. Practical approaches to evidence-based scheduling of shift-work are also discussed, with attention to improving patient safety.
Key Points
- •
Sleep deprivation remains an important factor contributing to medical errors.
- •
Physicians working extended-duration shifts as well as excessive cumulative hours show a decline in cognitive performance and are more likely to commit medical errors.
- •
Schedule changes that do not reduce total cumulative hours have not consistently resulted in improvements in patient safety or patient care measures.
- •
Sleep science research can guide schedule interventions to improve patient care and safety.
Introduction
During the past decade, increased attention has been paid to the number of hours worked by medical trainees (ie, residents and fellows). Before 2003, no regulations existed in the United States regarding the frequency, duration, or total number of hours that trainees could work. However, in response to early literature demonstrating the adverse effects of sleep deprivation on resident-physicians’ ability to interpret medical tests, in the 1970s to 1990s, many programs began reducing the frequency of in-house calls—extended shifts lasting between 24 and 40 hours—from every other night call (q2) to every third night (q3) or every fourth night (q4), with 8- to 12-hour day shifts occurring on the days between the extended shifts. As these changes were implemented, there were cries from many experienced physicians extolling the virtues of long hours in the hospital. A commonly heard mantra was, “the only problem with q2 call is all of the cases you miss on your night off.”
In the 1990s, based on the available evidence regarding the effects of fatigue on performance, European governments endorsed the European Working Time Directive, which currently restricts all workers, including health care trainees and experienced physicians, to a maximum of 13 consecutive hours and 48 total hours of work per week. All time spent in the hospital by physicians counts toward this limit.
In the United States, some reductions in trainee work hours have followed, reflecting a greater understanding of sleep cycles and sleep health, with a particular focus on reducing medical errors due to sleep deprivation. Initial modest limits for all residents and fellows advanced by the Accreditation Council for Graduate Medical Education (ACGME) in 2003 were followed by a more substantive limiting of hours for interns (first-year residents), in particular in 2011.
This article reviews the impact of sleep on overall performance, discusses the context and progress of duty hours regulations as an approach to decreasing fatigue, reviews associations of duty hour changes with medical errors, and provides recommendations for optimizing duty hours in health care settings.
Introduction
During the past decade, increased attention has been paid to the number of hours worked by medical trainees (ie, residents and fellows). Before 2003, no regulations existed in the United States regarding the frequency, duration, or total number of hours that trainees could work. However, in response to early literature demonstrating the adverse effects of sleep deprivation on resident-physicians’ ability to interpret medical tests, in the 1970s to 1990s, many programs began reducing the frequency of in-house calls—extended shifts lasting between 24 and 40 hours—from every other night call (q2) to every third night (q3) or every fourth night (q4), with 8- to 12-hour day shifts occurring on the days between the extended shifts. As these changes were implemented, there were cries from many experienced physicians extolling the virtues of long hours in the hospital. A commonly heard mantra was, “the only problem with q2 call is all of the cases you miss on your night off.”
In the 1990s, based on the available evidence regarding the effects of fatigue on performance, European governments endorsed the European Working Time Directive, which currently restricts all workers, including health care trainees and experienced physicians, to a maximum of 13 consecutive hours and 48 total hours of work per week. All time spent in the hospital by physicians counts toward this limit.
In the United States, some reductions in trainee work hours have followed, reflecting a greater understanding of sleep cycles and sleep health, with a particular focus on reducing medical errors due to sleep deprivation. Initial modest limits for all residents and fellows advanced by the Accreditation Council for Graduate Medical Education (ACGME) in 2003 were followed by a more substantive limiting of hours for interns (first-year residents), in particular in 2011.
This article reviews the impact of sleep on overall performance, discusses the context and progress of duty hours regulations as an approach to decreasing fatigue, reviews associations of duty hour changes with medical errors, and provides recommendations for optimizing duty hours in health care settings.
Sleep science and performance
Human beings, including physicians, are biologically wired to have sleep–wake cycles that correspond to day and night. Circadian rhythms, which are controlled by an endogenous pacemaker in the hypothalamus, drive alertness during the day and sleepiness during the night. The rhythms can respond to external influences to reflect changing external schedules (eg, traveling across time zones), but such adjustments take days to weeks depending on the degree of change required.
Independent of the impairments in performance that are induced by working at night (when the circadian system is promoting sleep), sleep deprivation also can impair performance. Sleep deprivation is known to impair many aspects of human functioning, including many that are essential to the practice of medicine (eg, cognitive performance, memory, and fine motor skills). Meta-analysis has shown that clinical performance and vigilance are particularly impaired by sleep deprivation. The impairment of cognitive functioning has been compared to alcohol intoxication, and research has shown that the cognitive performance decline after 17 hours of wakefulness mirrors a blood alcohol concentration (BAC) of approximately 0.05%.
Chronic low levels of sleep deprivation also impair performance. After 2 weeks of getting only 6 hours of sleep, psychomotor performance impairments are equivalent to those seen in subjects who are awake for 24 hours continuously. Recent research has demonstrated that when chronic sleep deprivation compounds acute sleep deprivation and circadian misalignment, performance is far worse than that which is induced by any of these 3 factors alone ( Fig. 1 ).
As a consequence of each of these factors, shift work, even of limited duration, affects sleep. A shift worker is defined as “anyone who works extended-duration shifts and other variable and nonstandard hours, including workers who work late into the night or start working very early in the morning.” Using this definition, virtually all trainees in all fields of medicine will be considered shift workers for some part of their careers. Shift work leads to circadian misalignment and decreased sleep. When an individual remains awake for more than 24 hours continuously, acute sleep deprivation compounds the sensation of “jet lag” caused by working at an adverse circadian phase. In addition, sleep the following day is often limited and of poor quality, as the circadian system promotes wakefulness by day.
It is also important to recognize that shift work is dangerous to the individual who is working, with multiple studies reporting that shift work may contribute to health problems, including increasing the risk of motor vehicle crashes, obesity, cardiovascular disease, and cancer. Based on the strength of this evidence, the International Agency for Research on Cancer concluded that shift work is “probably carcinogenic” to humans.
Sleep deprivation and physician performance
Physicians are in no way immune to the performance effects of sleep deprivation. In a study of pediatric residents, researchers evaluated task performance in residents at the end of a month of light call (4-week daytime clinic rotations averaging 44 hours per week, with limited night work) compared with residents at the end of a month of heavy call (intensive care unit [ICU] rotations with q4-q5 call, averaging 80 to 90 hours per week). Residents on the light call rotation were also tested with alcohol ingestion to a level of 0.04% to 0.05% BAC, for comparison with residents on heavy call who were alcohol-free. Measurements included reaction time, lapses, omission errors, as well as off-road events in a driving simulator. These psychomotor tasks are similar to tasks that a trainee might be expected to engage in as part of daily patient care or returning home after completing patient care. Overall, performance was comparable between the group on the heavy call month with placebo and the light call month with alcohol intoxication.
Physicians who are fatigued are at a higher risk for on-the-job accidental injuries as well. Percutaneous injuries (eg, needle-stick lacerations), although overall rare, are twice as common during nighttime hours compared with daytime. When asked to self-report contributing factors, physicians most commonly identified fatigue and inattention. This study also reported that injuries happening during daytime hours were more likely to occur on postcall days, when residents were likely to be suffering from fatigue. Residents have also been shown to have a higher risk of nodding off or falling asleep while driving or stopping in traffic as the number of extended-duration shifts increases each month. As would be expected based on the data comparing them to intoxicated residents, they also have an increased risk of motor vehicle accidents following extended-duration shifts.
There is limited information to guide the maximum safe duration of work shifts, and many factors contribute to performance on any individual shift. These factors may include start time of the shift, number of hours of sleep in the preceding day, transitions from day to night shifts, as well as use of countermeasures such as caffeine.
Early studies of duty hours and errors
Medical errors affecting patients are more common during months with multiple extended-duration shifts, compared with those with none. The first study associating fatigue with poor performance by medical trainees was done by Friedman and colleagues in 1971. In this study, interns were given a modified sustained attention task by being asked to identify and bracket arrhythmic episodes on a running EKG strip from a patient with arrhythmias. All subjects were evaluated both after a normal night of sleep (mean 7.0 hours sleep) and after completing a 24-hour shift (mean 1.8 hours sleep). The subjects made almost twice as many errors after the prolonged shifts.
Subsequent studies have likewise found that residents working 24 hours or more are more prone to errors across a range of neurocognitive and clinical tasks. A meta-analysis of 60 such studies found that resident performance across a range of clinical tasks dropped nearly 2 standard deviations below baseline rested performance following 24 hours of acute sleep deprivation ( Fig. 2 ).
In an early intervention study from 1991, patients cared for by internal medicine residents working on a shift schedule showed improvement in length of stay and decreased likelihood of a medication error after work hour reduction.
2003 ACGME guidelines
Before 2003, there were no national limits in the United States on resident duty hours. In the face of mounting public pressure and accumulating evidence regarding the hazards of long work hours, the ACGME released duty hours requirements in 2003, which applied to all accredited training programs. These requirements had 4 major components: (1) duty hours were limited to 80 hours per week, averaged over a 4-week period (inclusive of all in-house call activities); (2) 1 day in 7 must be free from all educational and clinical responsibilities, averaged over a 4-week period, inclusive of call; (3) a 10-hour rest period was required between all daily duty periods and after in-house call; and (4) extended-duration shifts were limited to a maximum of 30 hours of continuous on-site duties. The 30-hour shift was further subdivided as follows: after the first 24 hours, no new patients could be evaluated and treated by the trainee; the final 6 hours were for continuing care of current patients as well as handoffs of care.
Studies of the 2003 duty hour standards
Most studies evaluating the effects of the 2003 standards as a whole have found that they resulted in little if any change in patient-related outcomes. Three large retrospective national cohort studies found that in teaching hospitals as compared with nonteaching hospitals, mortality remained unchanged for surgical patients and minimally improved or unchanged for medical patients. One large study examined changes after the elimination of extended-duration shifts for residents, with additional coverage provided by nocturnists (hospital-based attending physicians who primarily provide care at night). The investigators reported improvement in some patient care measures but no improvement in readmission rates, adverse medication interactions, and mortality. The study did not report a decrease in total resident hours, noting only that all residents were within the 80-hour limits.
Other studies have shown improved compliance with guidelines for quality prescribing of discharge medications, reduced mean length of stay and decreased 6-month mortality, and improved perceived quality of care by nurses and patients after transitioning to ACGME-compliant schedules.
One reason for the limited effectiveness of the ACGME’s 2003 standards is that in most programs, they led to modest reductions in actual work hours or improvements in sleep, as most programs were at baseline not far from the ACGME’s new requirements. Nationwide, work hours decreased by only 5% to 6% after implementation of the standards, and sleep improved only 22 minutes per night ; in a tricenter pediatric study, work and sleep hours did not improve at all.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
