fb tracking

Petition Requesting Medical Residents Work Hour Limits

Honorable R. David Layne
Acting Assistant Secretary for Occupational Safety and Health
U.S. Department of Labor
Occupational Safety and Health Administration – Room S2315
200 Constitution Avenue, N.W.
Washington, D.C. 20210
(202) 693-2000 

Dear Mr. Layne: 

Public Citizen, a consumer and health advocacy group with 150,000 members, the Committee of Interns and Residents (CIR), a housestaff union representing over 11,000 medical residents, the American Medical Student Association (AMSA), a national organization representing over 30,000 physicians-in-training, Bertrand Bell, M.D., Professor of Medicine at Albert Einstein College of Medicine and author of New York State Health Code 405 restricting resident work hours, and Kingman P. Strohl, M.D., Professor of Medicine and Director of the Center for Sleep Disorders Research at Case Western Reserve University, hereby petition the Occupational Safety and Health Administration (OSHA) to implement the following regulations on medical resident and fellow work hours, with the primary intent of providing more humane and safe working conditions for medical residents and fellows, which will result in a better standard of care for all patients: 

  1. a limit of 80 hours of work per week,
  2. a limit of 24 consecutive hours worked in one shift,
  3. a limit of on-call shifts to every 3rd night,
  4. a minimum of 10 hours off-duty time between shifts.
  5. at least one 24-hour period of off-duty time per week, and
  6. a limit of 12 consecutive hours on-duty per day for emergency medicine residents working in hospitals receiving more than 15,000 unscheduled patient visits per year.

The petitioners request the Acting Assistant Secretary for Occupational Safety and Health to exercise his authority under section 3(8) of the Occupational Safety and Health Act,[1] on the grounds that work hours in excess of the requested limits are physically and mentally harmful to medical residents and fellows, and that a federal work-hour standard is necessary to provide them with safe employment.  Research has connected the typical resident work schedule to harms in three specific areas: motor vehicle accidents, mental health, and pregnancy. 

American medical residents work among the highest—if not the highest—number of hours in the professional world, regularly clocking 95 and as many as 136 out of the 168 hours in a week.[2]  Working these extreme hours for years at a time, predictably, has ill-effects on personal health and safety.  Multiple studies in the medical literature demonstrate that sleep-deprived and overworked residents are at increased risk of being involved in motor vehicle collisions, suffering from depressed mood and depression, and giving birth to growth-retarded and/or premature infants.  The signers of this petition believe that these are grave health outcomes and that any system allowing its workers to be subjected to such direct threats to their well-being is seriously flawed.  For OSHA not to regulate resident work hours is to abdicate its responsibility to protect the health of those who care for the nation’s sick and dying.

In order to fulfill OSHA’s mission “to send every worker home whole and healthy every day,”[3] you must act now to address the dangers that extreme work hours pose for medical residents and fellows.  Whereas previous appeals to limit residents’ work hours have focused on the well-documented risks patients face due to tired physicians, this petition concentrates on the often-overlooked health risks faced by the residents who endure those long hours.  As grounds for the work-hour standards requested above, we utilize a review of the medical literature to (1) characterize current working conditions in residency programs in the U.S., and to (2) provide evidence of the physical and mental harm that comes to residents as a result of their long work hours.  Next, we (3) review various responses to the problem of excessive work hours, including how other industrialized countries have acted to protect their medical residents through mandated work-hour limits, and how our own federal government already protects workers and the public through hours-of-service regulations in the transportation industries.  Finally, we examine (4) the relevance of resident work-hours reduction to patient protection, (5) arguments against the reduction of resident work hours, and (6) the role of OSHA in regulating resident work hours.  


The Resident Work Schedule 

Depending on the residency, physicians-in-training can work from 60 to 130 hours a week, with only one day off per week.  Specifics come from a 1991 mail survey of second-year residents conducted by Daugherty et al., in which a random 10% of all second-year residents (all specialties) listed in the American Medical Association’s research and information database were contacted.[4]  Seventy-two percent of the residents responded, for a total sample size of 1277.  Residents reported the following work schedule characteristics: 

  • Their longest period without sleep during their first year of residency was an average of 37.6 hours (standard deviation (SD) 9.9).
  • During a typical work week, they worked an average of 56.9 total hours (SD 30.19) in on-call shifts (as distinguished from the total average number of hours they worked per week).  An on-call shift is a continuous shift at the hospital allowing for little to no sleep; two on-call shifts are typically scheduled per week.
  • 25% reported being on-call in the hospital a total of over 80 hours per week.  Surgeons reported the highest average hours of on-call time per week (72.5). 
  • On a scale of 0 (never) to 4 (almost daily), residents most frequently gave a response of 3 for the amount of sleep deprivation experienced during the first year.  Over 10% of residents indicated sleep deprivation was an “almost daily” occurrence. 

The surgical residency program at Johns Hopkins Hospital provides anecdotal evidence further characterizing the resident work schedule.  A recent feature by ABC News, entitled Hopkins 24/7,[5] publicized what is occurring at many other teaching hospitals around the country.  The Chief of Surgery at Hopkins, Dr. John Cameron, reported that the average number of work hours per week for his residents is 94.  This translates to working two 12-hour days and two 35-hour shifts, with only four uninterrupted nights of sleep and one day off a week.  Some residents interviewed by the news program claimed to have worked weeks totaling an astonishing 140 hours.  Dr. Risa Moriarity, a third-year surgical resident, explained,  

I think the worst part of training as a surgical resident is, no question, the hours.  You routinely work 110 to 130 hours a week.  And I probably don’t have to tell you that there are 168 hours in a week.  So that leaves 38 or 48 hours in a week to commute, to get your entire week’s worth of sleep, to see your spouse, eat, clean your house, pay your bills, do everything else.  And it’s nearly impossible to do that.  It takes a very altered state of mind to get through it. 

This compares to the 128 hours off from work that most Americans enjoy.   

Moreover, surgical residents at Hopkins report that they can work on-call shifts of as many as 60 consecutive hours —  working straight for a total of two nights and three days, for example, from Monday morning to Wednesday night.  On-call shifts only allow a resident to sleep when his or her patients no longer need immediate attention; this results in only a few hours of sporadic sleep — if any — for the resident on-call.  In no other industry are workers forced to endure such long shifts.  During taping of the ABC news documentary, as Dr. Moriarity neared the 60th hour of her on-call shift, she decided to leave the Hopkins residency program primarily because of its long hours.  Concerned about how she was viewing her patients, and wanting to lead a more “balanced life” with her husband, she said: 

Obviously, part of the reason we all go to medical school is to take care of people, but after being awake for 56 hours or more, all you really care about, I think — or at least all I really cared about was me.  And so I — I just decided I wasn’t willing to live that way…. an 80-hour work week would have been much more reasonable and manageable than what I was doing.… I think in the long run, an 80-hour work week is sufficient to train good surgeons and would result probably in better patient care. 

Individual experiences from the surgery program at Hopkins encompass the same problems that many medical residents face working even 85 hours a week.  Residents must deal with such diverse and stressful demands as the responsibilities of patient care, patient death, the need for constant learning, the task of teaching, the requirements of attending physicians and senior residents, and the necessities of family and personal life — all while being subjected to the chronic sleep-deprivation and exhaustion caused by their excessive work schedules.  In a study conducted by Schwartz et al., 377 residents answered a survey in which they were asked to rate the most stressful factors in their residency.  Among choices such as “high mortality among patients” and “large patient load,” residents rated “lack of sufficient sleep” and “frequent overnight call” as the top two most negative factors of residency.[6]  Small has written of a “house officer stress syndrome” caused in large part by sleep-deprivation and excessive work load, in which physicians-in-training may suffer from (1) episodic cognitive impairment, (2) chronic low-grade anger with outbursts, (3) pervasive cynicism, (4) family discord, (5) depression, (6) suicidal ideation and suicide, and (7) substance abuse.[7]  

A number of studies have sought to identify residents’ own perceptions of their training.  Although not objective measures, resident opinions offer some insight into the problem.  Daugherty et al. found that medical residents had only a moderate level of satisfaction with their learning and work experience.[8]  When satisfaction with specific aspects of the internship year was reported on a scale from 1 to 7 (least to most), it was found that residents were most satisfied with what they learned (4.8) and their relationships within the hospital (4.8).  They were least satisfied with friend and family relationships outside the hospital (3.9), and with personnel support services (3.6).  Nearly 20% of respondents rated their internship experience as less than satisfactory. 

In a Letter to the Editor of the Journal of the American Medical Association (JAMA), Dr. Murad Alam of the College of Physicians and Surgeons at Columbia University wrote, 

While many types of work require precision (e.g., computer programming) or concentration (e.g., air traffic control), medical internship is unique in requiring both from workers in a chronic state of fatigue.  Living in the hospital for protracted durations prevents residents from obtaining respite from their job-related frustrations.  The intrusiveness of being on-call barely allows for even a few minutes of privacy.  For days at a time all personal activities are continually subordinated to the mantra of patient care.  On-call responsibilities make residents exhausted, lonely, and unable to meet their own needs … [9]

In a survey completed by 108 male and female residents, over 40% of those who defined themselves as being in a committed relationship believed that they had “major problems” with their spouse or partner, 72% of these believed that these problems were due to residency, and 61% reported that their spouse or partner agreed with this assessment.[10]  Multiple regression analysis revealed that 50% of the variance in relationship stress could be accounted for by the following ten variables: high number of hours spent in the hospital, year of training, lack of moderate athletic activity, sleep deprivation, lack of family and social contact, inability to do daily errands, indebtedness, large amount of time spent doing housework, high number of spouse’s working hours, and fewer awake hours at home (model F = 3.05; df = 14; 42; p < 0.002).  A paper describing a support group for the wives of medical residents found prevailing feelings of anger, depression, frustration, and impotence in coping with their husbands’ residency period.[11]   


In our review of harms to medical residents, we limited ourselves to studies that included residents (and not physicians in general) as the study population, and that used work hours or hours of sleep as the predictor variable and treated specific effects on health as the outcome variables.  Application of these selection criteria resulted in elimination of much of the literature on resident harm, including studies on marital stress, suicide, and substance abuse.  However, for three specific health hazards—motor vehicle crashes, mental health, and obstetric complications—some studies did match our selection criteria, and revealed substantial evidence of resident harm.  For each hazard, we present background information, describe relevant studies, and summarize key findings. 

A.  Motor Vehicle Crashes 

A 1998 American Medical Association (AMA) Council Report identified nonstandard schedules such as rotating and night shifts as risk factors in sleep-caused highway crashes: “The public health consequences of such [driving] errors can be substantial and are a concern for health care workers… ”[13]

The news is peppered with stories of medical residents dying from car crashes while driving from the hospital.  In January of 1999, a resident named Valentin Barbulescu had just finished a long shift in the cardiac care unit and was driving off to take an important certification exam.  He was killed in a one-car crash in rural Pennsylvania, an accident his colleagues reported was due in part to fatigue.[14]  An emergency medicine resident wrote a letter in 1998 to the editor of the Annals of Emergency Medicine, describing how he woke up at the wheel just in time to stand on the brakes and skid 73 feet before hitting the back of a 10-ton truck going 5 miles per hour.  He was on his way back home from the second of two 12-hour night shifts.  The resident suffered ten fractures and the failure of his new marriage as a result of the accident.[15]  In 1990, a New York City medical resident fell asleep at the wheel of his car after a sleepless 36-hour shift at New York City Hospital, and awoke to find his car stopped on FDR Drive, with traffic driving around him.[16]  Dr. Moriarity from the surgery program at Johns Hopkins relates, 

I think outside the hospital, you know, most residents fall asleep at the wheel driving home from work.  Almost every resident I know in that program [surgery] has fallen asleep at the wheel driving home from work.  And many of them have been in car accidents.  Inside the hospital, they fall asleep at the operating table…[17]

An informal survey published in JAMA in 1988 found that six out of seven surgical residents had fallen asleep at the wheel while driving to and from work, and three had been involved in car accidents.[18]  One doctor commented, “Falling asleep at the wheel post-call is virtually universal.  I have not found anyone who has not had this problem.”[19]

A MEDLINE search using the Medical Subject Headings (MeSH) “Internship and Residency,” “Accidents, Traffic” and “Sleep Deprivation” yielded three studies which have formally assessed the role of resident work schedules in automobile crashes.  In the most compelling study, researchers at Wayne State University sought to determine the prevalence of motor vehicle crashes (MVCs) and near-MVCs in emergency medicine residents, and the proportion which were caused by sleep deprivation.  Fifty-six percent of the residency programs surveyed responded, yielding a sample of 697 residents.  Seventeen percent (121) of these residents had a total of 157 MVCs.  The odds of having a crash caused by falling asleep at the wheel were higher during residency than before residency (19.3% vs. 4.1%, p < 0.001, odds ratio 6.7).  MVCs were most closely associated with being a first-year resident (50%), being on-call every four days (40%), and getting zero hours of sleep (27%).  For near-MVCs due to falling asleep at the wheel, residents were again more likely to experience these during rather than before their residency (60% vs. 26%).  The authors of the study concluded, “Emergency medicine residents are 6.7 times more likely to have a MVC due to falling asleep at the wheel during their residency [than before residency].”[20]

In a second study, researchers from Johns Hopkins Hospital examined how sleep deprivation affects driving in pediatric residents, and used pediatric faculty as the control group.  Questionnaires were administered to 70 pediatric residents who were on call every fourth night, and to 85 faculty members who were rarely disturbed at night.  The response rate was 87%.  Residents averaged 2.7 (SD- 0.9) hours of sleep while on-call vs. 7.2 (SD 0.8) hours of sleep when not on-call (p < 0.001).  Faculty slept 6.5 +/- 0.8 hours at night.  12.5% of pediatric faculty had fallen asleep while at the wheel at a stop light over the preceding three years, compared to 44% of residents (p < 0.001).  Forty-nine percent of residents reported falling asleep at the wheel (not necessarily at a stop light), and 90% of these events occurred while the residents were post-call.  The authors reported, “One [resident] wrote that she routinely used her emergency brakes when stopped at a light because of her sleepiness post-call.”  The report concluded, “We have demonstrated that [residents] have an increased incidence of falling asleep at the wheel when driving home post-call, and this probably has resulted in increased … motor vehicle accidents.”[21]

A third study involved a survey of 1554 residents in emergency medicine, 62% of whom responded.  Eight percent of residents reported crashes and 58% reported near crashes.  Nearly 75% of the MVCs and 80% of the near-MVCs followed the night shift.  Univariate analysis showed that the frequency of MVCs and near-MVCs was inversely related to the residents’ subjective tolerance of shiftwork and ability to overcome drowsiness.  Logistic regression analysis revealed that the number of crashes and near-crashes was positively related to the total number of night shifts worked per month.  The authors concluded: “The results of this study demonstrate that driving home after a night shift is an occupational risk for EM residents. Both sleep deprivation and fatigue may play a role in these events…”[22]

Yet more evidence comes from an abstract published in Anesthesiology, in which investigators surveyed 74 anesthesiology residents at the Hospital of the University of Pennsylvania.  Residents were asked to report traffic accidents, near-miss accidents, or traffic violations that occurred during their residency and which they attributed to fatigue.  Information on traffic conditions, time of day, call frequency, and hours of sleep on-call prior to the crash was obtained.  Seventy-eight percent (58) of the residents responded.  Crashes during residency were reported by 17% of the residents, with all crashes occurring between 8 and 9 am (driving home post-call) in “moderate” to “no” traffic.  “Near misses” or narrowly avoided crashes were reported by 72% of residents, with one-third of this group having had 5 or more.  Most residents (85%) indicated that they were specifically concerned for their safety because of fatigue while driving home post-call.  When compared to data from the National Highway Transportation Safety Administration (NHTSA) and the National Safety Council (NSC) on all drivers in 1994, investigators found that anesthesiology residents had over twice the national rate of “property-only” accidents (2.5% vs. 5.7%, respectively).[23]

Using the same questionnaire as the University of Pennsylvania study, CIR conducted a survey of over 100 residents at their 2000 Convention, receiving 62 responses, 60 of which were usable.[24]  Unpublished data from this survey indicate that most residents (78%) drive home from the hospital post-call; of these, 40% had a post-call MVC, and 75% had experienced a “close call.”  Most of these took place during the internship year (63% of MVCs and 75% of near-misses).  Sixteen percent had gotten no sleep and 53% had gotten less than 2 hours of sleep the night before the accident.  As with the previous study, both weather and traffic conditions were minimal contributors.  Although response bias may be a factor, these results report an even higher prevalence of post-call MVCs and near-misses associated with sleep-deprivation and long work shifts. 

Wendt has referred to chronically sleep-deprived residents as “misguided missiles.”[25]  Indeed, multiple studies establish that motor vehicle crashes are an occupational hazard for a significantly high number of residents under their current working conditions.  It is common sense that excessive work schedules contribute to resident fatigue, increasing the risk of serious motor vehicle crashes.  Moreover, residents unfit to drive due to fatigue and sleep deprivation represent a threat to other drivers on the road.  In reducing work hours for medical residents, it is clear that OSHA would be acting in the interest of both resident health and public health. 

B.  Mental Health 

Negative Effects of Work Hours on Mood and Affect  
Mood is defined as the prevailing subjective emotional state (such as happy, sad, euphoric, irritable, or agitated) of a subject, and affect is defined as how the subject’s mood is expressed.  Depressed mood in particular is characterized by general hopelessness, passivity, lifelessness, dysphoria, demoralization, and pessimism.  Subjects can be irritable, emotionally unstable, and argumentative; some can be agitated and anxious, while others are quietly apathetic and vegetative.[26]  A MEDLINE search with the MeSH terms “Internship and Residency,” “Work Schedule Tolerance,” and “Sleep Deprivation” yielded the following studies on the negative effects of work hours on mood and affect in residents.   

A case study of 14 residents at Columbia Presbyterian Hospital illustrates the alterations in mood and affect experienced by residents as a result of fatigue and sleep-deprivation.[27]  In this study, residents were on-call every other night, and often worked 60-hour shifts.  The following are comments taken directly from study subjects, all of whom experienced mood alterations: 

Difficulty Concentrating.  “When I’m tired, even though my mind is active, I can’t concentrate.  I can’t put things together in my mind so I don’t even try.  If a patient is really sick, I can pull myself together but I can’t write down what I’ve done in the chart.  What I write is a reflection of a fragmented thought process…. It gets me scared when this happens because it means that I am losing control of my ability to think.”  

Depressed Mood.  “My home life suffers and I miss my wife greatly.  I feel ashamed that I get tired and can’t live up to the tradition of the ‘iron men.’”

Irritability.  “As I lose sleep, I get more explosive and more irritable.  I snap at nurses and make them cry.  I pick on the nursing staff rather than my wife or my patients.  I’m ashamed of it in a way.”  “If you’re on two nights in a row, you want to do as little as possible.  You give bad care.  I am irritated all the time then … I give bad care to my patients, unfortunately.  When I’m tired I don’t give a Goddamn.” 

Inappropriate Affect (“Black” Humor).  “I laugh at things that aren’t really funny.  I’m giggly when I’m tired.  For instance last night a patient came in comatose.  Another intern asked the patient: ‘Do you have any parakeets at home?’  I found this enormously funny and I laughed and laughed … Things don’t seem so funny to me when I am rested.  Another example of my sense of humor when I am fatigued would be: An intern gets a patient with congestive heart failure and pulmonary edema.  He makes a wise comment like, ‘Give her some vitamins and send her home.’  That makes me crack up.  I would not find such a remark funny when I am rested.”

Memory Deficit.  “I would forget what I just said so my next sentence would make no sense.  I also stop sentences midway a lot because I forgot what I wanted to say.” 

Other studies confirm these mood effects using objective measures.  Researchers at Case Western Reserve University compared 34 pediatric residents before and after a night of call with 27 residents who were not on-call either day.[28]  The different groups were tested using the Profile of Mood State scale (POMS), a 65-item adjective-rating measure that assesses mood state on a five-point scale.  A total score is provided, as well as subscales for tension-anxiety, depression-dejection, anger-hostility, vigor-activity, fatigue-inertia, and confusion-bewilderment (higher score = worse mood state).  The on-call group of residents showed increased total negative mood state scores (pre-call: 54, SD 22; post-call: 74, SD 29), while the group not on-call any of the days demonstrated improving total mood scores (Day 1: 60, SD 33; Day 2: 49, SD 27).  The scores for all six subscales of the POMS also demonstrated increased negative mood for the on-call group post-call, while the off-call group demonstrated improved mood on the second day.  The consistent results of the subscales suggested that call duty affects a broader array of psychological responses than simply fatigue.  The authors observed: “Our findings are consistent with anecdotal reports of the effects of call… The fact that number of hours slept correlated with increased negative mood state, anxiety and perceived stress suggests that sleep deprivation affected psychological state.”

In another study utilizing the POMS, researchers compared 16 randomly recruited residents (8 males and 8 females) before and after 32-hour shifts at St. James Hospital in Dublin.[29]  The investigators found that the total POMS mood score significantly deteriorated post-call, from an average of 3 pre-call to 37 post-call (p < 0.0021; again, higher scores being worse).  The difference in depression-dejection score was not found to be significant comparing before call to after call; however, the investigators noted: “The lack of significance in the depression rating must … be treated with caution as the total mood disturbance score, which has been shown previously to correlate well with general psychological well-being, showed a marked deterioration after the period of duty in our subjects.”  Moreover, the study had a very small sample size.

German researchers performed psychological testing on 40 residents at the University Hospital in Tuebingen, who averaged 3 hours of sleep on call.  Subjects were tested in the morning after a night off-duty (with at least 6 hours of uninterrupted sleep), and once more at a similar time in the morning after a night on-call (24 continuous hours worked in the hospital).  Scores for emotional state tested via a 28-pair mood-adjective list were considerably worse for the post-call group compared to the off-call group (60, SD 9 vs. 54, SD 7; 2 points were assigned per negative adjective, 0 for positive adjective, and 1 for neither/nor).  The researchers noted, “[T]he emotional condition worsened after one night on-call with the following negative adjectives having been reported most often: tired (32 residents), feeble (24), tense (20), hesitating (15), lacking in verve (15), restless (12).”[30]

Another study randomly assigned 30 first-year internal medicine residents to sleep-deprived (n = 16) and nonsleep-deprived (n = 14) groups, and followed them from the fifth to the ninth month of their internship year at the Medical College of Virginia Hospitals.  The two groups had similar ages, sex ratio, and racial composition.  Residents in both groups were tested with the Multiple Affect Adjective Check List (MAACL) from 2 –3 pm after being on-call or after having been off-duty.  Sleep-deprived residents (mean 2.7 hours sleep) reported greater mood disturbance than the nonsleep-deprived residents (mean 7.9 hours of sleep) in the different categories of the MAACL (p < 0.05).[31]

In another study at Columbia University, investigators used the Nowlis and Green Mood Adjective Check List (MACL) to compare 14 interns when rested and fatigued.  The MACL consisted of 33 adjectives describing 11 mood factors: aggression, anxiety, surgency (feeling carefree, lively, talkative), elation, concentration, fatigue, social affection, sadness, skepticism, egotism, and vigor.  In the 32 hours before testing, rested interns slept a mean of 7.0 hours and fatigued interns slept a mean of 1.8 hours (p < 0.001).  Once more, tired residents reported worse scores than rested residents, with tired residents having statistically lower scores in positive mood factors (surgency, vigor, elation, egotism, and social affection) and significantly higher scores in negative mood factors (fatigue and sadness).[32]

Studies thus consistently show that medical residents experience negative, unhealthy alterations in mood as a result of their long work shifts.  The authors of a comprehensive review of the effects of sleep deprivation on residents concluded that the “accumulated evidence of studies performed over the past 30 years … suggests that the traditional system of 100-hour work weeks and 36-hour days may do harm.  Clearly, residents’ moods, affects and attitudes are altered unfavorably.”[33]

Distinct from depressed mood is depression, a clinical term requiring that at least five of nine defined criteria are met for a period of at least two weeks.  One of the five signs must include either depressed mood or anhedonia (loss of interest/pleasure in life), and the other four can include appetite disturbance with weight change, sleep disturbance, psychomotor disturbance, fatigue or loss of energy, feelings of worthlessness or guilt, diminished ability to concentrate, and recurrent thoughts of death/suicidal ideation.[34]  High rates of depressed mood place medical residents in a higher risk group for developing clinical depression.  Indeed, as many as 30% of medical residents experiences depression at one time during their residencies.[35] Female physicians have been shown to be especially vulnerable.[36]  Applying the MeSH terms “Internship and Residency,” “Work Schedule Tolerance,” and “Sleep Deprivation” to the MEDLINE database yielded the following two studies. 

In a study utilizing the Center for Epidemiological Studies-Depression (CES-D) scale, a predictor of depression, investigators surveyed 68 medical house officers at Rhode Island Hospital. [37]  They administered the test on a monthly basis for a year, with a response rate of 83%.  Twenty-one percent of respondents reported “depressed” scores, defined as a CES-D score equal to or greater than 16 (on a scale of 0 to 60).  When classified by year, 29% of first-year residents, 22% of second-year residents, and 10% of third-year residents reported depressed scores (p < 0.0001).  (Resident work schedules typically improve as residency progresses.)  When responses were examined by rotation (the specialty in which resident is currently working), depressed responses were most frequently received during those rotations that routinely required over 80 hours of work per week.  Twenty-five percent of residents reported depressed responses while on ward rotations, and 32% while on the intensive care unit, both of which require over 100 hours of work per week.  The author concluded, “The increased frequency of depressive symptoms on ward and intensive care rotations may be, in large part, caused by long working hours and sleep deprivation.”

In a study at two hospitals in St. Louis, investigators interviewed 53 interns at the end of their first year of training. [38]  Based on the Feighner criteria for clinical depression, 16 (30%) had an episode of depression during their internship, of which 13 were definite depressions and 3 were probable depressions.  The depressed and non-depressed groups were very similar in terms of age, sex, marital status, and type of internship.  Medical, social, and childhood histories were likewise not significantly different.  Eleven of the 16 interns became depressed within the first few months of their training.  Four of the 16 had suicidal ideation, 3 had a suicidal plan, and 6 experienced marital problems for the first time.  One subject who had made a suicidal plan thought of 5 or 6 ways to kill himself so that his wife could collect insurance.  Six depressed interns had feelings of hopelessness, 2 had a fear of losing their minds, and 3 called their spouses while on night call, crying, and saying they couldn’t go on.  Of the 11 whose depression began in the first two months, 7 were working more than 100 hours per week.  Of the 5 who became depressed later in their internship, 3 were working more than 100 hours per week at the time of the onset of their depression.

In sum, while it is certainly possible that the difficult nature of the medical work in residency contributes to the development of depression, evidence for the large role of excessive work hours is strong.  Fatigue and sleep deprivation caused by excessive work hours contribute to depressed moods in residents, placing them in a high-risk group for developing clinical depression, in turn increasing their risk for suicidal ideation/suicide.  Experts have agreed, “This combination of stress and fatigue may lead to severe psychologic repercussions, which may first appear as disappointment, loss of idealism, and isolation, and then progress to feelings of helplessness, impaired performance, and outright depression.”[39]  The authors of another study concluded, “In view of the special vulnerability of medical trainees to occupational stress, all efforts are warranted to reduce sleep deprivation in the medical profession.”[40]  It is reasonable to expect that reducing work schedules to allow for more sleep should reduce both the incidence of depressed mood and the likelihood of developing depression. 

C.  Obstetric Complications 

There is a growing number of women entering the medical profession, and almost 50% of married female residents will become pregnant during their residency.[41]  Moderate increases in work schedule over 40 hours are not consistently associated with adverse pregnancy outcomes.[42]  However, excessive levels of work level are associated with obstetric complications.  In addition to the selection criteria applied to all other studies in this petition, the studies reviewed in this section required a control population.  The following are a result of a MEDLINE search using the MeSH terms “Internship and Residency,” “Pregnancy Complications,” and “Pregnancy Outcome.” 

A study by Osborn et al. compared the outcomes of first pregnancies in 92 female residents and 144 spouses of male residents from 45 university-affiliated residency programs.[43]  Female residents reported working a mean of 62.5 hours per week and wives of male residents reported working a mean of 25.6 hours per week.  The authors found that female residents were as likely to give birth to a live, full-term newborn as the spouses of male residents.  For white cohort members, an increased risk of premature labor without delivery was identified, with a relative risk of 12.3 (95% confidence interval 2.4 – 61.6).  This study was limited by its low sample size.  The authors commented, “Although a 12-fold increased risk of premature labor was identified for white female housestaff, given the study’s limitations, this finding should be generalized with caution.”  Furthermore, “a much larger sample size would be required to detect subtle differences in outcome due to fatigue factors and occupational exposures.”

In the best-designed investigation, Klebanoff et al. sent questionnaires to 5096 female physicians who had graduated from medical school in 1985 and a random sample of 5000 of the 12,306 male physicians who graduated the same year.[44]  Eighty-seven percent (4412) of the women residents and 85% (4236) of the wives of male residents responded to the questionnaire, which included questions on outcome of each pregnancy and number of hours worked.  Women residents reported working twice as many hours per week during their pregnancies as did the wives of the male residents.  Between the two groups overall, investigators found no statistically significant differences in the proportion of pregnancies that ended in miscarriage, ectopic gestation, stillbirths, preterm delivery, or intrauterine growth retardation.  However, three important findings were identified.  First, premature labor requiring bed rest or hospitalization was nearly twice as common among the women residents as among the male residents’ wives (11.3% vs. 6.0%, p < 0.001).  This finding supports the similar finding of the Osborn study.  Second, preeclampsia or eclampsia was also twice as common among the women residents as the male residents’ wives (8.8% vs. 3.5%, p < 0.001).  Third, for those residents working 100 or more hours per week during the 3rd trimester, there was more than twice the risk of preterm delivery compared to those working fewer than 100 hours (10.3% vs. 4.8%, p = 0.04).  Premature labor, preeclampsia/eclampsia, and preterm delivery (in women residents working over 100 hours), were thus found to be significant problems for the pregnant residents.  Klebanoff et al. commented, “This increase suggests that the New York State law limiting residents to 80 hours of work per week is well advised with respect to pregnant residents.”  

Using the same data set, Klebanoff et al. also compared early-pregnancy complication rates in female resident physicians with those among partners of male resident physicians.[45]  The life-table probability of spontaneous abortion was 14.8% for female residents and 12.6% for the partners of male residents (RR 1.18, 95% confidence internal 0.96 – 1.45), a difference that was not statistically significant.

In another survey, investigators surveyed 1025 female board-certified obstetricians about their pregnancies before, during, and after residency.[46]  The response rate was 49%.  The mean number of hours worked during residency was 78.9, compared to 36.4 before residency and 46.5 after residency.  The average birthweight of firstborn infants delivered during residency was found to be significantly lower than the birthweight of firstborn infants delivered before residency (3146g, SD 696 vs. 3525g, SD 455, respectively; p < 0.001).  Although mean birthweights were lower in infants delivered after residency (3263g, SD 556; p < 0.005), they were still lower than birthweights of infants born before residency.   The low birthweight rates (defined as any birthweight below 2500g) were 3.7%, 11.6%, and 2.6% before, during, and after residency, respectively.  A second critical finding was that infants delivered during residency were more likely to be born with intrauterine growth retardation (defined as birthweight that was under the 10th percentile for a given gestational age) than those delivered before or after residency (rates of 1.2%, 8.2%, and 1.0%, before, during, and after residency.)

In summary, the results of the best-conducted study with the largest sample size point to increased risks for preterm labor requiring serious hospitalization, preeclampsia or eclampsia, and preterm delivery in those residents working greater than 100 hours a week.  A second paper with a large sample size suggests that residents and their children can also suffer from decreased birthweights and intrauterine growth retardation.  The author of one study agreed that available research includes “sufficient findings to suggest that heavy exertion and fatigue may cause premature deliveries, decreased birthweights, and other complications in pregnant residents.”[47]  In a review of the literature on pregnancy complications of medical residents, another author concluded, “[T]he greatest factor leading to poor pregnancy outcomes among physicians seems to be time spent working, particularly during late pregnancy.”[48]  The signers of this petition believe that even modest work-hour limitations would significantly reduce these adverse outcomes. 

D.  Limitations of Studies 

In general, we have restricted ourselves to studies that use work hours or hours of sleep as the predictor variables, and various aspects of health as the outcome variables.  There is, however, variation in the way the predictor variables were measured.  Some studies used hours worked (averaged over weeks or months), while others used number of hours of sleep or time on-duty as the predictor variable.  We did, however, exclude studies that simply used night shift work as the predictor variable.  The outcome variables were also assessed in different fashions; for example, depression was measured using both the CES-D scale and direct interviews.  Despite these limitations, the signers of this petition believe that the overall conclusions are valid and consistent with the common-sense notion that excessive work hours represent a threat to medical residents.


A.  Current Work Standards and Compliance 

The U.S. has responded to the problem of resident work hours in ways that have so far proved inadequate.  In 1989, following an 18-year-old woman’s death from medication error by an exhausted resident at New York Hospital, and the intensive investigation of hospital practices that ensued, the New York legislature passed section 405 of the New York State Health Code.  These are commonly known as the Bell Regulations, after the chair of the investigational commission, Bertrand Bell, M.D. (Dr. Bell is a signatory to this petition.)  The New York State Health Department was charged with enforcing the following standards:

  1. an 80-hour work week averaged over 4 weeks,
  2. shifts not exceeding 24 consecutive hours,
  3. shifts separated by at least 8 hours,
  4. maximum shifts of 12 hours per day for emergency medicine residents,
  5. and at least one scheduled 24-hour period of non-working time per week. 
  6. Surgical residencies were exempt from the 24-hour maximum shift, if surgeons met all of the following conditions:
    a. were scheduled to be on-call only every 3rd night,
    b. received a rest period of 16 hours following their on-call shift,
    c. could be documented to get adequate rest with infrequent patient interruption during the night hours of their on-call shift,
    d. and could be immediately relieved from duty if fatigued.

The other major component of the code was requiring better supervision of housestaff by attending faculty.[49]

Surprise inspections of 12 New York hospitals in 1998, however, found them in flagrant violation of the Bell Regulations: 94% of residents in New York City hospitals and 37% in all New York state hospitals were working more than 85 hours per week.  In surgical wards, 77% of residents in New York City and 32% in upstate hospitals were working more than 95 hours a week.  In the non-surgical wards, 38% of residents statewide were in programs violating the standard 24-hour shift. [50]  Frances Tarlton, spokeswoman for the New York State Health Department, commented, “we found violations at every single hospital we surveyed.  The hospitals themselves often didn’t know how much they were overworking the residents.”[51]  Since the initial inspection of 12 hospitals, a total of 26 of 36 (72%) inspected teaching hospitals in the state have been cited for violation of the section 405 regulations. [52]

In their original form, the section 405 regulations called for a maximum fine on hospitals of $2000 per violation, a sum inadequate to affect hospital policy.  Since the 1998 inspection, five hospitals have been fined a total of $82,000.[53]  Dr. Bell commented: “There is no real deterrent here.  When these people get fined, they get fined nothing.  The state should set up a roster in which all of the hospitals are audited all of the time on compliance with these laws and the results are published.”[54]  Since then, the Bell Regulations have been strengthened by the Health Care Reform Act of 2000, passed by the New York State Legislature.[55]  The state now requires that each year, all New York teaching hospitals submit plans describing how they will comply with the Bell Regulations.  A sum of $168 million has been apportioned to help crack down on hospitals which force residents to violate the hours limitations, and the Health Department can now issue fines of up to $50,000 for repeat violations.

So far, New York is the only state to have even attempted limiting resident work hours by enacting regulations.  Yet according to Mark Green, New York City’s Public Advocate, “the problem of exhausted and poorly supervised residents continues unabated.”[56]   

What comes closest to a national standard is a set of voluntary guidelines drawn up by the Accreditation Council for Graduate Medical Education (ACGME), the organization responsible for overseeing residency programs in the U.S.  ACGME does not apply a uniform set of guidelines to all specialties, but has instead formed a total of 26 Residency Review Committees (RRCs) which determine work-hour requirements for each of the medical specialties.[57]  For example, the RRC for internal medicine programs has determined that internal medicine residents should adhere to an 80-hour work week with call a maximum of every third night.  Surgery programs, on the other hand, do not have an overall work-hour limit (residency program directors are suggested to develop schedules with “appropriate” work hours), but should also schedule call only every third night.  (See Table 1 for more examples.) 

The signers of this petition believe that exemptions should not be made for particular specialties, and that — unlike the ACGME program guidelines — a uniform standard should be applied across all specialties.  OSHA’s mandate is to protect workers and there is no evidence that residents in surgery, for example, are any less prone to the adverse effects of long work hours reviewed in this petition.  We have therefore petitioned for uniform work-hour restrictions for all residents. 

In part because compliance with these guidelines is voluntary, a significant proportion of hospitals across the U.S. have failed to meet them.  In a 1999 inspection of 86 hospitals, 17 of them — nearly 1 in 5 — were found to be in violation of the ACGME work-hour standards.[58]  Of the 69 training programs in general surgery reviewed, 25 (36%) were cited for noncompliance.  Violations were found at 7 of the 13 programs in pediatric surgery (54%), 28 of the 92 programs in internal medicine (30%), and 20 of the 69 programs in orthopedic surgery (29%).  Dr. Marvin Dunn, ACGME’s director of RRC activities, commented, “The number of work-hour citations has increased every year for surgery and several other specialties.”[59]

ACGME citations call for programs in violation of the guidelines to come into compliance within a certain period of time, after which the program will be re-inspected.  Upon review, the program may receive a status of good standing, accreditation with warning, probation, or withdrawal of accreditation.  To date, no residency program has ever lost its accreditation based on violation of work-hour guidelines alone.  Programs have thus continued operating in violation of the guidelines.  Moreover, residents are not likely to seek enforcement of work-hour guidelines from the ACGME, as the only authority the ACGME has is withdrawal of accreditation of a residency program — hardly an outcome desired by residents who depend on their program for training. 

Investigations by the New York State Health Department and the ACGME, therefore, reveal that residents are still being subjected to excessive work hours despite restrictions that are in place.  The bodies meant to implement these restrictions lack the power to effect systemic reform.  Intervention by OSHA to regulate resident work hours is crucial to protecting residents, as organized medicine has been unable to police itself. 

B.  Other Countries’ Responses to the Problem of Excessive Resident Work Hours  

While the U.S. has been largely ineffective in addressing the long work-hour problem, other industrialized countries have taken active and successful steps to regulate resident work hours.  Indeed, most of the industrialized world already regulates resident work hours.  Over the past 15 years, legislation, directives, decrees, and collective agreements that limit average weekly hours worked by medical residents have been instituted in at least 6 countries or jurisdictions: Australia (70 hrs), Denmark (fewer than 45 hrs), United Kingdom (56 hrs), the European Union (48 hrs by the year 2003), Germany (56 hrs), and the Netherlands (48 hrs).[60], [61]  (See Table 2.) 

The restructuring of resident work schedules in these countries was achieved in most cases through collaboration between medical societies and the government.  Both entities were quick to acknowledge that long hours are counterproductive to resident learning, and that harm was being inflicted on both residents and patients.  In November of 1999, the social affairs ministers of the European Union approved a plan to shorten the phase-in period of a 48-hour work week to only 4 years.  Andrew Hobart, chairman of the British Medical Association’s Junior Doctors Committee, stated that he is “…pleased that the European Parliament has taken action to protect doctors and their patients against excessive hours of work.”[62]

Nations that have limited resident work hours have done so through designing new approaches to residency programs that emphasize both efficiency and learning, implemented through changes in shift schedules (such as the night float call system, in which rested residents take over the night shift), redefining of resident duties to include fewer menial tasks, hiring of additional ancillary staff, and transferring of more workload to faculty physicians.  Because these countries have made it a priority, they have demonstrated that it is indeed possible for hospitals to schedule reasonable work weeks for their residents.

C.  U.S. Work-Hour Regulations in Other Industries  

The federal government has long recognized the importance of regulating work hours in the transportation industry.  Under the jurisdiction of the Department of Transportation (DOT) and its daughter organizations, work-hour limits and rest-period requirements for the highway, aviation, railroad, and maritime industries have been established.  The Federal Aviation Administration (FAA) does not allow airline pilots to fly more than 30 hours per week or 8 hours in a single day.  The Federal Motor Carrier Safety Administration (FMCSA) limits drivers in commercial industries (trucking, bus drivers, etc.) to no more than 10 hours on the road or 15 hours on-duty, with a minimum of 8 hours rest per day.  The Federal Railroad Administration (FRA) and the U.S. Coast Guard (USCG) have likewise put work-hour regulations in place for operators in the railroad and maritime industries.  (See Table 3.)  These work-hour requirements have been instrumental in maximizing worker and public safety for many years.  The regulations currently in place, however, do suffer from certain loopholes allowing workers to be on-duty for long hours.  For example, some hours-of-service requirements do not operate on a 24-hour cycle: the FRA mandates a 10-hour rest period following 12 hours of work, allowing 14 hours of duty within a 24-hour period.  The second 12-hour shift can then extend 10 hours into the next 24-hour period, creating schedules that are staggered and irregular. 

In recent years it has become increasingly clear that fatigue plays a major role in transportation safety.  A large volume of research designed to further delineate the relationship between fatigue/sleep deprivation and performance/safety has been completed, and more studies are currently underway.  Following accident studies conducted in the 1980s, the National Transportation Safety Board (NTSB), the federal agency responsible for investigating significant accidents in transportation, issued a set of recommendations to the DOT in 1989: to investigate fatigue and its relation to safety, to educate transportation industry workers on work and its relation to health, and to revise current hours-of-service regulations to maximize the safety of its workers and the people they serve.   The result has been a collaborative effort among organizations within the DOT to “modify the appropriate Codes of Federal Regulations to establish scientifically based hours-of-service regulations that set limits on hours of service, provide predictable work and rest schedules, and consider circadian rhythms and human sleep and rest requirements.”[63]

As a result of major accident investigations, special investigations, and safety studies that identified operator fatigue as a factor, the NTSB has issued more than 70 fatigue-related safety recommendations to the DOT since their 1989 recommendation.  For the fiscal years 1990 through 1998, the DOT spent more than  $30 million on fatigue research.[64]  Organizations within the DOT are currently in the process of proposing new hours-of-service regulations, developing fatigue countermeasures, and forging partnerships with industry and labor to collaboratively study work-hour issues.  OSHA has, to our knowledge, conducted no research on resident fatigue, even though work hours of residents greatly exceed the regulated work hours in these other industries.

The Federal Motor Carrier Safety Administration  
The National Highway Traffic Safety Administration (NHTSA) estimates that each year, drowsy drivers may be responsible for as many as 103,000 crashes, [65] which result in more than 1,500 fatalities and 71,000 injuries.[66]  In the interest of highway safety, the Motor Carrier Act of 1935 resulted in hours-of-service regulations for commercial drivers, based on the rationale that “It is obvious that a man cannot work efficiently or be a safe driver if he does not have an opportunity for approximately 8 hours of sleep in 24.”[67]  The regulations have remained largely unchanged. 

Research conducted since that time confirms this claim, and has identified the link between sleepiness and crashes, accidents, and errors previously attributed to fatigue and inattention.  According to the FMCSA, there is evidence that “many crashes occur as a result of commercial motor vehicle (CMV) driver error, that driver error is often the result of inattention, that inattention can often be the result of fatigue, that the fatigue which causes inattention is often related to sleep deprivation, and that sleep deprivation is often related to working conditions of drivers.”[68]  In 1985, the American Automobile Association (AAA) Foundation for Traffic Safety in “A Report on the Determination and Evaluation of the Role of Fatigue in Heavy Truck Accidents” examined 250 accident reports of heavy truck accidents in six Western states.[69]  The study concluded that fatigue was the probable or primary cause of 41% of those heavy truck accidents.  The NTSB’s 1990 study of 182 heavy truck accidents that were fatal to the driver showed that 31% of the accidents in this sample involved fatigue.  Fatigue, drugs that are taken to counteract the symptoms of fatigue, drugs that aggravate fatigue, and the interaction of fatigue and drugs were found to be major factors in accident causation.[70]

Another set of studies has examined fatigue in terms of the role of continuous service time in causing accidents.  In the 1970s, the Federal Highway Administration (FHWA) conducted a study on truck and bus drivers which found that by the maximum allowed 10 hours of driving time, driver performance deteriorated, driver alertness diminished, rest breaks became less effective, and accident probability increased.[71]  Many studies have corroborated the finding that increased service time leads to increased accident risk.  A review of a series of truck driver fatigue research studies from the late 1970s to the early1990s found that time on task appears to have a limited effect on accidents for regular daily work periods less than 11 hours, but may have a more profound impact if the work periods are over 12 hours.[72]  In a survey of 1000 heavy-goods vehicle drivers undertaken in 1982-1983, the accident risk rate after 11 hours of work was nearly double (1.82 times baseline) that for work periods shorter than 11 hours.[73]  Saccomanno et al. found higher overall accident risk associated with 9.5 continuous hours of service or longer compared to baseline.[74]  A case-control study including truck crashes in Washington state from June 1984 through July 1986 matched to a comparison sample with similar roadway, time of day, and day of week characteristics, found higher relative risks associated with over 8 hours of driving.[75]

According to a chart published by the FMCSA in the May 2, 2000, Federal Register, a driver runs 16 times the risk of having a fatal accident during the 13th hour of driving than during the 1st.  (See Figure 1.)  A study by Lin et al. also described a rising risk curve: the first 4 hours of driving had the lowest accident risk, followed by increases in risk of 50% or more up until the end of the 7th hour, 80% until the 8th hour, and 130% until the 9th hour.[76]  In a mini-review by Folkard, it was determined that the safest work shift duration for commercial drivers is 6-9 hours.[77]  In terms of weekly service time, a 1996 study by Bowen found that based on a total of 173,110 reported hours of driving time, “…it seems that after 80 hours on duty [in an 8-day period], the accident rate rises precipitously.”[78]  

Additional research has identified factors other than just continuous driving time as causes of fatigue-related accidents.  A 1995 study of 107 accidents (62 of which were fatigue-related) found that the three most important factors that affected fatigue-related accidents were duration of sleep in the last sleep period, the total hours of sleep during the 24 hours prior to the accident, and the presence of split sleep periods. The truck drivers in fatigue-related accidents in this sample had an average of 5.5 hours of sleep in the sleep period prior to the accident, as compared to 8 hours for drivers in the nonfatigue-related accidents.  Many of the truck drivers involved in fatigue-related accidents did not recognize that they were in need of sleep and believed that they were rested when they were not—about 80 percent rated the quality of their last sleep before the accident as good or excellent. The study concluded that driving at night with a sleep deficit is far more critical in predicting fatigue-related accidents than simply nighttime driving.  Moreover, sleep accumulated in short time blocks (split sleep) was found to impede recovery of performance.[79]  The sleep characteristics of subjects in this study—decreased sleep in the last sleep period, decreased total hours of sleep in the last 24 hours, and split sleep periods—are all shared by medical residents under their present working conditions. 

Of the different modes of transportation, most data on the relationship between fatigue and safety are available for highway transportation.  In the highway transportation industry, research has confirmed the common-sense notion of restricting hours of service, as excessive work schedules have been shown to cause injuries and cost lives.  The evidence thus far has been so convincing that driver fatigue was voted the number one safety concern of the FHWA 1995 Truck and Bus Safety Summit, a meeting of over 200 drivers, motor carrier representatives, government officials, and safety advocates.[80]

The Federal Aviation Administration  
Aviation work-hour limits were addressed in the Civil Aeronautics Act of 1938 and the Federal Aviation Act of 1958.  The FAA reports that 21% of the accidents citing errors in the Aviation Safety Reporting System (ASRS) were related to general issues of fatigue.[81]  As a recent example, pilot fatigue is believed to have contributed to the crash of American Airlines Flight 1420, which skidded off the runway of Little Rock (Arkansas) International Airport on June 2, 1999, killing the pilot and 10 others.  The accident occurred after the crew had worked more than 13 hours.[82]

In a study entitled, “A Review of Flightcrew-Involved, Major Accidents of the U.S. Air Carriers, 1978 Through 1990,” the NTSB sought to learn more about flightcrew performance by evaluating characteristics of the operating environment, crewmembers, and errors made in major accidents. [83]  It found that crews comprising captains and first officers who had been awake longer than the median number of hours of others in their crew position made more errors overall, with significantly more procedural and tactical decision errors. 

In 1995, the NTSB examined operator fatigue in its safety studies on flight crew errors, commuter airlines, and aviation safety in Alaska.  Under Part 135.261 of the Title 14 Code of Federal Regulations, Alaska is permitted longer flight service hours than the rest of the states, due to its remoteness from the 48 contiguous states.  The Board concluded that “the consecutive, long duty days currently permitted [in Alaska] … for commuter airline and air taxi flight crews can contribute to fatigue and are a detriment to safety.”[84] 

The Federal Railroad Administration 
The Railroad Hours of Service Act was first enacted in 1907, substantially revised in 1969, and amended in 1976 and 1988.  In an oral statement at a September 16, 1998 Senate Safety Hearing, Administrator Joline Molitoris of the Federal Railroad Administration (FRA) commented,  

Fatigue and the railroad industry have been synonymous for over a hundred years.  In some industries, this might be only a quality of life issue.  In railroading, it is a life and death safety issue.  About one-third of train accidents and many employee injuries and deaths are caused by human factors.  We know fatigue underlies many of them.  Hundreds of communications from employees and their families eloquently testify to the devastating effects of fatigue…. all employees, contract and management, must be able to work within policies that assure them, their companies and the communities they serve, that they are alert and able to operate safely.[85]

As with the airline industry, freight and passenger rail operations are conducted 24-hours-a-day, subjecting employees to extensive night work, irregular work schedules, and extended work periods with few or no days off.  Fatigue has been thought to be a significant contributing factor to major train accidents in past years.  For example, on November 7, 1990, two freight trains collided in Corona, California.  The NTSB investigation concluded that “The engineer of train 818 failed to stop his train on the Corona siding at the stop signal because he was asleep or in a microsleep brought about by chronic and acute fatigue … a result of the irregularity and unpredictability of his work schedule.”[86]  Another widely publicized example of a fatigue-caused crash was the June 5, 1995 collision of two New York City Transit subway trains on the Williamsburg Bridge in Brooklyn. One person was killed and 69 treated for injuries, with total damages exceeding $2.3 million.  The train operator failed to take action to stop his train because he was asleep.[87]

The United States Coast Guard  
Work-hour regulations for the maritime industry date back to 1908.  In 1997, work-hour regulations from the Standards for Training, Certification, and Watch-keeping of the International Maritime Organization (IMO) also became effective.

A 1996 U.S. Coast Guard (USCG) analysis of 279 incidents showed that fatigue contributed to 16% of critical vessel casualties and 33% of personal injuries. [88]  Three factors were identified that could be combined to calculate a Fatigue Index score for casualty cases, which could then be used as a predictor for accidents: (1) the number of fatigue symptoms reported by the mariner, (2) the number of hours worked in the 24 hours prior to the casualty, and (3) the number of hours slept in the 24 hours prior to the casualty.  These findings parallel those found in commercial driving, demonstrating that the effects of limited sleep apply across different industries.

In a survey of 141 mariners from eight commercial ships (six tankers and two freighters), data on work and sleep patterns as well as other information pertinent to fatigue were collected.[89]  The incidence of critical fatigue indicators such as severely restricted sleep durations per 24-hour period, very rapid sleep onset times, and critically low alertness levels suggested that fatigue is a regular occurrence in commercial vessels. The study concluded that sleep disruption, reduced time between watches, fragmented sleep, and long workdays are the principal contributors to the problem.

One of the most high-profile examples of fatigue-related accidents which has compelled the USCG to examine its hours-of-service regulations is the Exxon Valdez oil spill.  Ranking among the worst environmental disasters to date, the spill resulted in untold death to wildlife, affected 1300 miles of shoreline, and cost $2.1 billion in clean-up efforts.[90]  The three-man crew was held responsible for running the ship aground on Bligh Reef on March 24, 1989.  The official NTSB Marine Accident Report states that, 

The performance of the third mate was deficient, probably because of fatigue, when he assumed supervision of the navigation watch from the master about 2350 [almost midnight].  The third mate’s failure to turn the vessel at the proper time and with sufficient rudder probably was the result of his excessive workload and fatigued condition, which caused him to lose awareness of the location of Bligh Reef.  There were no rested deck officers on the Exxon Valdez available to stand the navigation watch when the vessel departed from the Alyeska Terminal.[91]  

Relevance of Governmental Regulations in Other Industries to the Health-Care Industry  
The justifications for hours-of-service regulations in the transportation industries share many parallels with those in medicine.  At the DOT-sponsored Operator Fatigue Management Conference in August of 2000, representatives from all transportation modes agreed that “the incidence of fatigue is underestimated in virtually every transportation mode” and that “it is likely that fatigue is a bigger contributor to incidents, accidents, and fatalities than many realize.”[92]  The petitioners believe that this statement applies to medical residents and their patients as well, for several reasons.

First, as discussed above, workers in the four transportation industries commonly experience long shifts, long weekly hours, irregular shifts, work cycles that do not operate on a 24-hour schedule, and accumulated loss of sleep (sleep debt).  Medical residents experience similar, and in nearly all cases worse, work schedules.  Second, research conducted in the different transportation modes has converged on common concepts concerning the roles of duration of continuous service, duration of sleep in the last sleep period, duration of wakefulness since prior sleep, duration of sleep in the last 24 hours, split sleep schedules, and sleep debt, and their relation to human performance and safety.  Medical residents are not exempt from the outcomes predicted by the principles of sleep-wake biology.   Third, the efforts of the DOT and the NTSB exemplify the role of the federal government in creating regulations that protect workers and save lives.   The commitment to sleep research, implementation of hours-of-service regulations, efforts to revise hours-of-service regulations according to scientific principles, and collaboration with industry and labor, all serve as models for OSHA to follow. 


More than just harming medical residents, long work hours have been shown to have negative effects on patients by increasing chances for accidents and errors in the delivery of medical care.  A significant body of literature supports the common-sense notion that it is unsafe to subject patients to sleep-deprived and exhausted physicians-in-training.

Studies have shown that well-rested residents outperform their sleep-deprived peers on a wide range of tasks, including basic rote memory, language, and numeric skills;[93] retention of information and problem-solving skills;[94] tests of visual attention, short-term memory, and coding ability;[95] and tests of concentration.[96] The procedural skills of medical residents have also been shown to suffer under conditions of sleep deprivation or restriction, with decreases in residents’ performance of electrocardiogram interpretation,[97] anesthesia monitoring,[98] intubation of mannequins,[99] and simulated laparascopic cholecystectomy.[100]  A recent paper published in the journal Nature found that staying awake for 24 hours impairs cognitive psychomotor performance to the same degree as having a 0.1% blood alcohol level, a value above many U.S. legal driving limits (0.04% to 0.1%).[101]

In terms of actual effects on patients, one study found that hospital inpatients admitted during the night had increased mortality compared to those admitted during the day (relative risk 1.21, p = 0.002), an outcome possibly attributable to resident fatigue.[102]  In another study, work schedule changes at a Veterans Affairs hospital allowing residents to get more rest resulted in shorter in-hospital stays, fewer tests ordered, and 29% fewer medication errors.[103]  A third study retrospectively compared patient care under a traditional resident work-hour schedule and a reduced work-hour schedule. [104]  Although the authors found that more patients experienced at least one medical complication and/or diagnostic test delay when treated by residents working under the reduced work hour schedule, the study suffered from certain limitations, including being unblinded, involving a small number of residents, and taking place only a year after the reduced work schedule was put in place.

The medical profession is often reticent about acknowledging its mistakes.  Wu and colleagues sent out an anonymous questionnaire on medical errors committed during residency to 254 residents in three internal medicine training programs. [105]  One-hundred and fourteen (45%) of them reported mistakes, and another 56 (22%) returned a postcard acknowledging receipt but declining to complete the questionnaire.  Of the 114 who reported errors, 51% believed their most serious mistake in the last year was caused by having too many other tasks, and 41% attributed their mistake to fatigue.  Thirty-one percent of the residents reported that their mistakes had resulted in the deaths of their patients.

Green goes so far as to claim that the strongest argument for reducing resident work hours is an ethical one: overwork interferes with the development of professional values and attitudes that are an essential part of being a physician.[106]  Fatigue can cultivate anger, resentment, and bitterness—often directed at the patient—rather than kindness, compassion, or empathy.  As was evident from many of the studies on negative mood and its relationship to long work hours, this attitude is promoted when meeting a patient’s needs becomes incompatible with meeting a resident’s own needs.  The following entry from a resident’s diary illustrates the types of feelings engendered by typical resident work schedules:

1AM and I’m ready to go to bed: one should never be ready to go to bed in the ICU [Intensive Care Unit] — you’ll always be disappointed.  Anyway, I’m on my way to the EW [Emergency Ward] … when there’s a code [cardiac arrest].  Get up there and find [a resident] trying to intubate a lifetime asthmatic who is as blue as this ink.  I keep thinking — he’s blue enough to go to the ICU.  I keep hoping he’s going to be too blue to go anywhere.  Probably a nice man with a loving wife and concerned children, but I don’t want that SOB to make it because I’ve got one special who is going to keep me up 2 more hours.  I don’t need an intubated, blue, pneumothoraxed SOB coming to my unit… I don’t want the asthmatic SOB to live if it means I don’t sleep.  I don’t want the special to live if it means I don’t sleep.  I just want sleep.[107]

Although patient protection is not the principal focus of this petition, it is clear that in protecting medical residents from long work hours, OSHA would also be protecting patients. 


The following arguments are often made against reducing housestaff hours: (1) continuity of patient care will be disrupted, (2) long hours are necessary in order to sufficiently train physicians, (3) training under conditions of fatigue and sleep deprivation prepares residents to function should these conditions arise in future practice, and (4) long hours promote favorable character attributes in physicians, such as discipline, endurance, responsibility, self-reliance, confidence, collegiality, and humility.  Although there may be a degree of validity to each of these arguments, the signers of this petition believe that two important questions must be asked:  What is the incremental gain on any of these measures of working 100 hours a week vs. 80 hours a week, and what are the risks against which these incremental benefits are being weighed?  The risks have been reviewed in this petition, and have been shown to be substantial to both patients and physicians.  On the other hand, there are no data supporting the claims of proponents of the current situation. 

The continuity-of-care argument claims that (1) long continuous hours worked by an individual physician are necessary for the delivery of good care to his or her patients, and that (2) long continuous hours are necessary for the physician to receive adequate training.  First, it is inevitable that there will be breaks in the continuity of care for patients.  Attending physicians, for example, are scheduled to go home after shifts much shorter than those worked by residents.  Residents, on the other hand, have been traditionally expected to subordinate their needs to the “ideal” of spending as much time with their patients as possible.  With studies increasingly demonstrating serious harm to medical residents as a result of these long hours, the balance between patient care and resident well-being must be retuned to incorporate scientific evidence.  Second, residents need not spend 36 continuous hours at a hospital in order to learn from their patients.  Most diseases are seen a number of times in the course of multi-year residency programs, and residents can also learn from the experience or reports of other residents.  Finally, hospital stays have shortened dramatically in recent years, further weakening the continuity-of-care argument. 

Nor is there evidence that 80 total hours per week is insufficient to train proficient physicians.  The profession of medicine is a continuous learning process, with many opportunities for necessary skills to be obtained throughout a physician’s career.  With the rapidly changing nature of medicine, the ability to develop a well-balanced life with a well-established ability to continue to learn are at least as important as any experience to be gained between the 80th and 100th hour of a work week.  Moreover, much of a resident’s time is spent on non-educational activities.  In a time-motion study of internal medicine residents at two urban hospitals in New York City, Knickman found that 19% of internal medicine residents’ time was spent on activities that could be done by nurses, laboratory technicians, or other staff.[108]  A meager 3.1% of their time was actually spent exclusively seeing patients.  This raises the question of why more ancillary staff are not hired to allow residents added time for purely educational activities and to reduce work hours.  Green has argued that the real reason is that it is economically more favorable for hospitals to use residents as inexpensive labor to complete necessary tasks than to hire ancillary personnel who require higher salaries.[109]  It has been calculated that substituting ancillary staff to provide services currently provided by resident physicians would cost hospitals approximately $58,000 – $78,000 annually (in 1993 dollars) per resident, because residents’ salaries per hour actually worked are so low.[110]

The argument that, by training long hours, residents will learn to function under harsh conditions raises the issue of the mismatch between residency training and medical practice post-residency.  Very few physicians work 100 hours a week after residency, and on-call shifts occur less frequently than during residency.  The benefit of being able to function on the rare occasion post-residency when a physician has had minimal sleep must be weighed against the three-to-five years in which the fatigued, sleep-deprived, and error-prone resident may be harming both himself or herself and the patients he or she is taking caring of. 

Lastly, the consideration of benefit vs. risk also applies to the argument that training long hours engenders character in residents.  As Green points out, although residents may be building stamina and learning to be disciplined, self-reliant, and confident, they may also be undergoing socialization against their patients, dulling their compassion.  Moreover, there are many less harmful ways (to physicians and patients) for developing the same positive character attributes in physicians. 

In sum, whereas there is evidence that physical and mental harm comes to residents and  that performance decreases with sleep deprivation and fatigue, there is no evidence to suggest that reducing work hours would decrease the quality of patient care, that 80 hours per week is insufficient to train residents, that residents would not be able to provide care under harsh conditions without training greater than 80 hours per week, or that residents are better people because they trained more than 80 hours per week.  Moreover, there is positive evidence that medical errors decrease with reduced work schedules[111] and that residents’ moods get better with rest,[112] both of which can translate into better patient care.  If policy is to be based on evidence and not opinion, the federal government is obligated to protect medical residents and their patients by reducing resident work hours.


In November of 1999 the National Labor Relations Board (NLRB) overturned the 1976 Cedars-Sinai, St. Clare’s Hospital precedent that ruled that medical residents are primarily students rather than employees, stating that this previous determination was “flawed in many respects,” and that “Ample evidence exists here to support our finding that interns, residents, and fellows fall within the broad definition of ‘employee’…”[113]  The Board argued overall that medical residents’ “status as students is not mutually exclusive of a finding that they are employees.”

The NLRB’s rationale had four elements.  First, residents and fellows (called “housestaff”) were determined to work for an employer (the hospital) within the meaning of the Act.  Second, housestaff are compensated for their services by the hospital, and the hospital withholds Federal and state income taxes, as well as social security, from their salaries.  Further, housestaff receive fringe benefits reflective of employee status, such as worker’s compensation, paid vacations and sick leave, parental and bereavement leave, and insurance coverage such as health, dental, life, and malpractice.  Third, housestaff provide patient care for the hospital, the service that the hospital sells.  Lastly, housestaff were also found to be unlike students in that they do not pay tuition or student fees. 

The NLRB thus fundamentally changed the thinking applied to medical residents, asserting that housestaff are no longer considered primarily students: “That they [housestaff] also obtain educational benefits from their employment does not detract from this fact.  Members of all professions continue learning throughout their careers…”  The Board concluded that “house staff are employees … and … are therefore entitled to all the statutory rights and obligations that flow from our conclusion.”  The signers of this petition believe that these rights entitle housestaff not only the freedom to organize for collective bargaining, but also to expect that OSHA will protect them as employees from unsafe labor practices, the way they would workers in other industries.   The recent NLRB decision clearly brings regulation of resident work hours under  OSHA’s jurisdiction. 

Moreover, whereas hours-of-service regulations put forth by the DOT, for example, have been made possible by the existence of industry-specific statutes, there are no such statutes for resident working conditions, and OSHA is therefore the only government organization within whose purview regulation of housestaff hours falls.  Furthermore, OSHA has previously established standards for the protection of health-care workers when it promulgated the Bloodborne Pathogens Standard[114] to protect health-care workers from needlestick injuries. 


Petitioners’ Requests 

Public Citizen, the Committee of Interns and Residents, the American Medical Student Association, Dr. Bertrand Bell, and Dr. Kingman P. Strohl, request that the following hours-of-service regulations be applied to medical housestaff in all residency and fellowship programs: 

(1)   a limit of 80 hours of work per week,
(2)   a limit of 24 consecutive hours worked in one shift,
(3)   a limit of on-call shifts to every 3rd night,
(4)   a minimum of 10 hours off-duty time between shifts,
(5)   provision of at least one 24-hour period of off-duty time per week, and
(6)   for emergency medicine residents working in hospitals receiving more than 15,000 unscheduled patient visits per year, a limit of 12 consecutive hours on-duty per day. 

On-duty hours should be measured as time at the work-site.  Off-duty hours should be measured as time away from the hospital not on-call.  The regulations shall not be construed to require or to permit a physician to abandon his or her patient in need of emergency or critical care.  In an extreme situation proven by the hospital to be a patient care emergency and when that physician cannot be replaced by a rested physician or his or her skills are not replaceable, in order to complete that patient’s emergency care, the physician’s work that day may exceed the maximum 24 consecutive hour shift or 80 hour per week limit.  This does not mean that resident physicians may be scheduled to be present at a work-site in excess of the 24-hour-per-day or 80-hour-per-week limit or that new patients may be assigned at or near the end of that time. 

These requested regulations differ from the New York Bell Regulations in several ways.  First, whereas the Bell Regulations call for an 80-hour work week averaged over a 4-week period, the signers of this petition are calling for an 80-hour work week that is not averaged.  The signers believe that two 100-hour weeks and two 60-hour weeks are not consistent with minimizing harm to physicians and patients.  Second, the Bell Regulations include exemptions for surgical residencies, allowing them to exceed 80 hours of work per week.  As stated previously, we believe the same standard should apply to all specialties as all residents are  equally susceptible to the harms of long work hours.  We therefore request an 80-hour limit of work per week applied to all residency and fellowship programs.  Third, the Bell Regulations do not require all residency and fellowship programs to schedule on-call shifts only every 3rd night (the restriction exists as one of the conditions that must be met in order for surgery programs to schedule residents and fellows in excess of 80 hours per week).  The signers of this petition request that call be scheduled only every 3rd night in all residency and fellowship programs.  Fourth, the Bell Regulations mandate a minimum 8-hour rest period, whereas the signers of this petition call for a minimum 10-hour rest period.  Given that it is possible to arrange schedules incorporating the requested rest period, we believe that 10 hours off-duty between shifts will more adequately allow for residents and fellows to accomplish necessary tasks outside of the hospital, such as obtaining sufficient sleep, commuting, and attending to familial and domestic responsibilities. 


Furthermore, the signers of this petition request that strict enforcement accompany the regulations, as inadequate enforcement policies have been a source of failure in previous efforts to limit resident work hours:  

(1)   Resident and fellow schedules should be recorded and kept as public records by the hospitals, available for OSHA inspection.
(2)   There must be an official and confidential procedure for whistle blowers to report violations to OSHA.
(3)   Unannounced inspections should occur on a frequent basis.
(4)   OSHA must establish enforcement mechanisms so that violations incur fines sufficient to deter those violations.


Evidence convincingly demonstrates that excessive work schedules risk residents’ lives and health, in terms of automobile safety, mental health, and pregnancy complications.  OSHA can readily adopt the standards requested in this petition, which are based on those already in place in New York and to a lesser extent those put forth by the ACGME.  It has been demonstrated that implementation of an 80-hour work week is possible, and, in fact, well-received by residents.[115]  After establishing the requested regulations, as a second and necessary step, OSHA should then proceed to investigate resident work hours more fully to determine if even more stringent standards are required.  As elegantly defined in its enabling legislation,[116] the Occupational Safety and Health Act of 1970, OSHA’s mission is to “Assure so far as possible every working man and woman in the Nation safe and healthful working conditions.”[117]  The signers of this petition believe that for OSHA to immediately establish the requested work-hour regulations for medical residents and fellows would be to fulfill this mission. 

Anandev Gurjala
Medical Student
Northwestern Medical School 

Peter Lurie, M.D., M.P.H.
Deputy Director
Public Citizen’s Health Research Group 

Ladi Haroona, M.D.
Committee of Interns and Residents 

Joshua P. Rising
Legislative Affairs Director
American Medical Student Association 

Bertrand Bell, M.D.
Professor of Medicine
Albert Einstein College of Medicine 

Kingman P. Strohl, M.D.
Professor of Medicine and Director of Center for Sleep Disorders Research
Case Western Reserve University 

Sidney M. Wolfe, M.D.
Public Citizen’s Health Research Group


Table 1. ACGME requirements for selected specialties.



Weekly Work Hours

Longest Shift

Nights On-Call

Non-working hours

Diagnostic Radiology

None set

None set

No more than every 3rd night, on average

1 full day out of every 7

Emergency Medicine

60 hrs seeing patients and 72 hrs total

12 hrs, at least 1 equivalent period of time off between shifts

None set

1 full day out of every 7

Family Practice

None set

None set

No more than every 3rd night, averaged monthly

1 day out of 7, averaged monthly

General Surgery

Whatever is considered “appropriate” by residency directors

None set

No more than every 3rd night, on average

1 day out of every 7, on average

Neurological Surgery


None set

No more than every 3rd night

1 day out of every 7

Obstetrics and Gynecology

None set

None set

No more than every 3rd night

1 full day out of every 7, on average


80 hrs averaged over 4 wks

None set

No more than every 3rd night, on average

1 day out of every 7, on average

Orthopedic Surgery

None set

None set

No more than every 3rd night

1 day out of every 7


None set

On ER shifts no more than 12 hrs, with 8 hrs separation

Average of every 3rd or 4th night

Monthly average of 1 day out of 7

Preventive Medicine

80 hrs averaged over 4 wks

Adequate opportunity to rest and sleep when on duty for 24 hrs or more

No more than every 3rd night

1 day out of every7 on average

” src=”/CuteSoft_Client/CuteEditor/Images/anchor.gif”>

Table 2. Work Hour Regulations in Other Countries



Total Hours Duty per Week

Maximum Hours on Duty (Total and Consecutive)

Maximum Consecutive Shifts

Minimum Rest Hours


75 hrs/wk (Western and Victoria)
70 hrs/wk (Tasmania)
68hrs/wk (South)

24 consecutive hours for a shift (Capital Territory)




45 hrs/wk



11 hrs between shifts, although can be reduced to 8

United Kingdom*

56 hrs averaged (cycle time not specified)

72 hrs max/wk

14-16 total hrs/regular shift

16-24 consecutive hrs for a shift

32 consecutive hrs on weekdays and 56 consecutive hrs on weekends when on-call

12 regular shifts in a row

8 hrs during 32 hr period when on-call

12 hrs after being on-call

8 hrs between regular shifts

European Union

48 hrs/wk, including overtime†

Night work must not exceed 8 hrs on average


A rest break when the working day is >6hrs

11 consecutive hrs/day


56 hrs averaged/wk over 24 wks

7.5-10 hrs/day in addition to 12 hrs on-call or 24 hrs on-call when off-duty

24 consecutive hrs max

12 consecutive on-call duty periods

30 mins/duty period of 6-9 hrs or 45 mins when duty period >9 hrs

guaranteed 10 consecutive hrs off after duty >7.5 hrs


48 hrs averaged/wk over 13 wks

60 hrs max/week

13 or 15 hrs total/day, although can be extended up to 3 hrs

9.5 hrs for a night shift

24 max consecutive hrs

5 shifts worked consecutively/wk for a max 13 wks in 26

30 mins when worked 5.5 hrs

10 consecutive hrs between shifts

* by collective agreement.
† by the year 2003.

Table 3. Hours-of-Service Regulations in the Department of Transportation.


Motor Carrier (49 CFR Part 395)

• Drivers may drive for 10 hours or be on duty for 15 hours.
• Drivers must have 8 consecutive hours off following a 10/15 hour on-duty period.
• If drivers use a sleeper berth, they may split the 8-hour period into two periods as long as neither period is less than 2 hours.
• Drivers may not exceed 70 hours in 8 days, if the carrier operates 7 days a week.
• Drivers may not exceed 60 hours in 7 days if the carrier does not operate every day of the week.

Aviation (14 CFR Part 121; 14 CFR Part 135)

• Pilots flying domestic Part 121 operations may fly up to 30 hours per week, 100 hours per month, and 1,000 hours per year.
• Pilots flying domestic Part 135 operations may fly up to 34 hours per week, 120 hours per month, and 1,200 hours per year.
• If the scheduled flight time is less than 8 hours, the minimum rest period in the 24 hours preceding the scheduled completion of the flight segment is 9 hours. This time may be reduced to 8 hours if the following rest period, to begin no later than 24 hours after the commencement of the reduced rest period, is increased to 10 hours.
• If the scheduled flight time is 8–9 hours, the minimum rest period in the 24 hours preceding the scheduled completion of the flight segment is 10 hours. This time may be reduced to 8 hours if the following rest period, to begin no later than 24 hours after the commencement of the reduced rest period, is increased to 11 hours.
• If the scheduled flight time is equal to or greater than 9 hours, the minimum rest period in the 24 hours preceding the scheduled completion of the flight segment is 11 hours. This time may be reduced to9 hours if the following rest period, to begin no later than 24 hours after the commencement of the reduced rest period, is increased to 12 hours.

Rail (49 U.S.C. 211; 49 CFR Part 228)

• Maximum duty limit of 12 hours.
• Must be off-duty for 10 consecutive hours, after working 12 consecutive hours or off 8 consecutive hours if worked less than 12 consecutive hours.
• Time spent in transportation (deadheading) to duty assignment counts toward on-duty time.
• Time deadheading from duty assignment does not count toward on-duty or off-duty time.

Marine (46 U.S.C. 8104; 46 CFR Parts 15.705, 15.710, and 15.1111)

• Hours-of-service or watch requirements vary depending on type of vessel.
• An officer must be off duty for at least 6 hours within the 12 hours immediately before leaving port before taking charge of the deck watch on a vessel when leaving port.
• On an oceangoing or coastwise vessel of not more than 100 gross tons (GT), a licensed individual may not work more than 9 of 24 hours when in port or more than 12 of 24 hours at sea, except in an emergency.
• On a tanker, a licensed individual or seaman may not work more than 15 hours in any 24-hour period or more than 36 hours in any 72-hour period, except in an emergency or a drill.
• Officers in charge of a navigational or engineering watch on board any vessel that operates beyond the boundary line shall receive a minimum of 10 hours rest in any 24-hour period. The hours of rest may be divided into no more than two periods, of which one must be at least 6 hours in length. The hours of rest do not need to be maintained in an emergency. The hours of rest may be reduced to 6 hours if no reduction extends beyond 2 days and not less than 70 hours of rest are provided in each 7-day period.


CFR = Code of Federal Regulations; U.S.C. = United States Code.

Figure 1. Relative Risk of Fatigue Crash by Hours Driving, TIFA 1991-1996


[1] 29 USC § 652(8).

[2] Lagnado L.  Hospitals overwork young doctors in New York.  Wall Street Journal.  May 19, 1998.

[3] Occupational Safety and Health Administration.  OSHA 30-year milestones.  http://www.osha.gov/as/opa/osha30yearmilestones.html.

[4] Daugherty SR, Baldwin DC, Rowley BD.  Learning, satisfaction, and mistreatment during medical internship: a national survey of working conditions.  Journal of the American Medical Association 1998;279(15):1194-1199.

[5] Surgeons in training, running on empty; young doctors spend 110 hours per week or more doing their residencies.  ABC News: Nightline. September 28, 2000.

[6] Schwartz AJ, Black ER, Goldstein MG, Jozefowicz RF, Emmings FG.  Levels and causes of stress among residents.  Journal of Medical Education 1987;62:744-53.

[7] Small GE.  House officer stress syndrome.  Psychosomatics 1981;22(10):860-9.

[8] Daugherty SR, Baldwin DC, Rowley BD.  Learning, satisfaction, and mistreatment during medical internship: a national survey of working conditions.  Journal of the American Medical Association 1998;279(15):1194-1199.

[9] Alam M.  Mistreatment and maladaptations during medical internship.  Journal of the American Medical Association 1998;280(8):699-700.

[10] Landau C, Hall SA, Wartman S, Macko MB.  Stress in social and family relationships during the medical residency.  Journal of Medical Education 1986;61:654-660.

[11] Bergman AS.  Marital stress and medical training: an experience with a support group for medical house staff wives.  Pediatrics 1980;65(5):944-7.

[12] Peters RD, Kloeppel E, Alicandri E, et al.  Effects of partial and total sleep deprivation on driving performance. Federal Highway Administration Publication No. FHWA-RD-94-046.

[13] Lyznicki JM, Doege TC, Davis RM, Williams MA.  Sleepiness, driving, and motor vehicle crashes.  Journal of the American Medical Association 1998;279(23):1908-13.

[14] Henican, E.  “Their sleepless nightmares.”  Newsday.  February 3, 1999.

[15] Lowry RT.  An end-of-shift tale.  Annals of Emergency Medicine 1998;31(2): 287-288.

[16] Rovner S. Young M.D.’s long hours ill-advised?  Sleep deprivation leads to horror tales.  Rochester Democrat and Chronicle.  October 21, 1990.

[17] Surgeons in training, running on empty; young doctors spend 110 hours per week or more doing their residencies.  ABC News: Nightline. September 28, 2000.

[18] Wendt JR, Yen LJ.  The resident by moonlight: a misguided missile.  Journal of the American Medical Association 1988;259(1):43-44.

[19] Worth R.  Exhaustion that kills: why residents are still overworked—and what we can do about it.  Washington Monthly. Jan/Feb 1999;15-20.

[20] Kowalenko T, Haas-Kowalenko J, Rabinovich A, Grzybowski M.  Emergency medicine residency related MVCs—is sleep deprivation a risk factor?  Academic Emergency Medicine 2000;7(5):451.

[21] Marcus CL, Loughlin GM. Effect of sleep deprivation on driving safety in housestaff. Sleep 1996;19:763-766.

[22] Steele MT, MA OJ, Watson WA, Thomas HA, Muelleman RL. The occupational risk of motor vehicle collisions for emergency medicine residents. Academic Emergency Medicine 1999;610:1050-1053.

[23] Geer RT, Jobes DR, Tew JD, Stepsis LH.  Incidence of automobile accidents involving anesthesia residents after on-call duty cycles.  Anesthesiology 1997;87(3A):A938.

[24] Personal communication with the Committee for Interns and Residents.  Mark Levy.  April 20th, 2001.

[25] Wendt JR, Yen LJ.  The resident by moonlight: a misguided missile.  Journal of the American Medical Association 1988;259(1):43-44.

[26] Stoudemire A.  Clinical Psychiatry for Medical Students.  New York: Lippincott-Raven, 1998.

[27] Friedman RC, Kornfeld DS, Bigger TJ. Psychological problems associated with sleep deprivation in interns. Journal of Medical Education 1973;48:436-441.

[28] Berkoff K, Rusin W. Pediatric house staff’s psychological response to call duty. Developmental and Behavioral Pediatrics 1991;12:6-10.

[29] Leonard C, Fanning N, Attwood J, Buckley M. The effect of fatigue, sleep deprivation and onerous working hours on the physical and mental well being of pre-registration house officers.  Irish Journal of Medical Sciences 1998;167(1):22-5.

[30] Lingenfelser TH, Kaschel R, Weber A, Zaiser-Kaschel H, Jakober B, Kuper J. Young hospital doctors after night duty: their task-specific cognitive status and emotional condition. Medical Education 1994;28:566-572.

[31] Hart RP, Buchsbaum, DG, Wade JB, Hamer RM, Kwentus JA. Effect of sleep deprivation on first-year residents’ response times, memory and mood. Journal of Medical Education 1987; 62: 940-942.

[32] Friedman RC, Bigger JT, Kornfeld DS.  The intern and sleep loss.  New England Journal of Medicine 1971;285(4):201-203.

[33] Samkoff JS, Jacques CHM. A review of studies concerning effects of sleep deprivation and Fatigue on Residents’ Performance. Academic Medicine 1991;66(11):687-693.

[34] American Psychiatric Association.  Diagnostic and statistical manual of disorders: DSM-IV, 4th ed.  Washington D.C.: American Psychiatric Association, 1994. 

[35] Valko RJ, Clayton PJ.  Depression in the internship.  Diseases of the Nervous System. 1975;36:26-9.

[36] Pitts FN, Schuller AB, Rich CL, Pitts AF.  Suicide among U.S. women physicians 1967-1972.  American Journal of Psychiatry 1979;136:694-696.

[37] Reuben DB. Depressive symptoms in medical house officers: effects of level of training and work rotation. Archives of Internal Medicine 1985; 145:286-288.

[38] Valko RJ, Clayton PJ.  Depression in the internship.  Diseases of the Nervous System 1975;36(1):26-29.

[39] Reuben DB. Psychologic effects of residency.  Southern Medical Journal 1983;76: 380-383.

[40] Lingenfelser TH, Kaschel R, Weber A, Zaiser-Kaschel H, Jakober B, Kuper J. Young hospital doctors after night duty: their task-specific cognitive status and emotional condition. Medical Education 1994;28:566-572.

[41] Phelan ST. Pregnancy during residency: I. The decision “to be or not to be.”  Obstetrics and Gynecology 1988;72: 425-31.

[42] Mozurkewich EL, Luke B, Avni M, Wolf FM. Working conditions and adverse pregnancy outcome: a meta-analysis. Obstetrics and Gynecology 2000;95:623-635.

[43] Osborn LM, Harris DL, Reading JC, and Prather MB.  Outcome of pregnancies experienced during residency.  The Journal of Family Practice 1990;31(6):618-22.

[44] Klebanoff MA, Shiono PH, Rhoads GG. Outcomes of pregnancy in a national sample of resident physicians. New England Journal of Medicine 1990;323:1040-5.

[45] Klebanoff MA, Shiono PH, and Rhoads GG.  Spontaneous and induced abortion among resident physicians.  Journal of the American Medical Association 1991;261(21):2821-5.

[46] Grunebaum A, Minkoff H, Blake D. Pregnancy among obstetricians: a comparison of births before, during and after residency. American Journal of Obstetrics and Gynecology 1987;157:79-83.

[47] Silva BM. Pregnancy during residency: a look at the issues. Journal of the American Medical Women’s Association 1992;47: 71-75.

[48] Katz VL, Miller NH, Bowes WA. Pregnancy complications of physicians. Western Journal of Medicine 1988;149:704-707.

[49] New York State Health Code, Section 405.

[50] Kennedy R.  Residents’ hours termed excessive in hospital study.  The New York Times.  May 19, 1998; Section A:1.

[51] Henican E.  Their sleepless nightmares.  Newsday.  February 3, 1999.

[52] Mangan KS.  Cutting back to an 80-hour week.  The Chronicle of Higher Education.  September 15, 2000: STUDENTS;A43.

[53] Mangan KS.  Cutting back to an 80-hour week.  The Chronicle of Higher Education.  September 15, 2000: STUDENTS;A43.

[54] Kennedy R.  Residents’ hours termed excessive in hospital study.  The New York Times.  May 19, 1998; Section A:1.

[55] Committee of Interns and Residents.  Enforcement strengthened in 2000.  CIR-SEIU News.  March 2001.

[56] Mangan KS.  Cutting back to an 80-hour week.  The Chronicle of Higher Education.  September 15, 2000: STUDENTS;A43.

[57] Accreditation Council for Graduate Medical Education.  Program requirements for residency education.  http://www.acgme.org

[58] Dunn MR.  Programs and institutions cited for work hours and related requirements during 1999.  ACGME Bulletin.  April 2000.

[59] Greene J.  More residencies cited for work violations.  American Medical News.  March 6, 2000.

[60] Sharfstein JM.  Asleep on the job.  The New Republic.  June 21, 1999:17.

[61] Australian Medical Association.  AMA safe hours campaign.  http://www.domino.ama.com.au/dir0103/IRRemun.nsf/96ec31299990db634a2568e…. Now available at: http://www.ama.com.au/node/3757.

[62] Reuters.  European parliament backs limits on physician work hours.  November 17, 1999.

[63] National Transportation Safety Board.  NTSB Safety Recommendation A-99-45, as cited in National  Transportation Safety Board.  Evaluation of U.S. Department of Transportation efforts in the 1990s to address operator fatigue. Safety Report NTSB/SR-99/01.  May 1999.

[64] National  Transportation Safety Board.  Evaluation of U.S. Department of Transportation efforts in the 1990s to address operator fatigue. Safety Report NTSB/SR-99/01.  May 1999.

[65] Knipling, RR, Wang JS. October 1995.  Revised estimates of the U.S. drowsy driver crash problem size based on general estimates system case reviews.  In: 39th Annual Proceedings, Association for the Advancement of Automotive Medicine, Chicago, IL.  October 16–18, 1995.

[66] National  Transportation Safety Board.  Evaluation of U.S. Department of Transportation efforts in the 1990s to address operator fatigue. Safety Report NTSB/SR-99/01.  May 1999.

[67] Federal Register, Vol. 65, No. 85, May 2, 2000, pp. 25540-25611.

[68] Federal Register, Vol. 65, No. 85, May 2, 2000, pp. 25540-25611.

[69] AAA Foundation for Auto Safety.  A report on the determination and evaluation of the role of fatigue in heavy truck accidents.  October 1985. 

[70] National Transportation Safety Board.  Fatigue, alcohol, other drugs, and medical factors in fatal-to-the-driver heavy truck crashes. Safety Study NTSB/SS-90/01 and NTSB/SS-90/02. Washington, DC. February 1990, p. 70.

[71] Harris W, Mackie RR, et al.  A study of the relationships among fatigue, hours of service, and safety of operations of truck and bus drivers.  Human Factors Research, Inc. (Springfield, VA, National Technical Information Service, November 1972 [PB-213 963]).

[72] Brown, ID.  Driver fatigue.  Human Factors 1994;36(2): 298-314.

[73] Hamelin P.  Lorry drivers’ time habits in work and their involvement in traffic accidents.  Ergonomics 1987;30(9):1323-1333.

[74] Saccomanno FF, Shortreed JH, and Yu M.  Effect of driver fatigue on commercial vehicle accidents.  [Truck Safety: Perceptions and Reality.  Waterloo, Ontario: Institute for Risk Research, University of Waterloo.  1996] pp. 157-174.

[75] Jones IS, Stein HS.  Effect of driver hours of service on tractor-trailer crash involvement.  Arlington, VA: Insurance Institute of Highway Safety. September 1987.

[76] Lin TD, Jovanis PP, Yang CZ.  Time of day models of motor carrier accident risk.  Transportation Research Record 1467.  Washington, DC: Transportation Research Board. 1994;1-8.

[77] Folkard S.  “Time on shift” effects in safety: a mini-review.  Shiftwork International Newsletter 1995;12(1):16.

[78] Bowen V.  Utility service vehicle study. Richmond, VA: University of Richmond.  August 1996.  (FHWA docket number FHWA-97-2350-82)

[79] National Transportation Safety Board.  Factors that affect fatigue in heavy truck accidents.  Safety Study NTSB/SS-95/01 and NTSB/SS-95/02. Washington, DC. 1995.

[80] Federal Register, Vol. 65, No. 85, May 2,2000, pp. 25540-25611.

[81] Federal Register, Vol. 60, No. 244, December 20, 1995.

[82] Airlines unveil plan to reduce pilot fatigue.  http://www.cnn.com/2000/TRAVEL/NEWS/09/13/pilot.fatigue/index.html. Now available at: http://web.archive.org/web/20040420110316/http://www.cnn.com/2000/TRAVEL/NEWS/09/13/pilot.fatigue/index.html.

[83] National Transportation Safety Board.  A review of flightcrew-involved, major accidents of U.S. air carriers, 1978 through 1990. Safety Study NTSB/SS-94/01. Washington, DC. 1994.

[84] National Transportation Safety Board.  Aviation safety in Alaska. Safety Study NTSB/SS-95/03. Washington, DC. 1995

[85] Oral statement of Administrator Joline Molitoris for Senate Safety Hearing.  September 16, 1998.

[86] National Transportation Safety Board.  NTSB Railroad Accident Report, Atchison, Topeka and Santa Fe Railway Company (ATSF) freight trains ATSF 818 and ATSF 891 on the ATSF railway, Corona, CA, November 7, 1990.  August 23, 1991, as cited in National  Transportation Safety Board.  Evaluation of U.S. Department of Transportation efforts in the 1990s to address operator fatigue. Safety Report NTSB/SR-99/01.  May 1999.

[87] National Transportation Safety Board.  NTSB Railroad Accident Report, collision involving two New York City subway trains on the Williamsburg Bridge in Brooklyn, NY, June 5, 1995.  September 11, 1996, as cited in National  Transportation Safety Board.  Evaluation of U.S. Department of Transportation efforts in the 1990s to address operator fatigue. Safety Report NTSB/SR-99/01.  May 1999.

[88] McCallum, MC, Raby, M, Rothblum,AM.  Procedures for investigating and reporting human factors and fatigue contributions to marine casualties. CG-D-09-97.  U.S. Department of Transportation, U.S. Coast Guard, Marine Safety and Environmental Protection.  Washington, DC. 1996.

[89] US Coast Guard.  Fatigue and alertness in merchant marine personnel: a field study of work and sleep patterns.  US Coast Guard Research and Development Center, Groton, CT. Accession No. AD-A322126.  June 1996.

[90] Exxon Valdez Oil Spill Trustee Council.  http://www.oilspill.state.ak.us/history/history.htm

[91] National Transportation Safety Board.  NTSB marine accident report, grounding of the U.S. Tankship Exxon Valdez on Bligh Reef, Prince William Sound near Valdez, Alaska, March 24, 1989, September 18, 1990, as cited in National  Transportation Safety Board.  Evaluation of U.S. Department of Transportation efforts in the 1990s to address operator fatigue. Safety Report NTSB/SR-99/01.  May 1999.

[92] Department of Transportation.  Conference summary, partnering for transportation safety: human-centered systems: operator fatigue management.  Tysons Corner, VA.  August 29-30, 2000.  http://scitech.dot.gov/reeng/humperf.humperf.html.

[93] Hawkins MR, Vichick DA, Silsby HD, Kruzich DJ, Butler R.  Sleep and nutritional deprivation and performance of house officers.  Journal of Medical Education 1985;60:530-5.

[94] Hart RP, Buchsbaum DG, Wade JB, Hamer RM, Kwentus JA.  Effect of sleep deprivation on first-year residents’ response times, memory, and mood.  Journal of Medical Education 1987;62:940-2.

[95] Rubin R, Orris P, Lau SL, Hryhorczuk DO, Furner S, Letz R.  Neurobehavioral effects of the on-call experience in housestaff physicians.  Journal of Occupational Medicine 1991;33:13-8.

[96] Robbins J, Gottlieb F.  Sleep deprivation and cognitive testing in internal medicine house staff.  Western Journal of Medicine 1990;152:82-6.

[97] Friedman RC, Bigger JT, Kornfeld DS.  The intern and sleep loss.  New England Journal of Medicine 1971;285:201-3.

[98] Denisco RA, Drummond JN, Gravenstein JS.  The effect of fatigue on the performance of a simulated anesthetic monitoring task.  Journal of Clinical Monitoring 1987;3:22-4.

[99] Smith-Coggins R, Rosekind MR, Buccino KR.  Rotating shiftwork schedules: can we enhance physician adaptation to night shifts?  Academic Emergency Medicine 1997;4:951-961.

[100] Taffinder NJ, McManus IC, Russell RCG, Darzi A.  Effect of sleep deprivation on surgeons’ dexterity on laparoscopy simulator.  The Lancet 1998;352:1191.

[101] Dawson D, Reid K.  Fatigue, alcohol and performance impairment.  Nature 1997;388:235.

[102] Hillson SD, Dowd B, Rich EC, Luxenbeurg MG.  Call nights and patient care: effects on inpatients at one teaching hospital.  Journal of General Internal Medicine 1992;7:405-410.

[103] Gottlieb DJ, Parenti CM, Peterson CA, Lofgren RP.  Effect of a change in house staff work schedule on resource utilization and patient care.  Archives of Internal Medicine 1991;151(1):2065-70.

[104] Laine C, Goldman L, Soukup JR, Hayes JG.  The impact of a regulation restricting medical house staff working hours on the quality of patient care.  Journal of the American Medical Association 1993;269:374-378.

[105] Wu AW, Folkman S, McPhee SJ, Lo B.  Do house officers learn from their mistakes?  Journal of the American Medical Association 1991;265:2089-94.

[106] Green MJ.  What (if anything) is wrong with residency overwork?  Annals of Internal Medicine 1995;123(7):512-7.

[107] Groopman LC.  Medical internship as moral education: an essay on the system of training physicians.  Culture, Medicine, and Psychiatry 1987;11(2):207-27.

[108] Knickman JR, Lipkin M Jr, Finkler SA, Thompson WG, Kiel J.  The potential for using non-physicians to compensate for the reduced availability of residents.  Academic Medicine 1992;67:429-38.

[109] Green MJ.  What (if anything) is wrong with residency overwork?  Annals of Internal Medicine 1995;123(7):512-7.

[110] Stoddard JJ, Kindig DA, Libby D.  Graduate medical education reform.  Service provision transition costs.  Journal of the American Medical Association 1994;272:53-8.

[111] Gottlieb DJ, Parenti CM, Peterson CA, Lofgren RP.  Effect of a change in house staff work schedule on resource utilization and patient care.  Archives of Internal Medicine 1991;151(1):2065-70.

[112] Berkoff K, Rusin W. Pediatric house staff’s psychological response to call duty. Developmental and Behavioral Pediatrics 1991;12:6-10.

[113] National Labor Relations Board.  Boston medical center corporation and house officers’ association/committee of interns and residents, petitioner, Case 1-RC-20574.  November 26, 1999.

[114] 29 C.F.R. § 1910.1030

[115] Conigliaro J, Frishman WH, Lazar EJ, Croen L.  Internal medicine housestaff and attending physician perceptions of the impact of New York State Section 405 regulations on working conditions and supervision of residents in two training programs.  Journal of General Internal Medicine, 1993;8:502-507.

[116] P.L. 91-596

[117] Occupational Safety and Health Administration.  OSHA Strategic Plan.  http://www.osha.gov/oshinfo/strategic/pg1.html#intro. Now available at: http://web.archive.org/web/20030418083414/http://www.osha.gov/oshinfo/strategic/pg1.html.