Graduate Education

Graduate Education Long hours and little sleep: Work schedules of residents in obstetrics and gynecology Deidre M. Defoe, MD, MPH, Michael L. Power, PhD, Gerald B. Holzman, MD, Andrea Carpentieri, MA, and Jay Schulkin, PhD Objective: To investigate residents’ work schedules and their attitudes toward limiting their hours. Methods: An anonymous survey regarding resident work hours and call schedules was administered to the 4674 obstetric-gynecologic residents who took the year 2000 Council on Resident Education in Obstetrics and Gynecology in-training examination. Results: A total of 4510 surveys were analyzed (96.5%). Three of four (75.5%) respondents reported working between 61 and 100 hours each week. Most (71.3%) reported sleeping less than 3 hours while on night call. Eight of ten reported having postcall clinical responsibilities. The reported number of hours on call declined and the reported number of hours of sleep increased with year of residency. Three of four residents wanted limits on their work hours. Residents who reported longer on-call hours or less sleep during night shift were significantly more likely to want a restriction on work hours. Fatigue was the most commonly selected reason (77.6%) followed by “need more personal time” (76.3%), and “fear of compromising quality of care” (59.8%). Women were more concerned about fatigue than were men. Among residents who did not want work hour restrictions, “additional surgical experience” was the most commonly selected reason (69.0%). Conclusion: Residents in obstetrics and gynecology report working long hours, and experiencing periods of little sleep. Most want their work hours to be limited. Fatigue is a major concern among residents that want their hours limited. A sizable minority worries that such limits might also limit their experience. (Obstet Gynecol 2001;97:1015– 8.

From the American College of Obstetricians and Gynecologists, Washington, DC. This research is supported by grant MCJ 117016 from the United States Department of Health and Human Services Bureau of Maternal and Child Health, Bethesda, Maryland.

VOL. 97, NO. 6, JUNE 2001

© 2001 by The American College of Obstetricians and Gynecologists.)

Medical residency training has traditionally been a time of long work hours and little sleep. The long work hours serve to increase the continuity of care that patients receive, and to provide residents with the opportunity to witness the progression of disease and the effects of treatment. The periods of little sleep are unavoidable consequences of this training regimen. This tradition has come under criticism due to concerns over the effects of sleep deprivation on residents.1,2 Specific negative consequences on sleepdeprived medical residents have been evaluated using various outcome measures such as cognitive testing, resident alertness, and psychomotor performance.1,3–7 One common conclusion is that sleep deprivation can have detrimental effects on residents, and on their delivery of care. However, despite these findings, residency training continues to be a tradition through which long hours of work defines the educational experience, in part due to the perceived benefits of long work hours. This study was designed to investigate residents’ work and sleep schedules and their attitudes toward limiting their hours.

Materials and Methods An anonymous survey regarding resident work hours and call schedules was administered to the 4674 obstetric-gynecologic residents who took the 2000 Council on Resident Education in Obstetrics and Gynecology intraining examination within the United States of America. The survey contained questions on demography, current work hours, call schedules, postcall clinical responsibilities, and knowledge of and attitudes toward official limits on work hours for residents. Surveys were distributed to the residents before the test books were distributed and the proctors read a brief description of the survey and its purpose. Surveys were collected by the end of the day, and returned to ACOG in the same envelope as the test books. The data were analyzed using a personal computerbased software package (SPSS 10.0, SPSS Inc., Chicago, IL). Descriptive statistics were computed for the measures used in the analyses, which are reported as mean ⫾ SE. Two-tailed t tests were used to compare group means of continuous variables. Differences on categorical measures were assessed using ␹2. Group differences on ordinal measures were assessed using the Mann–Whitney U and Kruskal–Wallis tests. All analyses were tested for significance using alpha ⫽ .05.

0029-7844/01/$20.00 PII S0029-7844(01)01363-1

1015

Table 1. Respondents by Postgraduate Year and Gender

Table 3. Average Hours of Sleep on Call

Year of training

Males

Females

Total*

%

Year of training

⬍1 h

1–2 h

2–3 h

3– 4 h

⬎4 h

PGY 1 PGY 2 PGY 3 PGY 4 Total

314 339 383 384 1420

797 752 733 708 2990

1121 1121 1144 1124 4510

24.9 24.9 25.4 24.9 100.1

PGY 1 PGY 2 PGY 3 PGY 4 All residents

37.6 26.2 15.6 6.0 21.2

30.8 31.3 23.4 14.5 24.9

18.4 24.7 30.0 27.0 25.1

8.3 12.3 21.5 29.7 18.0

4.9 5.5 9.5 22.8 10.8

PGY ⫽ postgraduate year. * Male and female columns do not add to the total because 100 respondents declined to state their gender.

Results

PGY ⫽ postgraduate year. Values are percent of respondents in each category.

All differences significant at the P ⬍ .01 level.

sponding “yes” increased with year of training from 22.4% of first-year residents to 33.0% of fourth-year residents. The modal response for hours worked per week was 81–100 hours (42.1%), with three of four (75.5%) respondents reporting they worked between 61 and 100 hours each week. Three of four residents report getting one day off per week. After excluding those residents currently on night float, three of four (73.5%) residents reported they were on night call no more frequently than every fourth night; 24.6% reported they had night call every third night. Seven of ten (71.3%) residents reported sleeping less than 3 hours while on night call (Table 3). Eight of ten (82.4%) reported they had postcall clinical responsibilities. The year of residency had a significant effect on both on call hours and the amounts of sleep residents get during night call. In general the reported number of hours on call declined and the reported number of hours of sleep increased (Table 3) with year of residency. There were no differences between the genders concerning on-call hours during residency years 1, 2, and 3. Among fourth-year residents, however, men reported fewer hours on call and more hours of sleep (P ⫽ .049 and P ⫽ .023, respectively). After accounting for the effect of year of residency, residents older than 30 years worked more hours and got less sleep during night call. Three of four residents wanted a limit on their work hours (76.6% yes, 21.2% no, and 2.2% did not answer). There was a significant decline in the proportion of residents that wanted work hour restrictions with advancing year of residency, from 82.2% of first-year residents to 69.8% of fourth-year residents (P ⬍ .001). Residents who reported longer on-call hours or less sleep during night shifts were more likely to want a restriction on work hours (P ⬍ .001 for both). Women were significantly more likely than men to want to limit work hours (80.4% versus 68.7%, P ⬍ .001), and this result held when adjusted for age, year of residency, and residency setting. Residents in a community setting were significantly more likely to want to limit working

Residency Work Hours

Obstetrics & Gynecology

A total of 4583 surveys were returned from the 4674 residents taking the January 2000 in-training examination, a response rate of 98.1%. Seventy-three surveys were excluded from analyses because the year of training status was either missing or “other” (54 surveys) or because the respondents indicated that their residency was located in three or more states (19 surveys), leaving 4510 surveys to be analyzed (96.5% of the total possible). Respondents were equally distributed among the four year-of-training categories (Table 1). About twothirds of the residents were women (66.3%), 31.5% were men, and 2.2% declined to state their gender. Among respondents who reported their gender, the proportion of residents that were female steadily increased from 64.8% of fourth-year residents to 71.7% of first-year residents (Table 1). Residents older than 30 years of age, as a group, differed from those 30 and younger (Table 2). Specifically, they were less predominantly female, less predominantly found in a university setting and, among first- and second-year residents, more likely to get less than 1 hour of sleep while on night call. Most residents were not aware of any state regulations regarding resident work schedules. Residents from New York were a notable exception, with 98.9% of respondents from that state answering “yes.” Only 26.4% of residents responded that they were aware of Residency Review Committee requirements regarding resident workload. The percentage of residents reTable 2. Resident Characteristics by Age Grouping

Percent of residents Percent who are female Percent who are in a university setting Average number of residents in program Percent of first and second year residents who report getting less than 1 h of sleep while on night call

1016 Defoe et al

ⱕ30 y

⬎30 y

70.2 70.2 66.3 21.4 ⫾ 0.1 30.3

29.8 58.9 61.3 20.1 ⫾ 0.2 36.8

hours (79.0% versus 75.4% of residents in a university setting, P ⫽ .012). There was no age effect. Residents were given four reasons why they might want to limit work hours. They were able to select any or all of the choices. Among residents that wanted to limit their working hours, “fatigue” was the most commonly selected (77.6%), followed by “need more personal time” (76.3%) and “fear of compromising quality of care” (59.8%). A small number answered “need additional income” (7.3%). There were significant effects of year of residency, gender, age, and residency setting. Concern over fatigue declined with year of residency. Women were more likely to choose fatigue than were men and men were more likely to select “need additional income” than were women. Older residents were the most likely to select “fear of compromising quality of care.” Younger, female residents in a university setting were the most likely to select “need more personal time.” Residents were given three reasons why they might not want a limit on their work hours. Among those who did not want work hour restrictions, “additional surgical experience” was the most commonly selected (69.0%) followed by “opportunity to see rare cases” (46.5%) and “continuity with patients” (31.8%). There were no significant effects of demographic variables. About one in eight residents (12.7%) were currently “moonlighting” and 16.1% reported they had done so in the past. The proportion of residents who moonlight dramatically increased with year of residency, as might be expected. Virtually no first-year residents reported it (0.5%), compared with 24.3% of fourth-year residents. In residency years 1 through 3, men were more likely to moonlight than were women, but this difference between men and women disappeared among fourth-year residents. Residents who wanted to limit work hours were less likely to moonlight (10.5% versus 20.5%, P ⬍ .001).

Discussion The results of this survey support the contention that residents in obstetrics and gynecology work long hours and experience occasional periods of little sleep. This result has been documented previously in several other surveys of medical residents that found that residents in the surgical specialities, including obstetrics and gynecology, work longer hours and are more likely to go long periods without sleep than other residents.2,8 Sleep deprivation has been shown to be negatively correlated with satisfaction with the internship experience among first-year medical residents in the United States,8 and sleep loss has been suggested as an important predictor

VOL. 97, NO. 6, JUNE 2001

of stress and depression among general practitioners in England.9 A substantial majority of residents in this survey, more than 75%, were in favor of having restrictions on the number of hours they can work. Women were more in favor of work restrictions than were men, regardless of age, year of residency, or residency setting. However, even in the demographic subcategory of residents who were least likely to want work hour restrictions (ie, fourth-year male residents in a university setting) most (59.5%) favored restrictions. There was a sizable minority of residents, about one in five, that opposed limits on their work hours. The uniting perception among this group was that limiting hours might also limit their surgical experience. The fact that only residents from New York were aware of any state regulations regarding work hours is not surprising: only New York state has any restrictions. The fact that most residents were not aware that the Residency Review Committee stipulates that residents be on call no more frequently than every third night and have one off day per week is more troubling. Given the level of concern expressed over work hours by the residents in this survey, perhaps an educational awareness program needs to be incorporated into the fabric of residency training. The two most common reasons expressed for limiting work hours were “fatigue” (77.6%) and “need for more personal time” (76.3%). Concern over fatigue declined with year of residency. This finding is not surprising considering that the number of on-call hours declined and the self-reported amount of sleep during night call increased with year of residency. What is not as easily explainable is that women were more concerned about fatigue than were men, regardless of other demographic variables. Although studies of military aviators have demonstrated some small differences between men and women in performance after sleep deprivation,10 overall effects on performance and self-reported perceptions of the effects of sleep deprivation do not appear to differ between the sexes.10,11 Unfortunately we have no data on the personal lives of the residents, such as marital status and the existence of dependents, which might indicate whether these women residents in general have more “off-hour” responsibilities than do the men, which might explain their larger concern with fatigue. Most residents that favored work hour limits indicated that “fear of compromising quality of care rendered to patients” was a reason. A recent study of nonphysicians found that 24 hours of sustained wakefulness resulted in decrements of hand– eye coordination equivalent to those found in subjects with blood alcohol levels of .08%.12 With the current medical atmo-

Defoe et al


1017

sphere focusing on the reduction of adverse events resulting from medical errors,13 this is a timely period to invest in designing innovative pathways to restructure resident work hours and call schedules. One study found that modifying the traditional rotational overnight call schedule to one that was designed to reduce sleep deprivation, among other parameters, resulted in fewer medication errors by resident physicians.14 The concerns expressed by most residents indicate that reexamining the structure of the training experience of resident physicians might be warranted at this time. The proportion of women entering residency programs in obstetrics and gynecology, already a substantial majority, continues to rise. Among male residents, a higher proportion is older. Thus, the population that is being served by residency training programs differs in important ways from the not so distant past. Perhaps residency programs need to change as well. Adjusting call schedules is unlikely to be a “silver bullet” that improves the residency experience with no downside, however. Resident educators will need to be innovative to meet the concerns of their residents while preserving the inherent values of the training programs of the past. The use and continued development of modern educational techniques and technology, and innovative thinking regarding the more efficient use of all personnel, could play an equally important role in improving the residency experience.

References 1. Samkoff JS, Jacques CHM. A review of studies concerning effects of sleep deprivation and fatigue on residents’ performance. Acad Med 1991;66:687–93. 2. Daugherty SR, Baldwin DC. Sleep deprivation in senior medical students and first-year residents. Acad Med 1996;71:S93–5. 3. Robbins J, Gottlieb F. Sleep deprivation and cognitive testing in internal medicine house staff. West J Med 1990;152:82– 6. 4. Rubin R, Orris P, Lau SL, Hryhorczuk DO, Furner S, Letz R. Neurobehavioral effects of the on-call experience in housestaff physicians. J Occup Med 1991;33:13– 8.

1018 Defoe et al


5. Leung L, Becker CE. Sleep deprivation and house staff performance. Update 1984 –1991. J Ocup Med 1992;34:1153– 60. 6. Richardson GS, Wyatt JK, Sullivan JP, Orav EJ, Ward AE, Wolf MA. Objective assessment of sleep and alertness in medical house staff and the impact of protected time for sleep. Sleep 1996;19:718 – 26. 7. Nelson CS, Dell’Angela K, Jellish WS, Brown IE, Skaredoff M. Residents’ performance before and after night call as evaluated by an indicator of creative thought. J Am Osteopath Assoc 1995;95: 600 –3. 8. Daugherty SR, Baldwin DC, Rowley BD. Learning, satisfaction, and mistreatment during medical internship: A national survey of working conditions. JAMA 1998;279:1194 –9. 9. Firth-Cozens J, Greenhalgh J. Doctors’ perceptions of the links between stress and lowered clinical care. Soc Sci Med 1997;44: 1017–22. 10. Caldwell JA, LeDuc PA. Gender influences on performance, mood and recovery sleep in fatigued aviators. Ergonomics 1998;41:1757– 70. 11. Chelette TL, Albery WB, Esken RL, Tripp LD. Female exposure to high G: Performance of simulated flight after 24 hours of sleep deprivation. Aviat Space Environ Med 1998;69:862– 8. 12. Dawson D, Reid K. Fatigue, alcohol and performance impairment. Nature 1997;388:235. 13. Kohn LT, Corrigan JM, Donaldson MS, eds. To err is human: Building a safer health system. Washington DC: National Academy Press, 2000. 14. Gottlieb DJ, Parenti CM, Peterson CA, Lofgren RP. Effect of a change in house staff work schedule on resource utilization and patient care. Arch Intern Med 1991;151:2065–70.

Address reprint requests to:

Michael L. Power, PhD Department of Research American College of Obstetricians and Gynecologists 409 12th Street, SW Washington, DC 20024 E-mail: [email protected]

Received November 3, 2000. Received in revised form January 7, 2001. Accepted February 22, 2001. Copyright © 2001 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.

Obstetrics & Gynecology