Gabbe, Steven G. MD; Morgan, Maria A. PhD; Power, Michael L. PhD; Schulkin, Jay PhD; Williams, Sterling B. MD
Recent reviews have addressed the consequences of long work hours and fatigue on the health and performance of house officers. To reduce this stress, the Accreditation Council for Graduate Medical Education instituted substantial changes in scheduled work hours for residents, effective July 1, 2003.1,2 This policy includes a work week of no longer than 80 hours, one 24-hour day free every 7 days, shifts of no longer than 30 hours, and at least 10 hours off between duty periods. To assess the present status of resident duty hours in obstetrics and gynecology, a questionnaire was administered to residents throughout the country as part of the annual Council on Residency Education in Obstetrics and Gynecology examination in 2001. At the time of the study, women represented more than two thirds of all residents in obstetrics and gynecology. This analysis was undertaken to 1) assess the present status of work hours for residents in obstetrics and gynecology during pregnancy, 2) identify existing policies concerning work schedules during pregnancy, and 3) evaluate pregnancy outcome in female house officers and spouses or partners of male house officers.
MATERIALS AND METHODS
An anonymous, cross-sectional survey regarding resident work hours and call schedules was administered to the 4674 obstetrics and gynecology residents who took the 2001 Council on Residency Education in Obstetrics and Gynecology in-training examination in the United States. The survey contained questions on demography, knowledge of maternal and paternal leave policies, personal history regarding the first pregnancy outcome during the residency, and work hours and call schedules during pregnancy. Surveys were distributed to the residents before the test books, and the proctors read a brief description of the survey and its purpose. Surveys were collected by the end of the day, and returned in the same envelope as the test books.
The data were analyzed with a personal computer–based 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 ± andard error. Differences on categoric measures were assessed by χ2 analysis. All analyses were tested for significance (α = 05).
Surveys were returned by 4463 residents for a 95.5% response rate. However, 75 surveys did not indicate gender, 11 did not report residency year, and 20 indicated that the respondent was beyond the fourth year of residency training. After excluding these 106 questionnaires, 4357 remained for analysis, including 3034 women (69.6%) and 1323 men (30.4%). Residents were evenly distributed throughout the 4 years of training: first year 1100 (25.2%), second year 1089 (25.0%), third year 1067 (24.5%), and fourth year 1101 (25.3%) (Table 1). The largest proportion of women residents were in their first year of training (73.5%, 808 of 1100), whereas the largest number of male house officers were in the fourth year (34.2%, 376 of 1101). Of those residents recording an institutional affiliation (4157 of 4357, 95.4%), nearly two thirds were at university hospitals (65.9%).
Four hundred twenty-four women residents (14%) reported a pregnancy during training, whereas 353 male residents (26.7%) noted that their spouse or partner had had a pregnancy. Of female residents reporting a pregnancy, 94.8% had a spouse or partner who worked, whereas 68.5% of the spouses or partners of male residents who reported a pregnancy were employed (P < .001). Female residents were significantly more likely to have a spouse or partner who was a resident (P < .001) or a physician (P = .01) and significantly less likely to be a nurse (P < .001) (Table 2). Data on spouse or partner resident specialty or hours worked by spouses or partners in different professions were not collected. Half of the pregnant residents were between 26 and 29 years old at the time of gestation, and one third were in their early 30s. Nearly two thirds of the spouses or partners (64.5%) of male residents were younger than 30, compared with 58.2% of women residents (P = .008).
Ninety-one percent of the residents reported that there was a maternity leave policy at their institution, and 82% noted that the maternity leave was paid. For more than three quarters of the residents (76.5%), the maternity leave was reported to be 4–8 weeks. Residents reported that 60.1% of their programs have a paternity leave policy, and nearly half (48.8%) noted that this leave was paid.
Pregnancy was recorded most often in the fourth year of residency training for both women residents and spouses or partners of male residents (46.5%, 197 of 424, and 41.4%, 146 of 353, respectively). Relatively few pregnancies were reported in the first year of residency training for both women house officers and spouses or partners of male residents (Table 3).
Among residents who reported the number of hours worked per week during a pregnancy, women were significantly more likely than men with pregnant spouses or partners to work more than 80 hours per week (women: 72.1% [n = 380], 73.2% [n = 354], and 68.3% [n = 322] compared with men: 55.9% [n = 254], 57.1% [n = 238], and 57.3% [n = 232], for the first, second, and third trimesters, respectively [P < .01 for all trimesters]). However, when examined by postgraduate year, this result was only significant for third-year residents in the first trimester (P = .013) and for fourth-year residents throughout pregnancy (P < .01 for all trimesters). The lack of a statistical difference between women and men for the earlier postgraduate years might be due, in part, to fewer pregnancies reported in postgraduate years 1, 2, and 3 compared with year 4. Numerically, a greater proportion of pregnant women residents in all postgraduate years reported working 80 or more hours per week.
The modal work schedule for both male and female residents during a pregnancy was 81–100 hours. For men, the proportion that worked 81–100 hours per week varied little (42.1%, 39.4%, and 39.0%, for the first, second, and third trimesters, respectively). A higher proportion of women worked 81–100 hours per week during the first two trimesters (57.0%, 53.0%, 42.5%, for the first, second, and third trimesters, respectively). A consistently greater proportion of women residents reported that they worked more than 100 hours per week throughout pregnancy (women: 15.2%, 15.5%, and 16.4% compared with men: 13.8%, 14.7%, and 13.4%, for the first, second, and third trimesters, respectively).
Nearly 25% of women residents (24.7%) said that they were required to make up time they had missed at work after they completed their maternity leave. The time was made up most often when the resident returned from maternity leave or at the end of residency training. Nearly 95% of the respondents, both women and men, reported that they had to take over the work of a resident who was absent for maternity or paternity leave. More than three quarters of the women residents (76.5%) who reported a pregnancy outcome during training took no days off before delivery, whereas 10.3% missed 1–3 days of work before birth.
Pregnancy outcome was reported by 302 female residents (71.2%) and 274 male house officers (77.6%). Among residents who reported a pregnancy outcome, a singleton live birth occurred in 92.1% of women residents and 93.8% of spouses or partners of male residents (Table 4). A multiple gestation was reported in 11 women residents (3.6%) and in nine spouses or partners of male residents (3.3%). When compared with the spouse or partners of male residents, female residents were significantly more likely to report the following complications of pregnancy: premature labor (P = .03) but not preterm birth, preeclampsia (P = .01), and fetal growth restriction (FGR), defined as a birth weight below the tenth percentile for gestational age (P = .002). Stillborn fetuses occurred in only two female house officers and one spouse or partner of a male resident. When the hours worked by those female residents who had adverse pregnancy outcomes were examined, no statistical relationship was observed. For female residents reporting a singleton live birth, nearly 80% (221 of 278, 79.4%) had a spontaneous vaginal or assisted vaginal delivery, whereas 13.3% (37 of 278) underwent a primary cesarean delivery. For spouses or partners of male residents, these figures were 80.1% (206 of 257) and 10.1% (26 of 257), respectively.
A variety of opinions has been expressed regarding the impact of the changes in resident duty hours required by the Accreditation Council for Graduate Medical Education, effective July 1, 2003.3–6 Concern has been expressed that these changes will create a “shift work” approach to patient care among house officers and reduce clinical experience. How these changes might impact the work schedules of women residents during pregnancy or pregnancy outcome has not been considered. Our study, based on self-reported data from a questionnaire administered to residents in obstetrics and gynecology with the 2001 Council on Residency Education in Obstetrics and Gynecology examination, demonstrates that most residencies have both maternity and paternity leave policies, with maternity leave generally being 4–8 weeks long. This leave is usually paid. Nearly all of the residents reported that they assumed coverage of the work schedules of residents who were on maternity leave. Nearly one quarter of the women residents noted that they were required to make up time missed during maternity leave, usually when they returned from leave or at the end of their residency. More than three quarters of the women residents (76.5%) took no days off before their delivery. Pregnancy was most likely to occur during the fourth year of training for women residents. The spouses or partners of male residents were also more likely to be pregnant during their partners' final year of residency training.
Most women residents reported working more than 80 hours per week throughout pregnancy. A Council on Residency Education in Obstetrics and Gynecology survey conducted in 2000 also found that the modal work week for all residents was 81–100 hours per week, and that 75.5% worked between 61 and 100 hours per week.7 The proportion of residents reporting 81–100 hours per week in that survey (42.5%) was virtually identical to the proportions in this survey of male residents whose spouse or partner was pregnant and pregnant women residents in their third trimester, but less than the proportions of pregnant women residents in their first and second trimesters. Thus, it would seem that pregnant women residents, but not male residents with pregnant spouses or partners, are increasing their work hours during a pregnancy, at least during the first two trimesters.
Women residents have several work-related concerns regarding pregnancy.8 Although their institution is likely to grant them maternity leave, that time is generally required to be made up. Also, the women are aware that their peers are going to have to shoulder their work load during their leave. The data from this study imply that many women adopt a strategy of working long hours up to delivery. It is a reasonable hypothesis that this behavior is an adaptation to those two concerns. The new policy of restricting work hours will complicate this potential strategy for dealing with the time constraints of pregnancy during residency. Women will not be able to increase their work hours during pregnancy to accommodate the need to make up for maternity leave, yet how are other residents going to cover for residents on maternity or paternity leave if they are restricted to 80 hours per week?
Our data suggest that, for most female house officers, the long working hours associated with residency training in obstetrics and gynecology do not have an adverse affect on pregnancy outcome. More than 90% of women residents who provided information on pregnancy outcome indicated they had had a singleton live birth, and only two stillborn fetuses were described in this population. However, although the frequencies were relatively low, the rates of preterm labor, preeclampsia, and FGR were significantly higher for women residents when compared with the spouses or partners of male house officers. In 1990, Klebanoff et al9 performed a national questionnaire-based survey and reported pregnancy outcomes in 1293 women residents and 1494 wives of male residents. That study included 139 pregnancies in female residents in obstetrics and gynecology. The authors concluded that working long hours had little effect on pregnancy outcome in house officers. However, women residents were more likely to report preterm labor and preeclampsia, as observed in our study, but experienced no increase in preterm birth or FGR. We found no statistical association between hours worked and adverse pregnancy outcomes, most likely because most women worked long hours throughout pregnancy and because some complications resulted in the female house officer working fewer hours. It is of concern that pregnancy outcomes were not reported by 28.8% of female residents and 22.4% of male house officers. Given the extremely low rates for preterm labor, preeclampsia, and FGR in the spouses or partners of male residents, it is likely that significant underreporting complicates the interpretation of these data.
In summary, our data demonstrate that most women residents are working more than 80 hours per week during pregnancy and take little time off before their delivery. Most have a paid maternity leave of 4–8 weeks, and some are required to make up the time they have missed while on maternity leave. Nearly all residents reported they assumed the call missed by residents on pregnancy leave. How such coverage will impact the limitation on resident work hours set by the new Accreditation Council for Graduate Medical Education policy remains to be seen. In the future, it will be important to monitor the impact of the Accreditation Council for Graduate Medical Education requirements on resident duty hours, maternity and paternity leave, and pregnancy outcome.