A 199-item survey that queried all aspects of pregnancy, childbearing, and maternity leave as well as basic demographic information including training, residency, fellowship, work environment, and satisfaction was created. The survey was entered into an online survey engine, Zoomerang (http://www.zoomerang.com). A recruitment e-mail with an online link inviting participation was sent to the Ruth Jackson Orthopaedic Society (RJOS), female surgeon interest groups in other specialties, and the Women in Surgery Committee of the American College of Surgeons. Nine specialties were surveyed: general surgery, gynecology, neurosurgery, ophthalmology, orthopaedics, otolaryngology, plastic surgery, podiatry, and urology. Since comprehensive information on all board-certified female surgeons could not be obtained, we relied on these interest groups for distribution and recruitment. The survey site was open and recruitment was solicited for one calendar year, April 2009 to April 2010. Respondents included residents, fellows, and attending surgeons. Data regarding the duration of breastfeeding and maternity leave for first-born children, associated institutional policies, and respondent satisfaction were abstracted. Respondent options for the satisfaction questions included extremely satisfied, somewhat satisfied, neutral, somewhat dissatisfied, and extremely dissatisfied. For maternity leave, respondents were asked “Were you satisfied with the amount of time that you took off from work after each delivery?” Adoptions were excluded from the analysis. For breastfeeding, women were asked “Were you satisfied with the amount of time that you breastfed?” Given the personal nature of the survey, no respondent identifier or tracking mechanism was used, but members were given an option to enter an e-mail address.
Statistical Methods
Data are reported on the basis of the number of pregnancies carried through to live birth. Data regarding spontaneous and intentional abortions were collected, but these were not included as pregnancy complications in the analyses presented. Continuous parameters were analyzed with use of the Student t test with unequal variances, and analysis of categorical data was performed with use of the chi-square test. The normality of continuous variables was examined with use of histograms and box plots, and the Mann-Whitney U test was applied in instances in which normality assumptions were violated. Data comparisons were made both within our dataset of female surgeons as well as with the most recent national data from the American Pregnancy Association (APA). When possible, the APA data were manipulated to create an estimated comparison cohort matched according to demographic data. The APA data were based on a one-year period, which presumably reflects only one pregnancy per woman, and it was much more contemporary than that in our study, which included pregnancies over a forty-year period. We attempted to control for this by comparing the APA data with the data regarding only the first reported pregnancy in our respondents. No correction could be made for changes that have occurred in obstetric care over the years. Odds ratios for predictors of pregnancy and delivery complications and for satisfaction rates were determined with use of logistic regression analysis.
Source of Funding
No external funding source was utilized for this study.
A total of 1021 surveys were completed. Of these, 223 were from women trained in orthopaedics, representing 21.8% of all respondents, 48.6% of the women invited by the RJOS, and approximately 30% of all female orthopaedic surgeons in the U.S. One hundred and twenty-eight of the 223 orthopaedic surgeons reported a total of 263 individual pregnancies. The remaining 798 women reported training in other surgical subspecialties including general surgery, gynecology, neurosurgery, ophthalmology, otolaryngology, plastic surgery, podiatry, and urology. Five hundred and seventy-six of these women reported a total of 1180 pregnancies.
Individual characteristics of the orthopaedic surgeons were compared with those of the respondents in the other surgical specialties (Table I). The race distribution was similar between the groups, with the majority being White. The percentage of women who pursued fellowship training was higher among orthopaedic surgeons than among surgeons in nonorthopaedic specialties (77% [172 of 223] compared with 53% [426 of 798], p < 0.001). A greater percentage of orthopaedic surgeons reported working more than sixty hours per week (51% [111 of 219] compared with 32% [250 of 781], p < 0.001). Surgeons in orthopaedics also reported spending significantly more time in the operating room per week (mean, 17.3 compared with 13.5 hours; p = 0.0001).
Pregnancy-related characteristics are outlined in Table II. On average, the orthopaedic surgeons who had children had a total of 2.1 children, with most women having one to three. Seventeen respondents reported early-term miscarriages, which were not analyzed as pregnancy complications. Six women reported having abortions, three of which were specifically related to congenital abnormalities noted on screening examinations. The mean age of female orthopaedic surgeons at the time of delivery was considerably older than U.S. norms reported by the American Pregnancy Association13: 33.1 compared with 24.6 years for the first child, 35.1 compared with 27.9 years for the second child, and 36.7 compared with 29.2 years for the third child. Orthopaedic surgeons had their first child prior to completion of their training more commonly than women in other specialties (56.2% [seventy-two of 128] compared with 48.6% [280 of 576], p < 0.05). The percentage of all respondents (including those who did not have children) who had at least one child during their residency training was 25.1% (fifty-six of 223) in orthopaedics and 26.9% (215 of 798) in other surgical specialties. The number of births reported by orthopaedic surgeons as well as by the total cohort of all surgical subspecialists increased progressively by decade (Fig. 1).
One hundred and forty-eight of the 223 orthopaedic respondents attempted at least one pregnancy, and forty-five (30%) of these reported some problem with infertility. Of the orthopaedic surgeons who reported some problem with infertility, 78% (thirty-five of forty-five) sought evaluation from a medical specialist and 60% (twenty-seven) implemented fertility treatment. Of those who pursued treatment, 56% (fifteen of twenty-seven) went on to have a successful pregnancy.
The overall reported complication rate among orthopaedic surgeons was 31.2% (eighty-two of 263 pregnancies) (Table III). This was comparable with the overall complication rate of 35.3% (417 of 1180) reported by women in other specialties. The complication rate in the orthopaedic surgeon cohort was 35.2% (forty-five of 128) for first pregnancies and decreased with each succeeding pregnancy (e.g., 30.7% [twenty-seven of eighty-eight] for second pregnancies). Eleven of seventeen women with multiple complications during their first pregnancy reported a second pregnancy, and eight of the eleven had a complication during their second pregnancy as well. The most commonly reported complication was preterm labor with either preterm or term delivery. The overall complication risk in our orthopaedic surgeon cohort was not significantly associated with maternal age, stage of career, or time spent in the operating room per week. However, we did observe an increased risk of preterm labor and delivery in respondents who reported working more than sixty hours per week (odds ratio [OR], 4.95 compared with less than sixty hours a week; 95% confidence interval [CI], 1.4 to 36.6).
Delivery was induced prior to thirty-nine weeks of gestation in twenty-one of the 128 first pregnancies in the orthopaedic surgeons. A cesarean section was requested in 6% (eight) of these first births, which was not significantly different from the elective cesarean rate of 7.5% (forty-three of 576) reported by the combined cohort of other surgical subspecialists (p = 0.55). A cesarean section was considered elective if it was not specifically recommended by the treating obstetrician. Cesarean sections were elected for a variety of reported reasons: worries about the pelvic floor, worries about incontinence, in order to best time the delivery, worries about labor and delivery, desire for a repeat cesarean, and others. The nonelective cesarean rate was lower in female orthopaedic surgeons compared with the combined cohort of other surgical subspecialists (21.7% [fifty-seven of 263] compared with 37.7% [445 of 1180], p = 0.004). The indications for cesarean sections have changed over time, and our survey did not control for these changes. Activity restriction was imposed on orthopaedic surgeons during 23.2% (sixty-one) of the 263 reported pregnancies, and 10% (twenty-six) of the 263 pregnancies required some duration of bed rest (mean, 5.4 weeks; median, four weeks). Four percent (five) of 128 female orthopaedic surgeons required hospitalization prior to delivery of their first child because of complications, but no hospitalizations were required for subsequent births. The rate of congenital abnormalities reported for all births was 6.8% (eighteen of 263) for orthopaedic surgeons and 8.9% (105 of 1180) for the cohort of other specialists (p = 0.53).
The duration of maternity leave taken for three-quarters of the births (195 of 263) was less than eight weeks (mean, 7.8 weeks). They took less maternity leave during training than during clinical practice, with the median duration of maternity leave during residency and fellowship being four weeks compared with seven weeks during clinical practice. They took more than twelve weeks of maternity leave after only 5.3% (fourteen) of the births and three weeks or less after 11.8% (thirty-one). Dissatisfaction with the duration of leave was reported after 46% (102) of the births. We found no association between dissatisfaction with the duration of maternity leave and stage of training.
Breastfeeding was initiated and continued for at least one month after 90% (237) of the 263 births reported by orthopaedic surgeons. Breastfeeding was continued for at least four months after 71% (187) of the births. Most of the women who discontinued breastfeeding prior to six months reported problems with accessibility and/or time for breastfeeding after their return to work. The mean duration of breastfeeding was significantly shorter during training (4.7 months) than during clinical practice (8.3 months, p = 0.03). The rate of dissatisfaction with the duration of breastfeeding was also higher during training than during clinical practice (p = 0.04). There was a moderate correlation between the duration of maternity leave and the duration of breastfeeding (r = 0.45; 95% CI, 0.35 to 0.54). We found no association between the duration of breastfeeding and the decade of birth, despite the increased awareness and availability of nursing facilities in workplaces currently.
Satisfaction rates for first births were comparable in the two groups, with 85% (109) of the 128 orthopaedic surgeons and 83% (478) of the 576 women in the other surgical specialties reporting satisfaction with their overall pregnancy experience. Twelve percent (sixty-nine) of the 576 other surgical specialists surveyed and 7% (nine) of the 128 orthopaedic surgeons reported some element of dissatisfaction with their experience (p = 0.25). There was no association between satisfaction and stage of career, having a pregnancy complication, or the number of work hours. Regarding the timing of childbirth, 60% (134) of 222 female orthopaedic surgeons preferred having children at the time that they actually did, 37% (eighty-three) would have preferred an earlier time, and 2% (five) would have preferred a later time. Fifty-four percent (121) would have preferred the number of children that they actually had, 39% (eighty-seven) would have preferred more, and 7% (fifteen) would have preferred none.
Ninety-five percent (212) of all 223 orthopaedic surgeons reported satisfaction with their work, and 89% (199) were happy with their career choice. Seventy-one percent (158) were satisfied with the balance of their time management between family and career. The majority of all orthopaedic surgeons (with and without children) felt that their career affected their family life a moderate amount. The majority of women without children reported that family had no effect on their career, whereas women with children reported that their family affected their career a moderate amount. Fifteen percent (nineteen) of the 128 women who had children felt that having children had markedly slowed their career.
Many prior studies have been performed to assess the effect of working conditions on the risk of pregnancy complications. Overall, most studies support the absence of a difference in the risk of complications (including preterm birth, low birth weight, or maternal health issues) between employed and unemployed women14-16. Multiple studies have, however, raised concerns regarding adverse pregnancy outcomes in particularly demanding work environments. A case-controlled study involving seventeen European countries showed a moderately elevated risk of preterm birth in women working more than forty-two hours per week or standing more than six hours daily14. Physically demanding work involving prolonged standing and shift or night work has also been reported to be significantly associated with preterm birth and preeclampsia. High cumulative work fatigue scores have been specifically linked to the risk of preterm birth17. A demanding posture for at least three hours daily, whole-body vibrations, and high job-related strain combined with low or moderate social support have also been linked with preterm delivery. The combination of the last two conditions, high job-related strain and low to moderate social support, has been associated with severe preterm delivery18.
The present study further confirms the high rate of pregnancy complications among female surgeons. The complication rate of 31.2% (eighty-two) in the 263 pregnancies in the orthopaedic surgeon cohort was significantly higher than the 14.5% rate reported in the general U.S. population13 but was comparable with the rate of 35.3% (417) reported in the 1180 pregnancies in the combined group of other surgical subspecialists. This elevated rate of pregnancy complications is of particular concern given the higher education levels and socioeconomic status of surgeons; both of these characteristics are associated with decreased pregnancy complication rates in the general population. A previous retrospective chart review comparing physicians with nonphysicians of similar socioeconomic status who delivered to the same obstetrical team revealed that the relative risk of an adverse pregnancy outcome among the physicians was 1.8619. The authors reported significantly increased risks of both preterm labor and preterm delivery, and they recommended that physicians should be treated as a high-risk obstetrical group, a recommendation that is supported by our study.
Why surgeons, including orthopaedists, have higher rates of pregnancy complications is less clear and cannot be confirmed by the present study. However, many hypotheses can be made. Multiple studies have shown an adverse effect of prolonged standing on pregnancy outcomes18,20-22. Prolonged, uninterrupted periods of standing and physically demanding work, involving lifting and repetitive strenuous activity, during surgical procedures are inherent to orthopaedics as well as to other surgical specialties. Despite this, we did not find any association between the amount of time spent in the operating room and the rate of pregnancy complications in either group, even though orthopaedic surgery is a particularly physically demanding specialty. Most surgeons work long hours (defined as more than forty hours per week or more than eight hours per day), and this, along with strenuous physical activity, has been linked to adverse pregnancy outcomes23-26. A recent quality-of-life study of orthopaedic surgeons during residency and academic practice revealed that residents worked a mean of 70.2 hours per week and attending surgeons worked a mean of 62.8 hours per week27. Many of our respondents reported working a similar amount of sixty-one to eighty hours per week. We did find an increased risk of preterm delivery in female orthopaedic surgeons who reported working more than sixty hours per week (OR, 4.95; 95% CI, 1.4 to 36.6). There was no significant association between work hours and preterm delivery in groups working less than sixty hours per week.
Although it is tempting to presume that the most likely underlying factor for the increased rate of pregnancy complications among the female orthopaedic surgeons in our study was advanced maternal age, the majority of our respondents reported births before the age of forty. The most commonly seen complications in our study included preterm labor with term delivery and with preterm delivery; however, in the FASTER trial (a large multicenter study of 36,056 single births), these two complications were only noted to be significantly increased in women over forty years of age2. Therefore, the increased overall complication rate in our study is likely multifactorial, with advanced maternal age, long work hours, physical activity, and high job stress all contributing. Controlling the data from the general U.S. population13 according to race (White), age (thirty to forty years), education level (graduate degree), and health status (nonsmokers) to more closely resemble our surveyed population, we found the relative risk of preterm delivery in orthopaedic surgeons to be 2.5 (95% CI, 1.3 to 4.6). In addition, we found the relative risk of gestational diabetes (a complication not known to be linked to work environment or physical demands) to be 0.5 (95% CI, 0.13 to 2.0) compared with the general population.
The rate of reported infertility problems in the orthopaedic surgeons in our study was high (30.4%, forty-five of 148 respondents who attempted pregnancy); however, this rate was comparable with that for an age-matched control population, suggesting that advanced maternal age was the predominant cause. Nevertheless, 8% (twenty-one) of the 263 reported pregnancies among orthopaedic surgeons were conceived by means of assisted reproductive technology (ART); this rate is higher than the national norm of 1%28. The success rates for ART in our study (56% [fifteen of twenty-seven] among the orthopaedic surgeons and 65% [109 of 167] among the other surgical specialists) were higher than those in the general population (approximately 30%29). This difference could have arisen from a possible tendency for female surgeons to pursue and/or be evaluated for treatment earlier in their attempts to conceive (rather than waiting until after the age of thirty-five), female surgeons’ medical knowledge, the theoretically greater ability of female surgeons to access the health care system through personal contacts, statistical error resulting from our small sampling size, and/or bias of the survey respondents (with respondents who had infertility problems possibly being more likely to complete the survey than those without such problems).
Thirty-nine percent (fifty) of the 128 orthopaedic surgeons who reported a pregnancy had their first child during residency, and 56% (seventy-two) of the 128 first children were born to women prior to their completion of training. A previous survey of obstetrics and gynecology residents taking the 2001 in-training examination revealed a significantly greater rate of premature labor, preeclampsia, and fetal growth restriction in pregnancies reported by the female residents compared with the spouses or partners of the male residents30. Pregnant residents worked more hours than their nonpregnant and/or male counterparts during their first and second trimesters in preparation for taking maternity leave. Seventy-six percent of female residents worked until the day of delivery, and 10.3% missed only one to three days before delivery. Not surprisingly, 95% of all residents in obstetrics and gynecology programs across the country who were surveyed after implementation of the new duty hour restrictions reported doing increased work to cover the call responsibilities of residents who were away on maternity or paternity leave. A survey of fifty directors of plastic surgery residency programs revealed that 36% actively discouraged pregnancy during training9. Two primary reasons were given: a pregnant resident imposes hardships on other residents and interferes with the smooth functioning of the training program, and time away because of pregnancy is detrimental to the resident’s own surgical experience and training. To our knowledge, there have been no studies indicating that orthopaedic training programs or department chairs actively discourage pregnancy among their female residents.
The rate of congenital anomalies in our study (6.8% [eighteen of 263] for the offspring of the orthopaedic surgeons and 8.9% [105 of 1180] for the offspring of the other specialists) was higher than that in the general U.S. population (3.8%13). The causes of this difference are unknown, with advanced maternal age being a probable substantial contributor; however, given the rarity of congenital anomalies, our sample size was not sufficiently powered to analyze the trend further. In addition, definitive conclusions cannot be made because estimates of associated exposure factors are based on memory rather than proven exposures. Although we did not specifically ask about exposure to ionizing radiation, this is a concern for many surgical specialties. Zadeh and Briggs investigated the risk of ionizing radiation on the offspring of orthopaedic surgeons and found no significant difference in the rate of congenital anomalies compared with a control group of offspring of obstetricians and gynecologists31. We also did not evaluate exposure to methylmethacrylate. However, a prior study in mice indicated no evidence of teratologic effect or fetal toxicity with inhalation of higher doses of methylmethacrylate fumes than those typically present in the operating room during total hip arthroplasty32.
Overall, 85% (189) of 223 female orthopaedic surgeons reported satisfaction with their pregnancy experience. They were, however, more dissatisfied with the duration of maternity leave and the duration of the time during which they breastfed their infants. The reported rate of dissatisfaction with the time they took for maternity leave was 46% (102 of 223), which was comparable with the rate in the rest of the study cohort as well as that previously reported by urologic surgeons33. In the prior survey of urologic surgeons, 70% of women reported taking less than nine weeks of leave. A similar rate was reported for the 263 births among the orthopaedic surgeons in the present study, with the mothers having taken eight weeks or less after 74% (195) of the births. Prior studies have shown that taking less than nine weeks of maternity leave compared with more than twelve weeks has a negative impact, including increased rates of depression and anxiety, loss of general positive affect, and decreased life satisfaction34.
The rates at which female orthopaedic surgeons in our study both initiated and continued breastfeeding were higher than those previously reported among resident physicians35. Breastfeeding was initiated and continued for at least one month after 90% (237) of the 263 births among the orthopaedic surgeons, compared with an 80% initiation rate noted in the previous study of resident physicians. Breastfeeding was continued for at least four months after 71% (187) of the 263 births and for more than six months after 42% (111). The medical recommendation is for infants to be exclusively breastfed for four to six months unless major complications arise. It is encouraging that although the reported maternity leave duration was rarely longer than twelve weeks, nearly one-half of the orthopaedic surgeons were able to continue nursing after returning to work until their infant was six months of age.
As with any survey-based study, our study had limitations. Despite multiple attempts, we were unable to obtain a comprehensive list of all practicing female surgeons for recruitment. We therefore relied on individual interest groups for dissemination of the survey, but we were unable to solicit responses from all practicing female surgeons. Our study was also limited by historical recall bias. We collected data on pregnancies reported over a span of forty years, and inability to recall all details is a limitation. Our overall response rate was low and our data may also have been skewed by selection bias, with women without children or women who had uncomplicated pregnancies being less likely to respond. Our comparison with the average U.S. population was also limited by the fact that our study included data on pregnancies that had occurred over a broad time span but the comparison data were from a single recent year. To control for this, we compared data from only the first reported pregnancy. Nevertheless, we recognize that prenatal and obstetric care has advanced over time and that a pregnancy from forty years ago may not be equivalent to a pregnancy from three years ago. However, some context is necessary in order to interpret the presented results, and we believe that this was the most appropriate way to do so despite these limitations.
In summary, female surgeons, including orthopaedists, had an increased risk of pregnancy complications compared with the general U.S. population. The most commonly increased risk among orthopaedic surgeons was preterm delivery. Working more than sixty hours per week during pregnancy was associated with an increased rate of preterm labor and delivery. Although this study cannot determine exactly how surgeons should change their training and/or practice environments to reduce their risks, it is the first step toward bringing awareness of the issues. Restricted residency hours likely already pose challenges to training programs and to the residents themselves. Restricting hours even further during pregnancy, or at least during part of pregnancy, could have a substantial impact on training programs and on the training experience of the pregnant resident. Indeed, restricted hours during pregnancy could require that the residents extend their training by several weeks or limit their maternity leave.
These are complicated issues that cannot be solved easily. At the very least, medical schools, advisors, and program directors should share this information with their female mentees and allow them to make their own informed decisions. We hope that department chairs, program directors, and other medical leaders will discuss these sensitive issues and be amenable to change as necessary. Ideally, prospective studies will be conducted that can better characterize the issues for the health of the surgeon, the health of the fetus, and the overall health of the orthopaedic training program and of orthopaedic practice.
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. None of the authors, or their institution(s), have had any financial relationship, in the thirty-six months prior to submission of this work, with any entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, no author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.