Fertility and Obstetric Outcomes Among Procedural and Nonprocedural Physicians: A Call for Improved Fertility Education for Medical Students and Trainees

Abstract

Background:

Most physicians are in training during the time of known fertility decline, and trainees postpone parenthood for a variety of reasons. The aim of this study is to describe differences between the experiences of proceduralists versus nonproceduralists.

Methods:

An 11-item anonymous survey addressing experiences with pregnancy, fertility, and assisted reproductive technology (ART) was distributed to physicians (n = 2098) at 2 academic medical centers.

Results:

There was a 20% (n = 423) response rate. The median age was 31 years. Of the respondents, 293 (69%) were female. Two hundred seventy-six (66%) were trainees and 154 (36%) were proceduralists. Of 293, 137 (46.8%) female respondents reported having ever been pregnant. Of those, 55 (40%) reported pregnancy or obstetric complications. The overall birth rate was 2.57 (SD 0.95) live births per woman. Proceduralists had a lower number of live births (mean [SD], 2.3 [0.93] versus 2.7 [0.94]; P < 0.014). Of all respondents, 302 (72%) of 423 reported no prior education about fertility or ART. Proceduralists were significantly more likely to report never having received information on fertility and ART during residency (123 of 154 [80%] versus 179 of 269 [67%]; P = 0.007).

Conclusions:

We found a high prevalence of pregnancy and obstetric complications as well as pregnancy loss, with significantly fewer live births among proceduralists versus nonproceduralists. Medical school and residency curricula should include relevant fertility data and education to allow trainees to make informed decisions regarding family planning and fertility preservation.

Takeaways.

Question: Are physicians in procedural specialties at increased risk of infertility and poor pregnancy outcomes?

Findings: In this survey of 423 physicians at 2 academic medical centers, physicians in procedural specialties experienced fewer total pregnancies and live births than those in nonprocedural specialties. In addition, proceduralists were less likely to have received any information on fertility, cryopreservation, or assisted reproductive technologies during training.

Meaning: Female proceduralists may be at increased risk for infertility. Formalized educational curricula and institutional support may help support physician family planning goals.

INTRODUCTION

It is well known that fertility declines with age, with female fertility sharply declining from a woman’s late 20s into her 30s.^1–6 Most physicians complete residency training in their early to mid 30s, with longer programs graduating trainees in their late 30s and sometimes into their 40s.^7,8 Most female physicians are in residency training during the time of known fertility decline, particularly as more candidates are adding research or gap years. When in residency, trainees purposefully postpone parenthood due to the demands of clinical training.^9–12

A study from 2016 surveyed female physicians on fertility and childbearing history and reported high rates of infertility and regret regarding family planning and career decisions.¹³ Female surgeons are known to have twice the amount of miscarriages compared to the general population, and significantly more major obstetric complications, emergency cesarean sections, and postpartum depression compared to the female partners of male surgeons.¹⁴ Studies show that fertility awareness is low among residents across specialties,^15,16 and concerns surrounding fertility may be significant enough to alter career trajectory,¹⁷ with 1 survey of 347 female surgical trainees reporting that 39% strongly considered leaving surgical training specifically due to the concern of balancing motherhood and pregnancy with residency.¹⁷

As women represent a growing proportion of the physician workforce—particularly among surgical and procedural specialties—conducting further research in this area is imperative.¹⁸ To date, no studies have surveyed fertility and obstetric data and perceptions among all trainees and attendings within a single hospital or across multiple institutions and compared the experiences of proceduralists and nonproceduralists. Our study aims to address this gap.

METHODS

This study and associated materials were reviewed and approved by the institutional review boards of the 2 participating academic medical centers, one of which is a publicly funded county hospital (RUMC: 23082501-IRB01-AM02 and CCH: 23-061X-Stroger).

A short, anonymous survey addressing experiences with pregnancy, obstetric outcomes, and fertility and assisted reproductive technology (ART) was developed with a content expert experienced in survey design and was pilot-tested. Our survey was based on similar surveys described in the literature^14,19; surveys were publicly available or provided by the original paper authors. (See appendix, Supplemental Digital Content 1, which displays the survey, https://links.lww.com/PRSGO/F345.)

E-mail invitations containing the survey link via the RedCap platform (Research Electronic Data Capture Consortium, Nashville, TN) were advertised via e-mail to be distributed to all trainees and attendings across both institutions over a 1- to 2-month period between January 2024 and February 2024 with a total of 4 reminder e-mails sent at 1-week intervals. Several departments, including internal medicine, orthopedic surgery, radiology, and anesthesia, declined to formally send the survey to their faculty and/or their trainees. However, orthopedic surgery and anesthesia residents received the survey through resident-to-resident dissemination. Reasons for nonparticipation by certain departments were not systematically collected. In some cases, survey access may have occurred through informal resident-to-resident distribution. Participants were offered a modest financial incentive (entry into a raffle to win 1 of 10 $25 gift cards) to encourage survey completion.

Data were analyzed using descriptive statistical methods and univariate analyses using chi-square and Mann-Whitney U tests with R (The R Foundation for Statistical Computing, Vienna, Austria). Data were further stratified by “procedural” versus “nonprocedural” specialties. The procedural specialties were those listed by the American College of Surgeons²⁰ as “surgical specialties,” as well as interventional radiology and anesthesiology due to time spent in the operating room or procedure suites and similar physical demand. We acknowledge that this classification may not fully capture all procedural fields (eg, interventional cardiology and interventional gastroenterology).

This study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. (See appendix, Supplemental Digital Content 2, which displays strengthening the Reporting of Observational Studies in Epidemiology guidelines checklist, https://links.lww.com/PRSGO/F346.)

RESULTS

Participant Characteristics

All faculty, resident, and fellow physicians at both included medical centers were invited to participate in the study (n = 2098). As previously noted, a minority of divisions did not disseminate the survey, and some physicians received the survey informally. Thus, the response rate was calculated using the total number of eligible physicians across both institutions as the denominator to account for informal sharing of the survey. There were 423 responses, for a response rate of 20.2%, with 154 (36%) respondents belonging to procedural specialties. The median age of respondents was 31 years (interquartile range [IQR]: 29–39 y)—likely reflecting the composition of our respondent pool, which included a substantial proportion of trainees (66%). Two hundred ninety-three (69%) were female and 129 (31%) were male. Participant characteristics are displayed in Table 1.

Table 1.

Participant Characteristics

	Overall (N = 423)	Nonprocedural (n = 269)	Procedural (n = 154)	P
Age, median (IQR)	31 (29–39)	32 (29–40)	31 (29–35)	0.055
Sex
Male	129 (31%)	69 (26%)	60 (39%)	0.005
Female	293 (69%)	200 (74%)	93 (61%)
Race				0.35
White	248 (59%)	147 (55%)	101 (66%)
Black	31 (7.3%)	21 (7.8%)	10 (6.5%)
Asian	90 (21%)	63 (23%)	27 (18%)
Native Hawaiian/Pacific Islander	36 (8.5%)	26 (9.7%)	10 (6.5%)
Other	18 (4.2%)	12 (4.5%)	6 (3.9%)
Level of training				0.043
Resident	264 (63%)	158 (60%)	106 (69%)
Fellow	12 (3%)	9 (3%)	3 (2%)
Faculty	142 (34%)	98 (37%)	44 (29%)
Relationship status				0.90
Single	100 (24%)	60 (23%)	40 (27%)
Married	248 (59%)	159 (59%)	89 (58%)
Other partnership	74 (18%)	50 (18%)	24 (15%)

P < 0.05 (bolded) indicate statistical significance.

Female Respondents Who Have Been Pregnant: Overall

Specific characteristics and outcomes for those who have been pregnant are displayed in Table 2. Of 293, 137 (46.8%) female respondents reported having ever been pregnant. The median maternal age at birth of the first child was 32 years (IQR: 29–34 y). The most common time to have a child was during residency or fellowship (64 of 137 [44.6%]), followed by assistant professorship (36 of 137 [26%]), and medical school (22 of 137 [17%]). At survey completion, 18 (13%) were pregnant. One hundred five respondents (80%) reported working over 40 hours/wk during their pregnancy.

Table 2.

Characteristics of Those Who Have Been Pregnant

	Overall (N = 137)	Nonprocedural (n = 94)	Procedural (n = 43)	P
Age at first birth, median (IQR)	32.00 (29–34)	31.00 (29–33)	32.00 (30–35)	0.082
Career stage at first child				0.35
Before medical school	3 (2.3%)	2 (2.2%)	1 (2.4%)
During medical school	22 (17%)	19 (21%)	3 (7.1%)
PGY 1	3 (2.3%)	3 (3.4%)	0 (0%)
PGY 2	17 (13%)	11 (12%)	6 (14%)
PGY 3	12 (9.2%)	9 (10%)	3 (7.1%)
PGY 4	14 (11%)	6 (6.7%)	8 (19%)
PGY 5	5 (3.8%)	3 (3.4%)	2 (4.8%)
PGY 6	0 (0%)	0 (0%)	0 (0%)
PGY 7	0 (0%)	0 (0%)	0 (0%)
Research resident	3 (2.3%)	2 (2.2%)	1 (2.4%)
Fellowship	7 (5.3%)	5 (5.6%)	2 (4.8%)
Faculty	45 (34.45)	29(33%)	16(38%)
Hours worked during pregnancy				0.023
40 or fewer	26 (20%)	23 (26%)	3 (7.1%)
Greater than 40	105 (80%)	66 (74%)	39 (93%)
Hours worked on feet during pregnancy				0.092
40 and below	82 (62%)	61 (67%)	21 (50%)
Above 40	51 (38%)	30 (33%)	21 (50%)
Time off work during pregnancy				0.78
No	36 (69%)	24 (67%)	12 (75%)
Yes	16 (31%)	12 (33%)	4 (25%)

P < 0.05 (bolded) indicate statistical significance.

The overall birth rate was 2.57 (SD 0.95) live births per woman with the average number of total pregnancies being 3.6 (SD 1.0) pregnancies per woman. Overall, 19 (14%) respondents reported that none of their pregnancies resulted in a live birth. Thirty-two (24%) respondents reported having 1 or more miscarriages (defined as pregnancy loss between 5 and 20 wk gestation), and 55 (40%) reported having complications during pregnancy. The most common obstetric complication (Table 3) was preeclampsia (8.0%), followed by gestational diabetes (6.6%). The median age among those who had pregnancy complications was 38 (IQR: 32–45 y). Complications are displayed in Supplemental Digital Content 3. (See appendix, Supplemental Digital Content 3, which displays the obstetric complications, https://links.lww.com/PRSGO/F347.)

Table 3.

Pregnancy Outcomes

	Overall (N = 137)	Nonprocedural (n = 94)	Procedural (n = 43)	P
Live births/woman, mean (SD)	2.57 (0.95)	2.70 (0.94)	2.28 (0.93)	0.014
Live births, median (range)	2 (0–4)	2 (0–4)	1 (0–4)	0.017
Total pregnancies/woman, mean (SD)	3.60 (1.00)	3.75 (1.00)	3.28 (0.93)	0.010
Had no live births	19 (14%)	10 (11%)	9 (22%)	0.20
One or more stillbirths	3 (2%)	3 (3%)	0 (0%)	0.50
One or more miscarriages	32 (24%)	25 (27%)	7(17%)	0.50
One or more chemical pregnancies	16 (12%)	9 (10%)	7 (16%)	0.53
One or more ectopic pregnancies	3 (2.2%)	2 (2.2%)	1 (2.4%)	>0.99
One or more elective terminations	11 (8%)	10 (11%)	1 (2.4%)	0.34
Currently pregnant	18 (13%)	7 (7.4%)	11 (26%)	0.008

P < 0.05 (bolded) indicate statistical significance.

Female Respondents Who Have Been Pregnant: Procedural Versus Nonprocedural

The median age at birth of the first child was 31 (IQR: 29–33 y) among nonproceduralists and 32 (IQR: 30–35 y) among proceduralists (P = 0.082). There were no statistically significant differences in stage of career at birth of first child between proceduralists and nonproceduralists; however, the majority (22 of 43 [51.2%]) of proceduralists and a plurality (39 of 90 [43.3%]) of nonproceduralists had their first child during training (Table 2). Those in procedural specialties were significantly more likely to work over 40 hours/wk (39 of 42 [93%] versus 66 of 89 [74%]; P = 0.023). However, the time spent working on their feet was not statistically significantly different between the 2 groups (21 of 42 [50%] versus 30 of 90 [33%]); P = 0.92).

When comparing pregnancy outcomes, proceduralists had a lower average number of total pregnancies per woman (mean [SD], 3.28 [0.93] versus 3.75 [1.0]; P = 0.01) and live births per woman (mean [SD], 2.28 [0.93] versus 2.7 [0.94]; P = 0.014) than nonproceduralists (Table 3). The median number of live births was 1 among proceduralists and 2 among nonproceduralists (P = 0.017); this difference was statistically significant. A fraction of both proceduralists (9, 22%) and nonproceduralists (10, 11%) reported that none of their pregnancies resulted in a live birth (P = 0.20). There were no statistically significant differences in rates of stillbirths, miscarriages (loss of gestation 5–20 wk),²¹ chemical pregnancies (loss of gestation <5 wk),²² or elective terminations between proceduralists and nonproceduralists (Table 3).

Among proceduralists, the most common complication was preeclampsia (12%) followed by postpartum depression (9.3%) and small gestational size (9.3%) (Supplemental Digital Content 3, https://links.lww.com/PRSGO/F347). Among nonproceduralists, the most reported complication was gestational hypertension (7.4%) followed by preeclampsia (6.4%) and gestational diabetes (6.4%). There were no statistically significant differences in obstetric complications, difficulty conceiving, or use of ART among proceduralists and nonproceduralists who have been pregnant.

Female Respondents Who Have Never Been Pregnant

Over half (53%, n = 156) of female respondents reported that they had never been pregnant (median age: 30 y; IQR: 28–32 y). Of those, 94 (60%) reported that they are married or in a significant relationship; 2 women reported that they had children but did not give birth themselves. Over two-thirds (67%, n = 105) reported that they would like to have children, whereas 28 (18%) were unsure. Only 9 (5.8%) women who had not been pregnant reported undergoing cryopreservation.

Male Respondents

Of male respondents, 89 (69%) reported that they are married or in a significant relationship; 48 (38%) reported that they had children, and 16 of 28 (57%) reported that they had children during training. Of the 79 men who did not have children, 60 (76%) reported that they would like to have children, whereas 10 (13%) were unsure.

Difficulty Conceiving and Use of ART: All Participants

Among both male and female respondents, 221 (52%) reported attempting to conceive, but 63 (28.5%) had difficulty, with 32 (51%) of those having reported use of ART. Overall, 14.5% (32/221) of those who tried to conceive used ART. Of those with difficulty conceiving, respondents in procedural specialties reported higher rates of ART use (15 of 22 [68%] versus 17 of 41 [41%]); however, this difference was not statistically significant. Methods of ART use are displayed in Figure 1. Only 26 of 420 (6.2%) respondents reported that they or their partner used cryopreservation. Procedural specialists reported higher rates of cryopreservation (13 of 153 [8.5%] versus 13 of 267 [4.9%]), but this difference was not statistically significant.

Fig. 1.

Methods of ART use.

Prior Knowledge and Perceptions of Fertility Information and Assisted Reproductive Technologies

A large majority of participants, 302 of 423 (72%), reported that they never received any education regarding fertility data, cryopreservation, or the different types of ART during residency. Proceduralists were significantly more likely to have never received such information during residency (123 of 154 [80%] versus 179 of 269 [67%]; P = 0.007).

Participants were asked to provide free response answers to questions regarding perceptions of fertility, cryopreservation, and fertility education during training. Respondents endorsed positive views of cryopreservation and broadly supported fertility education curricula, with a prior subgroup analysis of surgical trainees done by this group demonstrating a particular emphasis on institution-specific insurance policies and fertility resources.²³

DISCUSSION

We examined fertility challenges across specialties, focusing on the experiences of procedural versus nonprocedural specialties. We found decreased total pregnancies and live births among physicians in procedural specialties. Previous studies have observed decreased fertility and increased pregnancy complications among physicians, particularly among those in surgical specialties.^14,19,24,25 Our study, in agreement with a previous study, did not observe an increase in complication rates among physicians in procedural specialties.²⁶ However, we did identify several trends that may contribute to difficulties and complications associated with pregnancy and fertility among physicians.

Our study revealed a median age at birth of the first child of 32 years. Although there was no significant difference between the ages at birth of the first child for proceduralists and nonproceduralists, this age is substantially delayed when compared to the age of 27 found among the general population.¹¹ Our results are comparable to those of other studies evaluating fertility among female physicians. One study of 5238 female Canadian physicians found an age at birth of first child of 32.¹¹ Another reported that 75% of female physicians in their study population delayed childbearing due to their careers.²⁷ Delayed fertility may particularly affect those in procedural specialties who undertake lengthy training programs.²⁸ Furthermore, although the age at birth of first child was similar between the 2 groups, the decreased number of total pregnancies and live births among proceduralists may elucidate difficulties in family planning among those undergoing or having completed potentially more physically demanding training programs.²⁶

We also found that proceduralists were more likely to work more than 40 hours/wkk during pregnancy when compared to nonproceduralists. Most proceduralists (93%) reported working greater than 40 hours/wk during pregnancy, with the majority working over 60 hours/wk while pregnant. Although we used a threshold of greater than 40 hours/wk to assess workload during pregnancy, more granular stratification of work hours did not demonstrate meaningful differences in outcomes within our dataset, likely due to limited subgroup sample sizes. The use of this threshold provides a pragmatic comparison to standard full-time work outside of medicine. Additionally, as some specialties transition toward more shift-based scheduling models, the structure and distribution of work hours—not just total hours worked—may represent an important factor in understanding these relationships. Increased work strain during pregnancy has been associated with adverse pregnancy outcomes, likely due to both psychological and physical stressors, as well as logistical constraints.^14,29,30 Studies among the general population have demonstrated that working greater than 40 hours/wk is associated with increased rates of adverse pregnancy outcomes such as miscarriage, preterm labor, low birth weight, and small for gestational age newborns.^25,29 Female surgeons who operate more than 12 hours/wk during the third trimester of pregnancy are at increased risk for major complications.¹⁴ Working greater than 40 hours/wk, with a significant amount of time spent on the feet, may have contributed to the higher rate of preeclampsia we noted in our study (8.5%) when compared to the general population (3.1%).³⁰ Although our study did not directly observe increases in complications among physicians who worked greater than 40 hours/wk, these findings do provide insight into work conditions during pregnancy and their potential to cause adverse pregnancy and fertility outcomes.

Another factor in adverse fertility outcomes among physicians may be a lack of education. Our data demonstrate a striking lack of fertility education, particularly among physicians in procedural specialties. Overall, 72% reported that they received no information from their residency programs regarding fertility or assisted reproductive technologies. Among those in procedural specialties, this number grew to 80%. These findings support the notion that there is a general lack of education around fertility and related institutional policies among medical professionals. Although our data demonstrate a small proportion of respondents reporting education regarding fertility or assisted reproductive technologies, our survey did not assess the timing, content, or quality of such education, or its impact on behavior or decision-making. Free-text responses, however, suggested a perceived gap in knowledge and a strong interest in structured educational resources, specifically institution-specific insurance policies and fertility resources.²³ Previous work has demonstrated that most residents, fellows, and program directors are unaware of their institutions’ insurance coverage for fertility preservation.^31,32 Other studies have demonstrated a general lack of knowledge on fertility among medical professionals.^15,33–36 Surveys of medical students, trainees, and attendings have shown that barely half of respondents were able to correctly answer the questions on female fertility, whereas only 12% of respondents were able to correctly answer questions on male fertility while expressing concern over their own fertility.³⁷

Trainees who are unaware of factors affecting fertility or policies surrounding fertility-related coverage may be less likely to recognize that fertility-related issues exist or use available resources. This may explain the lower-than-expected rate of cryopreservation found in our study. We observed an overall rate of cryopreservation of 6.2%, with similar rates between procedural and nonprocedural specialists. This is lower than the findings of a previous survey study of 1056 female physicians that reported an 11.5% fertility preservation rate.²⁷ Lacking awareness of options may contribute to these findings. Previous studies have demonstrated that physicians who received education on the risks of infertility and delayed pregnancy were more likely to have children during residency and earlier in their careers and were less likely to have miscarriages or require fertility treatments.³⁸ These findings suggest that improved fertility education may support informed decision-making and reproductive planning.^39,40

Limitations

The major limitation of our study is the response rate, albeit similar to that of similar published survey studies. This limitation is inherent to most studies designed around a survey, and even with a cash incentive and promotion among residents across specialties, it was difficult to mitigate, which raises the possibility of a response bias. Selection bias may have been introduced due to nonparticipation of certain departments and reliance in some cases on informal survey dissemination, which may have resulted in differential representation across specialties. Reasons for departmental nonparticipation were not collected, limiting our ability to assess potential systematic differences between participating and nonparticipating groups. Furthermore, individuals with strong personal experiences related to fertility may have been more likely to participate. In addition, we did not perform a formal nonresponder analysis, limiting our ability to assess differences between respondents and nonrespondents.

The classification of “procedural” versus “nonprocedural” specialties may have introduced misclassification bias, as certain procedure-based specialties were not explicitly included. We were also unable to perform detailed specialty-level comparisons due to limited sample sizes within individual specialties, which may obscure important differences in training structure, workload, and reproductive outcomes. Finally, this study was conducted at only 2 academic institutions, which may limit generalizability to other practice settings or geographic regions.

Future Directions

Future studies should include larger, multi-institutional cohorts and evaluate differences across training level, specialty, and sex. Further work is also needed to assess the impact of fertility education interventions on physician decision-making and reproductive outcomes.

In addition, fertility education curricula for trainees should be developed that may be tailored to any specialty at any hospital. Fertility education may be most effectively implemented through a combination of medical school curricula and graduate medical education initiatives, potentially integrated into orientation and supported by institutional leadership, residency program directors, and other subject matter experts. Delivery methods may include workshops or didactic sessions, as well as accessible and well-advertised resources. Fertility education and awareness have previously been suggested as a strategy to address the issue of physician infertility,^{16,34,41–43} and a medical student and resident education curriculum may be the optimal mechanism to do so.⁴⁴ However, further research is needed to evaluate the content, delivery, and impact of such interventions on physician decision-making and reproductive outcomes. In addition to addressing educational gaps, addressing structural barriers to accessing care during training may also be important. For example, the ACGME resident survey includes items related to trainee access to care, highlighting that structural barriers may have important implications for access to fertility-related care.

CONCLUSIONS

Our findings demonstrate a high prevalence of pregnancy and obstetric complications and pregnancy loss, with significantly fewer live births in those in procedural specialties versus nonprocedural specialties. In addition, our findings revealed an overwhelming lack of fertility awareness and education during training, and an interest in increased and formalized fertility education. As medical students and trainees are beginning training later, and as more women are entering the physician workforce,¹⁸ it is critical that we offer early formal education around these topics.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

ACKNOWLEDGMENTS

Chat-GPT (OpenAI, 2026) was used for editorial assistance during manuscript revision, including language refinement and formatting. All content was reviewed and verified by the authors.

Supplementary Material

gox-14-e8197-s001.pdf ^{(115.3KB, pdf)}

gox-14-e8197-s002.pdf ^{(143.2KB, pdf)}

gox-14-e8197-s003.pdf ^{(166.7KB, pdf)}