Gender Disparity in Awards in General Surgery Residency Programs

Key Points

Question  Are female general surgery residents underrepresented as award winners?

Findings  In this survey of 24 general surgery residency programs from 1997 to 2017, female residents were significantly less likely to win an award compared with male residents. This outcome factors into account the proportion of female residents in each program each year.

Meaning  These findings suggest that female general surgery residents are underrepresented as award winners.

Importance  Women are disproportionately underrecognized as award winners within medical societies. The presence of this disparity has not been investigated in training programs.

Objective  To determine the presence of a gender disparity in award winners in general surgery residency programs.

Design, Setting, and Participants  In this retrospective survey study, 32 geographically diverse academic and independent general surgery residency programs were solicited for participation. The 24 participating programs (75.0%) submitted deidentified data regarding the gender distribution of residents and trainee award recipients for the period from July 1, 1996, to June 30, 2017. Data were analyzed from September 11, 2017, to December 21, 2018.

Exposures  Time and the proportion of female trainees.

Main Outcomes and Measures  The primary outcome was the percentage of female award winners. A multilevel logistic regression model accounting for the percentage of female residents in each program compared the odds of a female resident winning an award relative to a male resident. This analysis was repeated for the first and second decades of the study. Award winners were further analyzed by type of award (clinical excellence, nonclinical excellence, teaching, or research) and selection group (medical students, residents, or faculty members).

Results  A total of 5030 of 13 760 resident person-years (36.6%) and 455 of 1447 award winners (31.4%) were female. Overall, female residents were significantly less likely to receive an award compared with male residents (odds ratio [OR], 0.44; 95% CI, 0.37-0.54; P < .001). During the first decade of the study, female residents were 70.8% less likely to receive an award compared with male residents (OR, 0.29; 95% CI, 0.19-0.45; P < .001); this improved to 49.9% less likely in the second decade (OR, 0.50; 95% CI, 0.42-0.61; P < .001). Female residents were less likely to receive an award for teaching (OR, 0.33; 95% CI, 0.26-0.42; P < .001), clinical excellence (OR, 0.44; 95% CI, 0.31-0.61; P < .001), or nonclinical excellence (OR, 0.69; 95% CI, 0.48-0.98; P = .04). No statistical difference was observed for research award winners (OR, 0.76; 95% CI, 0.42-1.12; P = .17). The largest discrepancies were observed when award recipients were chosen by residents (OR, 0.23; 95% CI, 0.14-0.39; P < .001) and students (OR, 0.32; 95% CI, 0.25-0.42; P < .001) compared with faculty members (OR, 0.52; 95% CI, 0.42-0.66; P < .001).

Conclusions and Relevance  This study found that female residents were significantly underrepresented as award recipients. These findings suggest the presence of ongoing implicit bias in surgery departments and training programs.

Gender disparities in academic medicine are pervasive, affecting differences in academic rank,^1-7 compensation,^6-13 research funding,^14-17 and receipt of recognition awards from academic specialty societies^18-21 and institutions.²² Female physicians are less likely than male physicians to lead specialty societies,^23-25 journal editorial boards,²⁵ academic divisions and departments,^26,27 medical schools,²⁸ and health systems.^3,27,29 These disparities have been described in nearly all specialties and for academic medicine overall and have persisted for decades.

At what stage do these disparities begin? Benchmarks used to compare faculty achievement, such as salary or advancement, do not exist similarly for all trainees. Gender disparity research at the medical student and resident levels has largely investigated qualitative differences in words used in medical student performance evaluations and other evaluation and recommendation tools.^30-36 A recent study on the attainment of milestones in the clinical competency framework³⁷ determined that male residents score higher and attain milestones at a faster rate than female residents and hypothesized that gender bias contributed to this disparity. However, this study was limited to emergency medicine residents at a single institution.

Although evaluations and recommendation letters are inherently important because of the effect on an individual’s future, the content of these documents is private. In contrast, the public nature of academic rank and leadership positions highlights gender disparities in an inescapable way. For female trainees and physicians, witnessing systemic gender inequities can contribute to a poorer sense of belonging and acceptance at their institution and can ultimately damage an individual’s well-being.^38-41 Awards in particular highlight a specific accomplishment, and award winners are often regarded as examples to emulate. Gender disparities exist in the award recipients from a wide range of specialty societies but have not been investigated at the trainee level. We hypothesized that gender disparities would be present among award recipients from surgical residency programs.

After the receipt of exempt status from the Partners Health Research Committee for this study, we emailed 32 surgery residency program directors to solicit participation (eAppendix in the Supplement). The programs were chosen to achieve geographic (Northeast, South, Midwest, and West) and program type (university-affiliated vs independent) diversity. Once a program director agreed to participate in the study, a templated spreadsheet was sent to the program director or the director’s designated administrator. The gender distribution of the residency for each year from July 1, 1996, to June 30, 2017, was queried. Information was also requested for each award given by the department to surgical residents of all postgraduate year levels: what the award was for, including the formal description if possible, who selected the award winner (faculty, residents, or medical students), and the gender of each recipient from 1997 to 2017. Completed spreadsheets were then returned. Programs were contacted a second time for clarity or completeness of the data as needed. To ensure the privacy of the residency programs and their alumni, no programs were identified. Furthermore, because only gender information was collected from individual programs, no personal identifiers were included.

Data were analyzed from September 11, 2017, to December 21, 2018. The number of resident person-years in all surgical programs for the entire study period was counted, and the percentage of female resident person-years was calculated. Resident person-years were used because residents were counted for each year of participation in the study: for example, in a 5-year program contributing 5 continuous years of data, a single resident could be counted 5 times and therefore would be considered to contribute 5 resident person-years. Programs with research residents were included in the overall person-years count. The percentage of female award recipients was calculated and compared with the number of female resident person-years. This calculation was then repeated for the first and second decades of the study. Four programs were unable to provide demographic information for the whole residency, so the demographic characteristics of the chief resident class each year were used as a proxy.

Awards were then categorized in 2 ways for subset analysis. Awards were first grouped by type of award or the reason it was given: clinical excellence, nonclinical excellence (for example, humanism in medicine), research, or teaching. This categorization was performed independently by 2 authors (L.E.K. and H.G.L.) based on the description of the award provided by the residency program. Areas of disagreement were clarified through further discussion with additional authors (D.S.S. and N.L.C.).

Second, awards were categorized by who determined the winner: faculty, residents, or medical students. In some cases, these groups overlapped. Within each subset, the percentage of awards given to female residents was calculated and compared with the percentage of female residents.

For each analysis, a multilevel logistic regression analysis accounting for the percentage of female residents in each program was performed to compare the odds of a female resident winning an award relative to a male resident in that program; that is, the model factors in the presence of female residents within each program when determining whether female or male residents are more or less likely to win an award. Therefore, a program’s increased likelihood of having a female award recipient does not simply reflect an increased percentage of female residents within the program. A mixed-effects logistic regression model was used for this study. Data were analyzed with the statistical software STATA, release 14 (StataCorp LLC). Two-sided P < .05 indicated significance.

Of the 32 programs solicited for participation in the study, 24 joined (75.0%). Four programs did not respond to the request for participation, and 4 programs were unable to participate because no residency awards were given at those programs. Of the participating programs, 10 were located in the Northeast, 4 in the South, 5 in the Midwest, and 5 in the West. Two programs were independent; the remaining 22 were university based.

Residency programs provided 5 to 21 years of data. Residency programs conferred 1 to 9 awards annually; some awards were given to an individual, whereas others were given to multiple recipients each year. In sum, 112 awards were given to 1447 recipients during the study period.

Programs varied in the proportion of awards distributed to women, as seen in Figure 1. The lowest percentage of awards given to women was 9.7% (3 of 31 awards) from program 5, compared with 57.7% (15 of 26 awards) from program 24. In general, programs with a greater proportion of female residents distributed a greater proportion of awards to female residents (Figure 1).

Female residents accounted for 5030 of 13 760 resident person-years (36.6%) during the study. Female residents constituted 455 of 1447 award recipients (31.4%). The absolute yearly percentages of female residents and female award recipients for the entire cohort can be seen in Figure 2. When accounting for the presence of female residents in each residency program, female residents were overall 55.6% less likely to receive an award compared with male residents (odds ratio [OR], 0.44; 95% CI, 0.37-0.54; P < .001).

In the first decade of the study, women constituted 1140 of 4032 resident person-years (28.3%) and received 69 of 315 awards (21.9%). During this time, female residents were 70.8% less likely to receive an award (OR, 0.29; 95% CI, 0.19-0.45; P < .001) when compared with male residents. In the second decade of the study, women constituted 4412 of 9583 resident person-years (46.0%), received 386 of 1132 awards (34.1%), and were 49.9% less likely to receive an award (OR, 0.50; 95% CI, 0.42-0.61; P < .001).

Taken together, our results show that although female underrepresentation has improved over time, women are still significantly less likely to win awards compared with men. Of note, the ORs represent the mean difference in the odds of receiving an award for all years combined and are adjusted for the proportion of residents who were women in each program year. In contrast, Figures 2, 3, and 4 illustrate the absolute proportion of awards granted to female residents over time.

During the study period, female residents were 67.3% less likely than male residents to receive a teaching award (OR, 0.33; 95% CI, 0.26-0.42; P < .001). Female residents were 56.4% less likely to receive an award for clinical excellence (OR, 0.44; 95% CI, 0.31-0.61; P < .001) and 31.5% less likely to receive an award for nonclinical excellence (OR, 0.69; 95% CI, 0.48-0.98; P = .04). Female residents were 23.7% less likely to receive a research award, but the difference was not significant (OR, 0.76; 95% CI, 0.42-1.12; P = .17). The absolute yearly percentage of female residents and award recipients by award type is shown in Figure 3. Our findings suggest that although the absolute proportion of research and nonclinical awards given to female residents is greater than the absolute proportion of residents who were female, between-program variation in gender distribution fully accounts for these differences.

Award recipients were chosen by medical students, residents, and faculty members. Compared with male residents, female residents were 77.0% less likely to receive an award chosen by residents (OR, 0.23; 95% CI, 0.14-0.39; P < .001), 67.8% less likely to receive an award chosen by medical students (OR, 0.32; 95% CI, 0.25-0.42; P < .001), and 47.8% less likely to receive an award chosen by faculty (OR, 0.52; 95% CI, 0.42-0.66; P < .001). The absolute yearly percentage of female residents and award recipients according to who chose the award is shown in Figure 4. The relative difference in the odds of receiving an award was smaller when awards were granted by faculty than when awards were granted by residents or students.

This is the first study in any specialty, to our knowledge, to examine gender disparities in awards given to trainees. These results clearly demonstrate that female surgical residents are underrepresented as award recipients, despite the increasing prevalence of women in surgical residencies. Although the underrepresentation as award winners has improved over time, an inequality is still present.

The degree of disparity differed by the type of award given: significant gender differences were noted in award recipients for teaching, clinical excellence, and nonclinical excellence, but not for research. One explanation for this finding is that research awards are based on research productivity or quality, which may be measured more objectively because there is a definitive body of work to review. Programs may consider the addition of research or ABSITE (American Board of Surgery In-Training Examination) awards as more objective measurements of excellence among their residents. In contrast, an individual’s teaching ability, technical skills, clinical judgment, and humanism can only be interpreted subjectively. These findings are not unique; a multitude of other studies^34,37,42,43 have demonstrated that female trainees are evaluated more poorly than male trainees across a variety of medical specialties and over domains including history taking, procedural skills, and clinical judgment.

The degree of disparity also differed by the type the evaluator. Awards chosen by residents were the least likely to be given to women, whereas awards chosen by faculty were the most likely to be given to women, although women still received significantly fewer awards than men. Previous studies⁴³ have shown that trainees and faculty alike exhibit gender biases in evaluating men and women, but this is the first study, to our knowledge, that provides quantitative data on the degree of disparities between trainees vs faculty. This result suggests that younger generations, rather than eradicating gender biases because of increasing awareness of gender-based discrimination, are actually propagating these biases. Perhaps faculty, who have a broader experience with evaluating trainees, are more able than residents and students to evaluate and award residents objectively. It is also possible that faculty have had more formal training in implicit bias than residents or students or more personal experience with biases during their careers. More research is needed to identify generational differences in gender biases and to examine their origins.

There are 2 broad potential explanations for the pervasive disparities seen in this study: (1) implicit biases that prevent the appropriate recognition of women’s success as surgeons, teachers, researchers, and leaders and (2) graduation of female surgical residents as poorer surgeons, teachers, and leaders than male surgical residents. These 2 explanations are not mutually exclusive; in fact, they may not only coexist, but also influence one another.

Implicit bias is an unconscious attitude or belief that influences behavior. Implicit biases result from gender schemas, which are culturally ingrained gender-based behaviors. Women are penalized for acting outside of traditional gender schemas, which call for women to be nurturing and communal. Women who are more assertive, confident, or independent—skills that are valued in surgery—are punished for not acting in a more “female” way.⁴⁴ Furthermore, when a woman succeeds in a historically male field, such as surgery, she is less liked and more disparaged, resulting in poorer evaluations.⁴⁵ Two publications^37,43 highlight the role of implicit bias in medical education evaluations. In a study by Dayal et al,³⁷ female emergency medicine resident milestone scores universally decreased throughout training compared with those of male residents. In a study by Thackeray et al,⁴³ female gastroenterology fellows were evaluated significantly more poorly by attending physicians when compared with male fellows, and female gastroenterology physicians were evaluated more poorly than male physicians by fellows. The uniform differences in female and male evaluations suggest widespread bias is at work, rather than an individual’s “diminished competency or skill.”^37(p655)

To acknowledge all explanations, we must consider the possibility that female surgical residents are not as well trained as male surgical residents by general surgery residency programs. This is not to suggest that female surgeons are less skilled than male surgeons; on the contrary, recent studies have shown that female surgeons have superior or equivalent outcomes when compared with male surgeons.^46,47 However, residency training may treat male and female trainees differently. In 2 studies investigating operative autonomy given to trainees, the first investigating thoracic surgery residents at 7 institutions and the second investigating general surgery residents at 14 institutions, Meyerson et al^48,49 determined that female residents received fewer opportunities to operate independently or to lead an operation than male residents. A study investigating the role of resident gender on autonomy during laparoscopic general surgery cases⁵⁰ also determined that female residents received more intraoperative guidance than male residents. Consequently, there are fewer chances for female trainees to advance technical skill and intraoperative decision-making and gain surgical confidence. There may instead be a maladaptive loop in which decreased autonomy begets less skill and confidence.⁵¹ In a recent review of emergency medicine resident evaluations,³⁵ male residents were given clear and consistent feedback, whereas female residents were given inconsistent and contradictory feedback; vague feedback lacks specific guidance on areas for improvement and prevents reinforcement of positive behaviors.⁵² When extrapolating these findings to general surgery trainees, it is easy to see how female residents may not receive useful, actionable advice about their strengths and weaknesses, hampering overall growth.

On a positive note, some residency programs appropriately recognize female achievement, as demonstrated through proportional or, rarely, overrepresentation of women as award winners. In general, residency programs with a higher proportion of female residents demonstrated more equitable disbursement of awards. As the national proportion of women in surgical residency increases with time, we hope that the presence of female award winners will also increase within individual programs. Furthermore, although the residency programs that participated in this study are deidentified, future studies should examine characteristics of residency programs with appropriate recognition of female residents to understand the cultures and environments that produce these equitable attitudes and perspectives. Whether improvements in gender equity reflect an increased proportion of female residents, female leadership, or, equally important, male sponsorship would provide important insights into promoting effective change. Until that time, individual residency programs—the programs studied herein as well as programs not included—should carefully review their own award history to determine whether female residents are appropriately recognized for their achievements.

There are strengths to this study. Participating programs are geographically diverse, indicating that our findings are not limited to a subgroup of residency programs. All collected data are historical, preventing a Hawthorne effect or social desirability bias on the part of individual residency programs.³⁵ Furthermore, this study captures contemporary data. Although some trainees and physicians experience gender discrimination in the workplace, others may believe that disparities are part of a bygone era.³⁸ Instead, our findings demonstrate a concerning, persistent disparity.

There are also limitations to our study. The 24 residency programs that participated may not adequately represent all residency programs. Notably, only 2 independent residency programs participated. Not all residency programs provided data for the entire study period, potentially weighting the results toward more recent years when female residents were more prevalent. Gender was categorized as binary male or female, which may not accurately capture an individual’s gender identity. Last, male evaluator gender has been shown elsewhere to enhance gender disparities in evaluations.⁴² However, gender information about the faculty and medical students conferring the awards was not available, limiting the ability to assess the role of evaluator gender in the decision-making process.

This nationwide study of award winners in surgical residency programs determined the presence of a widespread gender disparity in award recipients persisting across residency programs, types of awards, and categories of award conferrers. This disparity exists despite improving and near-equalizing gender demographics in surgical training and is particularly notable among students and trainees. Attention should be paid within residency training programs to gender disparities in award winners: well-publicized, systemic gender biases have potential to cause burnout, limit future career potential of female surgical trainees, and discourage women from choosing a general surgical career or training program.

Accepted for Publication: May 31, 2020.

Corresponding Author: Nancy L. Cho, MD, Department of Surgery, Brigham and Women’s Hospital, 75 Francis St, ASB III, Boston, MA 02115 (nlcho@bwh.harvard.edu).

Published Online: September 2, 2020. doi:10.1001/jamasurg.2020.3518

Author Contributions: Dr Cho had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Kuo, Lyu, Melnitchouk, Smink, Cho.

Acquisition, analysis, or interpretation of data: Kuo, Lyu, Jarman, Doherty, Smink, Cho.

Drafting of the manuscript: Kuo, Cho.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Lyu, Jarman.

Administrative, technical, or material support: Melnitchouk, Doherty, Smink.

Supervision: Kuo, Melnitchouk, Smink, Cho.

Conflict of Interest Disclosures: Dr Jarman reported receiving grants from National Institute on Aging and the US Department of Defense outside the submitted work. No other disclosures were reported.

Disclaimer: Dr Doherty is the Review Editor of JAMA Surgery, but he was not involved in any of the decisions regarding review of the manuscript or its acceptance.