Schroen, Anneke T. MD, MPH; Brownstein, Michelle R. MD; Sheldon, George F. MD
The growth of women physicians and scientists in academic leadership positions has been inadequate over the last 25 years. In its 2002 report, the Increasing Women's Leadership Project Implementation Committee of the Association of American Medical Colleges (AAMC) found that the percentage of women who hold full professorships in medical school faculties grew from 9.9% in 1985 to only 10.9% in 2001.1 Although modest increases in the proportion of women faculty at all ranks in U.S. medical schools have been noted, significant disparities in academic advancement and professional opportunities persist between men and women.
As in other medical disciplines, women in surgery are less likely to attain high academic positions, become tenured, or publish.2–5 Proposed reasons for this disparity in academic productivity include disproportionate childcare responsibilities, lack of mentoring, sexual harassment or other gender-based discrimination, professional isolation, and different career motivations or priorities. Understanding the possible barriers women in medicine face in reaching their full professional potential takes on additional importance given the increasing proportion of women in the medical workforce. In 2002, 24% of surgical residents are women despite the fact that 46% of medical students are women.6 Our understanding of the causes behind this persistent imbalance is limited. The environment in which women surgeons train and ultimately practice affects the career's appeal for female medical students, particularly if their first impressions of fully trained female surgeons are from their academic faculty.
Prior studies addressing the barriers to women's advancement in surgery have several limitations. One included female physicians in all surgical subspecialties, including obstetrics-gynecology and ophthalmology, resulting in general surgeons representing only 12% of the study population.2 Others focused only on particular surgical subspecialties other than general surgery.3,4 Still other studies did not include men as a comparison group.2,5 Finally no study included women who had left academia, thereby leaving the account of their promotion and retention problems incomplete. The purpose of this study was to better describe the professional activities and accomplishments, career aspirations and perceptions, and personal lives of both women and men general surgeons using a detailed questionnaire. Of the larger study, this report focuses specifically on:
* the similarities and differences between men and women general surgeons in academic careers;
* the factors associated with differing academic productivity and career satisfaction between men and women; and
* the responses regarding influences for intending to leave academia by women surgeons.
The study population was created using the American College of Surgeons (ACS) membership roster in January 1998. All female members (1,076) and a randomly generated 2:1 sample of male members (2,152) listed as general surgeons, matched for fellowship status, were selected. Members in residency or fellowship training, as well as foreign or honorary members, were excluded. The total sample (3,228) represented approximately 17% of the total ACS's general surgeon membership.
A self-administered instrument consisting of 131 questions was developed7 using existing instruments for surveying women physicians8,9 and assessing physicians’ career satisfaction.10 Adaptations to these instruments were made based on suggestions from focus groups of men and women practicing surgery in various settings. The final instrument was piloted with a focus-group of practicing surgeons (three women and three men). Detailed questions regarding current practice attributes, surgical training, practice activities, harassment, malpractice, career satisfaction, and personal life characteristics were included.
The factors influencing their practices or their wish to change practices were assessed on a five-point Likert scale (1 = very important; 3 = somewhat important, 5 = not important). These questions specifically addressed the respondents’ reasons for viewing their number of publications or grants as insufficient for career stage, reasons for leaving academic practice by year 2000, and reasons for delaying childbearing. A different five-point Likert scale (1 = strong agreement, 5 = strong disagreement) measured the respondents’ relations with colleagues, career satisfaction, and perceptions of personal time. These validated questions were adapted with permission from the Physician Worklife Survey.10
The survey was conducted in three direct mailings between September 1998 and March 1999. Responses received by May 1, 1999, were used for the study. Respondents either returned the printed questionnaire or answered the questions electronically on a specially designed, secure Web site. The questions were identical in both formats. All data were double-entered into a database to minimize errors.
Responses were screened for eligibility. Responses were deemed ineligible if the surgeon was not in active practice (active practice did not include residency or fellowship training), the surgeon declined to participate, or if the surgeon was unable to be located with forwarding addresses.
Responding surgeons were categorized by practice type, age, and gender for analytical purposes. Surgeons with an academic appointment, designated as “academic surgeons” or “tenure-track surgeons,” are reviewed in this paper. Academic surgeons were those who held an academic appointment and were not concurrently in a private practice. Tenure-track surgeons were a subset of academic surgeons who also held a tenure-track position. “Young academic surgeons” were defined as those younger than 45 years.
Analysis of each questionnaire item was performed by gender due to the oversampling of women. A t test or one-way analysis of variance was used for continuous data and the chi-square test was used for categorical data. In instances of nonnormal data distribution, Wilcoxon rank sum test was used and median values reported. Likert scale data were viewed either as continuous with mean values reported or consolidated into three categories (agree, neutral, or disagree). Multiple regression models were used to verify the significance of associations found on bivariate analysis adjusting for related factors. Because of multiple testing performed for certain outcomes, values of p > .01 were interpreted more cautiously. Analytical tests were performed with Stata 6.0 software (STATA Corp., College Station, TX). This study was approved for exemption by the University of North Carolina at Chapel Hill Medical School Institutional Review Board.
The response rate was 1,843 of 3,228 (57.1%). Responses were roughly equivalent for male (56%) and female (58%) surgeons. Two hundred twenty-one responses were excluded based on eligibility criteria, with 1,622 eligible responses (1,050 from men, 572 from women) remaining for analysis. Of the total respondents, 80% of men were primarily in private practice compared with 67% of women (p < .001).
The subset of respondents referred to as academic surgeons and tenure-track surgeons are described in Table 1. Academic surgeons constituted 20% of the total respondents (317) and tenure-track surgeons comprised 9% of the total respondents (150). The remaining results referring to academic surgeons relate to these 317 surgeons.
Most academic surgeons were white (89%), and the remainder described their ethnicity as black/African American (4%), Asian American (4%), and Hispanic (3%). Over 99% of the academic surgeons were board certified in general surgery and approximately 30% were board certified in a surgical subspecialty, with no significant difference between men and women. The majority of academic surgeons (74%) stated that research is expected as part of their jobs, yet only 48% of respondents were in tenure-track positions. The proportions of men and women who held tenure-track positions, even when limited to surgeons 45 years and older, did not differ significantly. However, men were significantly more likely to be tenured than were women, even when considering only more senior academic surgeons (see Table 1).
To delineate career goals, respondents were asked their current academic position as well as their ultimate goal for academic rank. Among the academic surgeons, men, even young surgeons, held higher academic positions than women. In the smaller subset of tenure-track surgeons, differences in academic rank between men and women 45 years or older were not statistically significant. For young tenure-track surgeons, however, men were more likely than women to have attained the rank of professor (see Table 1).
Men expressed higher ultimate goals for academic rank than women. For example, 25% of men declared chairmanship to be their ultimate goal compared with 5% of women. When limited to tenure-track surgeons, 36% of men and 8% of women wished to chair a department. Women in tenure-track positions most commonly set professor (41%) or division chief (23%) as their ultimate goal, whereas men in tenure-track positions most commonly aspired to chairman (36%), professor (31%), or division chief (13%).
To ascertain one measure of academic productivity, respondents were asked to estimate how many first author or senior author articles they had published in professional journals, excluding book chapters, abstracts, or case reports. The number of articles published by academic surgeons ranged from zero to 380. Differences in the median number of publications between male and female academic surgeons was pronounced, which persisted when more accomplished, senior surgeons were removed from the comparison (see Table 2). Among the tenure-track surgeons, the median number of publications for professors, associate professors, and assistant professors was 70, 30, and 10, respectively. Most men (81%) and women (68%) in tenure-track positions had published within the last year. Marital status or parenthood did not influence number of publications for either men or women. There was no statistical difference for the median number of publications by academic rank between female surgeons with and without children.
Fifty-seven percent of academic surgeons and 51% of tenure-track surgeons indicated feeling their number of publications was not sufficient for their career stage. Among the latter group, 35% of men and 72% of women felt dissatisfied in this regard. For those tenure-track surgeons younger than 45 years, 42% of men and 77% of women felt his or her number of publications was inadequate. Among the male tenure-track surgeons, 57% of assistant professors were dissatisfied with their publication quantity, but this fell to 43% for associate professors and 20% for professors. With women in tenure-track positions, this dissatisfaction remained common in all ranks. Eighty-two percent of female assistant professors, 59% of associate professors, and 64% of professors replied that they felt their number of publications to be insufficient for their career stage.
Respondents were similarly asked to estimate their grant funding awarded over the course of their careers. Forty-eight percent of the academic surgeons had not received any grant funding to date in contrast with 16% of the tenure-track surgeons. Therefore, grant totals for tenure-track surgeons only are listed in Table 3 by gender, age, and academic rank. Nearly half of all tenure-track surgeons felt their level of grant funding was sufficient for career stage. Tenure-track women were less likely to be satisfied with total grant funding than were men (39% versus 56%, p = .04).
Respondents who felt their publication rate or their grant funding was insufficient indicated to what degree certain perceived factors affected their productivity. Using a Likert scale (1 = very important, 5 = not important), tenure-track surgeons graded the effect of the following factors (listed here with overall average score): inadequate time (1.3), inadequate funding (2.3), inadequate mentoring (2.5), inadequate collaboration opportunities (3.0), inadequate training (3.3), and no interest in publishing (4.4). Although female academic surgeons overall indicated that inadequate mentoring and training had a greater effect on publishing than did men, these differences between men and women were no longer significant among tenure-track surgeons. Using the same scale to rank factors that influenced amounts of grant awards, tenure-track surgeons voiced similar opinions irrespective of gender. These surgeons ranked the following factors as affecting grant awards: inadequate time (1.5), inadequate mentoring (2.1), inadequate departmental support (2.4), inadequate training (2.8), inadequate collaboration opportunities (2.9), and no interest in pursuing funding (4.1).
Intentions to Leave Academia
Eighteen percent of the academic surgeons overall and 15% of tenure-track surgeons admitted to seriously considering leaving academic medicine by the year 2000. Among the tenure-track surgeons, 11% of men and 20% of women (p = .10) reported considering leaving academia. Reported in average Likert-scale scores, the most important reasons for considering leaving included personal time requirements (2.4), overall stress (2.7), inadequate career advancement (2.8), and inadequate mentoring (2.9). The least important factors influencing the decision were teaching responsibilities (4.5), malpractice concerns (4.4), and inadequate case diversity (4.1). The remaining factors were changing career interests (3.0), uncertainty of grant funding (3.0), family responsibilities (3.1), sense of isolation (3.2), harassment (3.2), inadequate job opportunities (3.3), inadequate reimbursement (3.4), managed care environment (3.5), research responsibilities (3.9), and general disinterest in academia (3.9). Men and women generally graded potential factors similarly. The two exceptions were that men in tenure-track positions graded inadequate mentoring relatively unimportant (3.9) while women viewed it as quite important (2.3) (p = .01), and women rated a sense of isolation as having much greater impact than men (2.3 versus 4.3, p = .02).
Approximately 10% of both men and women among tenure-track professors and associate professors were considering leaving academia. However, among assistant professors, 29% of women compared with 5% of men (p = .03) expressed an interest in leaving their academic careers by year 2000. For this group of women, the most important reason for leaving was inadequate mentoring, with over 75% rating this factor as a 1 or 2 on the Likert scale. The next most important reason for this group of academic women was inadequate career advancement.
Harassment was defined as unwanted physical attention, propositions, hostilities or threats. Among academic surgeons, 8.3% of men and 4.7% of women (p = .19) reported having harassment charges brought against them at some point during their careers. In contrast, 10% of men and 61% of women reported experiencing harassment (p < .001). The two forms of harassment most commonly experienced by women were sexual (59%) and gender based, nonsexual (85%), and harassment stemming from lifestyle (13%) or ethnicity (10%) was much less frequent. Of men academic surgeons who had experienced harassment, on the other hand, most (61%) described different forms, which were not described on the questionnaire. The other harassment men experienced was sexual (33%), gender based but nonsexual (22%), lifestyle based (11%), and ethnically based (22%). For women, harassment was experienced fairly evenly throughout their professional careers with occurrences during medical school (47%), graduate training (75%), and practice (64%). In contrast, men encountered harassment predominantly in practice (83%), with only 17% claiming harassment in medical school and 22% in graduate training. For women, the harasser was commonly a senior surgeon (77%), a surgeon-peer (55%), or a nonsurgeon physician (46%). For men, it was senior surgeons (44%), surgeon-peers (33%), nurses (39%), and ancillary staff (33%). Approximately 25% of both men and women who had experienced harassment perceived this as coming from patients or their families.
Marriage and Parenthood
Women academic surgeons were more likely to have never been married compared with men (23% versus 4%) and were more likely to be living alone than men (24% versus 4%). Among the female married academic surgeons, 28% were married to another surgeon, 14% were married to a physician of another specialty, and nearly 90% lived in a dual-career household. Fifteen percent of men in academia were married to other physicians, and just over 50% were in dual-career households.
Childbearing and childrearing issues highlight other differences in the experiences of men and women academic surgeons. Among academic surgeons, 62% of women and 90% of men had children. Women intentionally delayed childbearing more commonly than men (63% versus 41%, p < .001). The two most important reasons cited for delaying childbearing were surgical training responsibilities and personal career demands for both men and women. Training responsibilities had a slightly higher impact on women, but personal career demands were viewed equally by both genders. Women were more likely to postpone all childbearing until the completion of all surgical training than were men (67% versus 37%, p < .001). Childcare predominantly took place in the surgeon's home for about 75% of both men and women academic surgeons. In the households of male surgeons, the spouse (63%) or an employed person (29%) was the primary childcare provider. For children of female surgeons, an employed person provided the childcare in 88% of these families and the spouse or other family member was the primary childcare provider in only 5% of cases.
Respondents were asked to estimate their annual personal income. Seventy-three percent of academic surgeons made between $100,000 to $300,000 annually. A total of 57% of men versus 24% of women made over $200,000 per year. Personal income did not differ between men and women at the full professor level among tenure-track surgeons. However, among tenure-track associate professors, 73% of men versus 12% of women earned more than $200,000 annually; among tenure-track assistant professors, 90% of men versus 69% of women made over $100,000 per year (see Table 4). Academic rank and gender were independently associated with income when adjusted for surgeon's age and weekly hours spent on various clinical activities, including inpatient surgeries. Only 26% of women versus 46% of men claimed no student debts at medical school graduation. Approximately 40% of both men and women in academic surgery had some educational debt under $50,000.
To gauge job satisfaction, participants rated 20 different statements addressing general career satisfaction, professional relationships, intellectual and technical challenge, practice environment, and personal life on a five-point Likert scale. The responses on 11 representative statements for men and women tenure-track surgeons are shown in Table 5. Career satisfaction overall was very high; a large majority replied they would choose to become surgeons again and would recommend it as a career to medical students. Despite this, women were significantly more likely to feel career advancement opportunities were not equally available to them and to feel a sense of isolation from their colleagues. Both men and women expressed mixed sentiments on the value of mentor relationships in their careers and on the support of surgical colleagues in efforts to balance work and family responsibilities. Women in tenure-track positions were not more likely to state that their career goals had been influenced by a need to maximize control over personal time. Finally, half of both men and women expressed that their work schedules did not leave them enough time for their personal family lives.
By focusing solely on general surgeons and by asking extensive questions about many facets of their professional and personal lives, we have tried to provide a more comprehensive and distinct view of the experiences of dedicated academic general surgeons. Although the differences in the academic careers of men and women depicted here may not be unique to general surgery, their pertinence to fields like general surgery is increasing. Interest in this specialty faces a potential downward trend among senior medical students, nearly half of whom are women, and an apparently stagnant level of interest among women medical students. The most influential impressions medical students receive of general surgery are likely from the academic faculty and residents they encounter during their clerkships. Recognizing and improving existing disparities may improve the recruitment and retention of competitive men and women to general surgery. Furthermore, educating a surgeon costs a significant amount. Not capitalizing on this investment through inadequately emphasizing career development for all general surgeons constitutes an unfortunate waste.
Although more women than men entered academia upon finishing training, women were less likely to attain higher academic ranks and achieve promotion at a young age. This phenomenon has been noted repeatedly in other medical specialties and academia in general.11,12 Explanations include differing career aspirations and motivations, lower productivity by the most commonly used measures, fewer or less effective mentoring and networking opportunities, fewer institutional resources, and disproportionate parenting responsibilities. Our study's results show different career aspirations among women than among men. Attaining the rank of professor, however, was a common aspiration among general surgeons. Data from the AAMC indicate that the proportion of women in associate and full professor positions in academic surgery departments has not significantly changed between 1995 and 2001.1 In 2001, 10% of associate professors and 4% of full professors in surgery were women.1 A cohort analysis of male and female graduates over a 15-year period showed that fewer women than expected were promoted from assistant to associate professor and from associate to full professor in surgery, as in most other clinical specialties. The discrepancy in actual versus expected promotions, however, was much greater in the advancement to associate professor, suggesting this to be the limiting step for many women in academic general surgery.11 Our study indicates that nearly a third of women at the assistant professor level contemplate leaving academia, a considerably higher proportion than of men at comparable rank or women at higher ranks.
Number of publications, a highly valued productivity measure, differed dramatically between men and women. Women in academic general surgery published one third to one half as many journal articles as men. Eliciting a specific reason for this discrepancy proved difficult. Within our study, marital status and parenthood did not significantly influence publication number for either gender. Gender did not affect overall hours worked per week (median = 50 hours) or hours dedicated to research activities (median = 5 hours). Women were no more likely than men to attribute insufficient publications to inadequate time, funding, mentoring, or collaboration opportunities. Yet, women reported fewer career advancement opportunities and feeling isolated from surgical peers. Other publications, such as book chapters, were not queried, but it is unlikely their inclusion would significantly reduce the differences we found.
Our results suggest the environment of academic experiences for men and women in general surgery is not equivalent. Prior studies, which have shown a similar disparity in publications or advancement in academic medicine, have identified that parental status affects women's productivity by limiting their expandable work hours.13 Others have found a difference in institutional support14 or in networking opportunities.15 Female faculty appear to have fewer colleagues in higher ranking positions or from former institutions available for networking, resulting in a smaller net social capital. In addition to access to fewer or less effective mentoring relationships, studies have identified a discrepancy in the institutional support afforded to female faculty. This includes being offered less office or laboratory space, less dedicated time for research, and less assistance with grant support early in their careers.16
Although the influx of women into higher academic ranks will naturally take time, it is concerning that the AAMC has shown that the proportion of men to women at higher academic professional levels has remained essentially unchanged for over 15 years.17 Bickel concludes that women in academic medicine do not succeed at the same rate as men because of a complex constellation of factors, including women's career choices, discrimination, gender stereotypes, isolation, and limitations in balancing family and professional responsibilities, which results in a “cumulative career disadvantage.”17 Often these factors are difficult to isolate, deeply entrenched in organizational culture to the degree that they are hardly recognized or questioned, or occur with no intended harm. Organizational structures in medicine, as well as in other professions and in business, have grown around a paradigm of men whose spouses provide significant support at home, allowing the men to spend additional time at work. Organizations reward such unlimited availability for work as evidence of a man's dedication to the job. How efforts to balance work and personal life are perceived to influence work effectiveness may also differ between men and women. A survey of a large surgery department found that most men and women surgeons thought that female faculty who were parents were less able to meet job expectations, but that only one third of men and none of the women felt that male faculty who were parents were less able to meet these career expectations.18 Further studies have shown that the conflict of balancing work and personal lives has an increasing impact on both male and female physicians and contributes to professional burnout.19 With current medical students weighing lifestyle issues more heavily in their specialty choice and with evolving social constructs forming more dual career families, addressing the conflict of professional and personal life is becoming increasingly critical for academic surgery. It would seem more pragmatic to address these issues on a larger, institutional scale, rather than piecing together solutions for faculty on an individual basis.
This study has several limitations. The response rate of 57% overall is comparable to other large physician surveys.5,10 Furthermore, a review of published physician survey research indicates that 54% is the mean response rate for this study genre.20 Nonetheless a significant nonresponse rate raises issues of generalizability. Respondents and nonrespondents were similar in gender composition. Comparisons of responders and nonresponders based on practice characteristics were not possible because practice setting information was not available for nonrespondents. Finally, the study respondents appeared representative of academic surgeons in terms of academic rank. AAMC data demonstrate that 53% of academic surgeons were professors or associate professors in 1999.21 In this study, 54% of respondents held professor or associate professor positions.
To deal with gender-related issues in academic surgery, the recommendations put forth by the AAMC's Increasing Women Leadership Project Implementation Committee could serve as a valuable guide. These include emphasizing faculty diversity in departmental reviews and evaluating department chairs in their faculty development of both women and men; targeting the professional development needs of women within the context of helping all faculty to make the most of their appointments and potential; conducting assessments of institutional practices that tend to favor academic success for men over women; enhancing the effectiveness of search committees in attracting women candidates, including reassessing how candidates’ qualifications are defined and evaluated; and financially supporting institutional Women in Medicine programs and the AAMC's Women Liaison Officer.1 Considering the large amounts of resources that are expended to train each surgeon, regardless of gender, it would behoove academic departments to view career development of its faculty as a high priority. By strengthening its faculty, the specialty as a whole benefits, not only in terms of academic pursuits in surgery but also in recruiting the most competitive students to surgery.
This work supported by grants from the Department of Surgery at the University of North Carolina-Chapel Hill, the Association of Women Surgeons, and the Robert Wood Johnson Foundation. We appreciate Carol Porter from the Cecil G Sheps Center for Health Services Research at the University of North Carolina at Chapel Hill for her assistance with dataset management.