eAppendix. Survey Questionnaire
eTable. Pairwise Comparisons of the Distribution of Resident Responses to Survey Questions Based on Gender and Visible Minority Status
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Mocanu V, Kuper TM, Marini W, et al. Intersectionality of Gender and Visible Minority Status Among General Surgery Residents in Canada. JAMA Surg. 2020;155(10):e202828. doi:10.1001/jamasurg.2020.2828
How is diversity associated with resident wellness and resiliency?
In this survey study of 210 residents enrolled in general surgery training programs in Canada, being a woman and identifying as a visible minority were associated with adverse implications for the training experience during general surgery education.
The results of this study serve as a call to training programs, accrediting bodies, and the medical community that new strategies focused on the intersectionality of gender and race/ethnicity are needed to improve the training experience of at-risk residents.
Within medical specialties, surgical disciplines disproportionately and routinely demonstrate the greatest underrepresentation of women and individuals from racial/ethnic minority groups. Understanding the role that diversity plays in surgical resident training may identify strategies that foster resident resiliency, optimize surgical training, and improve patient outcomes.
To examine the implication of gender and visible minority (VM [ie, nonaboriginal people who are not White individuals]) status for resiliency and training experiences of general surgery residents in Canada.
Design, Setting, and Participants
In this survey study, a 129-item questionnaire was emailed from May 2018 to July 2018 to all residents enrolled in all Canadian general surgery training programs during the 2017-2018 training year. Survey responses were extracted and categorized into 5 major themes. The survey was designed by the Resident Committee and reviewed by the Governing Board of the Canadian Association of General Surgeons. French and English versions of the survey were created, distributed, and administered using Google Forms.
Main Outcomes and Measures
Survey questions were formulated to characterize resident diversity and training experience. Self-perceptions of diversity, mentorship, and training experience were evaluated using a 5-point Likert scale (1 for strongly disagree, 2 for disagree, 3 for neither agree or disagree, 4 for agree, and 5 for strongly agree) and open-ended responses. The frequency of perceived unprofessional workplace encounters was evaluated using a 5-point scale (1 for daily, 2 for weekly, 3 for monthly, 4 for annually, and 5 for never).
Of the 510 general surgery residents invited, a total of 210 residents (40.5%) completed the survey. Most respondents were younger than 30 years (119 [56.7%]), were women (112 [53.3%]), reported English as their first language (133 [63.3%]), did not identify as a VM (147 [70.0%]), had no dependents (184 [87.6%]), and were Canadian medical graduates (178 [84.8%]). Women residents who identified as VM compared with male residents who did not identify as a VM were less likely to agree or strongly agree that they had a collegial relationship with staff, (21 [63.6%] vs 61 [89.7%]; P = .01), to feel like they fit in with their training programs (21 [63.6%] vs 56 [82.3%]; P = .003), and to feel valued at work (15 [45.4%] vs 47 [69.1%]; P = .03). Both female residents and female residents who identified as VM described significant concerns about receiving fewer training opportunities because of their gender vs their male peers (54 [48.2%] vs 3 [3.0%]; P < .001). Ninety-one of 112 female residents (81.2%) reported feeling that their medical expertise was dismissed because of their gender at least once annually, with 37 women (33.0%) experiencing dismissal of their expertise at least once every week (P < .001). In contrast, 98% of male residents reported never experiencing dismissal of their medical expertise because of their gender. Similarly, residents with VM status vs those without VM status reported at least monthly dismissal of their expertise because of their race/ethnicity (9 of 63 [14.3%] vs 1 of 147 [0.7%]; P < .001).
Conclusions and Relevance
In this study, female sex and VM status appeared to be associated with adverse implications for the training experience of general surgery residents. These findings suggest that new strategies focused on the intersectionality of gender and race/ethnicity are needed to improve the training experience of at-risk residents.
Although less scrutinized than in the corporate sector,1-9 diversity within the health care workforce does not reflect the diversity in the general population.10-19 In the United States, Black, Hispanic, and Native American individuals compose one-third of the total population but only 9% of all practicing physicians.20 Less than one-third of all full-time academic physicians are women, and only 4% belong to underrepresented racial/ethnic groups.21 Minority and female faculty members are less likely to be promoted, obtain fewer grants, and are compensated less than their White or male counterparts.21-27 Within medical specialties, surgical disciplines routinely demonstrate the greatest underrepresentation of women and visible minorities.15,28-33 The term visible minority (VM) is defined by the Canadian government as “persons other than aboriginal peoples, who are non-Caucasian in race or non-white in color.”34
Recent research has found that mistreatment because of gender and racial/ethnic biases is common and is associated with burnout and suicidal thoughts among general surgery residents in the US.35 Although these experiences seemed to be more prevalent in female residents, information on VM status was not captured,35 limiting the assessment of an important at-risk group of residents. Additional details on the experiences of surgical residents are needed to inform the development of strategies that can help all minority residents overcome the barriers that they encounter during surgical training. Currently, the diversity within Canadian surgical programs and its implication for resident training are not known, limiting efforts to develop such strategies.
The goal of this study was to use a national survey to investigate the implication of gender and VM status for resiliency and training experiences of general surgery residents in Canada. We hypothesized that gender and VM status were associated with barriers to resident wellness and resiliency. We characterized the diversity and training experiences in a large Canadian surgical cohort to identify strategies for overcoming these barriers and improving surgical training.
From May 2018 to July 2018 (during the 2017-2018 training year), a 129-question electronic survey was emailed to all residents enrolled in general surgery training programs across Canada (n = 519). The survey was designed by the Resident Committee and reviewed by the Governing Board of the Canadian Association of General Surgeons. French and English versions of the survey were created, distributed, and administered using Google Forms; the final survey is presented in eAppendix in the Supplement. The survey was not validated prior to distribution. Three reminder emails were sent to nonrespondents. No stipends or gifts in-kind were provided for survey completion. Written informed consent was obtained from all participants before survey initiation. This study received ethics approval from the University of Alberta Research Ethics Board.
Survey questions were formulated to characterize resident diversity and training experience to identify factors associated with resident wellness and resiliency. Basic demographic data included age, sex, sexual orientation, location of training program, postgraduate year of training (1-5, research), relationship status (single, separated, divorced, or in a relationship), number of dependents, entry program (Canadian Medical Graduate [CMG], International Medical Graduate [IMG], or visa [trainee who required visa sponsorship]), VM status, first language, number of years since immigration, and race/ethnicity of origin. Visible minority status was self-identified and categorized according to the racial/ethnic groups in the 2016 Canadian census. Categories with fewer than 3 respondents were collapsed to minimize resident identification. Self-perceptions of diversity, mentorship, and training experience were evaluated using a 5-point Likert scale (1 for strongly disagree, 2 for disagree, 3 for neither agree or disagree, 4 for agree, and 5 for strongly agree) and open-ended responses. The frequency of perceived unprofessional workplace encounters was evaluated using a 5-point scale (1 for daily, 2 for weekly, 3 for monthly, 4 for annually, and 5 for never).
To compare the sex distribution between respondents and all general surgery residents, we obtained residency match data from 2015 to 2019 through the Canadian Resident Matching Service. All CMGs, IMGs, and US medical graduates who were matched to a Canadian general surgery training program were characterized by match year and sex. Data on VM status were not available.
Categorical variables were expressed as frequencies and percentages, and continuous data were expressed as mean (SD). Demographics were compared between female and male respondents and between those who did and did not identify as a VM using a paired, 2-tailed t test, χ2 test, and the Wilcoxon rank sum test as appropriate. Members of the Resident Committee of the Canadian Association of General Surgeons reviewed individual open-ended responses and identified themes and barriers through group consensus.
The study cohort was divided into 4 groups defined a priori according to gender and VM status. Differences in responses to Likert-type questions between the 4 groups were evaluated using the Kruskal-Wallis test. Pairwise comparisons for statistically significant responses were then performed using the Dunn test, and the resulting P values were adjusted using the Bonferroni correction for multiple comparisons (eTable in the Supplement). Several questions pertained specifically to differential experiences based on gender or race/ethnicity. The Wilcoxon rank sum test was used to compare responses between female and male respondents and those who did and did not identify as a VM. All submitted surveys were fully completed.
A 2-sided α < .05 was considered statistically significant. All statistical analyses were performed with Stata, version 15 (StataCorp LLC). Data analysis was conducted from September 1, 2018, to April 1, 2020.
Of the 510 general surgery residents invited, a total of 210 completed the survey, for a response rate of 40.5%. Most respondents were younger than 30 years (119 [56.7%]), were women (112 [53.3%]), reported English as a first language (133 [63.3%]), did not identify as a VM (147 [70.0%]), had no dependents (184 [87.6%]), and were CMGs (178 [84.8%]) (Table 1). All postgraduate year training levels and training programs had nearly uniform representation after accounting for the size of respective programs. The 3 most common self-identified race/ethnicity groups among residents were European (105 [49.8%]), Asian (57 [27.4%]), and nonaboriginal North American (51 [24.4%]). The proportion of female respondents was 53.3% (n = 112), which was identical to the proportion of 53.4% women (n = 238) who were matched to training programs from 2015 to 2019 (Table 2).
Unadjusted bivariate analysis revealed that IMGs and those with visa sponsorship were more likely than CMGs to be men (IMG: 13 [61.9%] and visa: 9 [81.8%] vs CMG: 76 [42.7%]; P = .02) and to identify as a VM (IMG: 6 [28.6%] and visa: 8 [72.7%] vs CMG: 49 [27.5%]; P = .006) (Table 1). Female residents were more likely than their male counterparts to be single (42 [37.5%] vs 24 [24.5%]; P = .01) and to be separated or divorced (3 [2.7%] vs 0 [0%]; P = .01). Residents who identified as a VM were more likely than those who did not have a VM status to be immigrants (30 [47.6%] vs 11 [7.5%]; P < .001) and were less likely to speak English or French as their first language (32 [50.8%] vs 136 [92.6%]; P < .001). Of those residents who identified as a VM, the most common race/ethnicity groups were Asian (45 [71.4%]) and African or Caribbean (6 [9.5%]) (Table 3).
Theme 1 was defined by this open-ended response: “I generally have a more formal relationship with my staff than the male residents… this I do feel contributes to women losing out on opportunities such as networking, research initiatives, and actual operative training.” Most residents (173 of 210 [82.4%]) suggested that fostering diversity would increase the overall strength of surgical training programs and that programs should strive to increase diversity among staff general surgeons. Most respondents (164 [78.1%]) acknowledged that a diverse workforce was important for both resident training and patient care. Residents agreed or strongly agreed that their cohort was diverse in age (140 [66.7%]), sex (185 [88.1%]), training (133 [63.3%]), and race/ethnicity (143 [68.1%]) (Figure, A). Although residents agreed or strongly agreed that their staff was diverse in age (178 [84.8%]), sex (148 [70.5%]), and training background (124 [59.0%]), they disagreed or strongly disagreed that staff was diverse in race/ethnicity (76 [36.2%]) or sexual orientation (71 [33.8%]) (Figure, B).
Gender and VM status were factors in perceived quality of interpersonal relationships. Although most residents reported having a collegial relationship with staff, female residents with a VM status were less likely than male residents who did not identify as a VM to agree or strongly agree with this statement (21 [63.6%] vs 61 [89.7%]; P = .01), to feel like they fit in with their training programs (21 [63.6%] vs 56 [82.3%]; P = .003), and to feel valued at work (15 [45.4%] vs 47 [69.1%]; P = .03) (Table 4; eTable in the Supplement).
Similarly, gender and VM status were also factors in resiliency. More than one-third of residents (86 of 210 [41.0%]) revealed that they thought about quitting residency. Overall, 56 women (50.0%) reported thinking about quitting residency at least sometimes compared with only 20 men (29.4%) who did not identify as a VM. Women with a VM status were more likely to “at least usually” consider quitting than their male counterparts (5 [15.2%] vs 1 [1.5%]; P < .001).
Theme 2 was reported in statements like the following: “There lacked much support in the form of mentors or role models (in particular young female role models).” Female residents vs their male peers agreed or strongly agreed with the importance of having role models of the same gender (99 [88.4%] vs 48 [49.0%]; P < .001), and residents who identified as a VM vs those who did not identify as a VM agreed or strongly agreed with the value of having role models of the same race/ethnicity (33 [52.3%] vs 20 [13.6%]; P < .001) (Table 4). Yet, more than 1 in 4 female respondents compared with male respondents strongly disagreed or disagreed with having such a mentor within their institution (33 [29.5%] vs 19 [19.4%]; P < .001) (Table 4), and those who identified as a VM were less likely to report having this type of mentorship than their peers without a VM status (36 [57.1%] vs 21 [14.2%] P < .001). Women with a VM status were found to be particularly disadvantaged, with nearly 1 in 2 women strongly disagreeing or disagreeing with having a mentor to help them navigate through residency.
Theme 3 was suggested in the following response: “…occasionally I feel that I and others of my gender have to work harder to receive equal recognition as the opposite gender.” Female residents were more likely than their male colleagues to strongly agree or agree with worrying about receiving fewer opportunities because of their gender (54 [48.2%] vs 3 [3.0%]; P < .001) (Table 4). Women were also less likely than men to agree or strongly agree with receiving equal surgical training opportunities (62 [55.4%] vs 80 [81.7%]; P < .001). Similarly, residents who identified as a VM were less likely than peers without VM status to agree or strongly agree with perceiving equal surgical training opportunities (45 [71.4%] vs 127 [86.4%]; P = .049) and were more likely to worry about receiving fewer work opportunities because of their race/ethnicity (20 [31.7%] vs 2 [1.4%]; P < .001). Women (2 [6.0%] vs 0 [0%]; P < .001) and women without VM status (1 [1.3%] vs 0 [0%]; P < .045) were also more likely than men who did not identify as a VM to agree or strongly agree with perceiving fewer opportunities at work because of religion.
Theme 4 was described in the following response: “Patients… trust the opinion of my male medical students over my own.” Female residents reported that patients often dismissed their medical expertise because of their gender, and those with a VM status were dismissed because of their race/ethnicity (Table 4). A total of 98% of male residents reported never experiencing dismissal of their medical expertise by patients because of their gender. In contrast, 91 of 112 female residents (81.2%) reported that their expertise was dismissed at least once annually, with 37 women (33.0%) experiencing dismissal of their expertise at least once every week (P < .001). Furthermore, female residents reported being called “doctor” less frequently than other residents compared with their male colleagues (82 [73.2%] vs 3 [3.1%]; P < .001). Women also identified that they were more likely than their peers to be confused for a nonphysician or allied health worker at least monthly by both patients (108 [96.4%] vs 20 [20.4%]; P < .001) and hospital staff (76 [67.9%] vs 8 [8.2%]; P < .001), an event that occurred daily for some female residents (Table 4). Similarly, residents with VM status reported at least monthly dismissal of their expertise because of their race/ethnicity compared with residents who did not identify as a VM (9 [14.3%] vs 1 [0.7%]; P < .001).
Theme 5 was illustrated by the following open-ended statement: “I’ve been told … that a female resident in another program was a ‘really good resident’ but then ‘dumb enough to get pregnant’ in residency.” Reporting of discrimination was a concern for all residents. A total of 90 residents (42.9%) did not feel confident that a trusted authority within their program would take steps to prevent an incident of discrimination from reoccurring after it was reported. Fifty-four residents (25.7%) were not confident that they could safely speak to an authority figure about discrimination if they experienced it at work. Women residents who identified as a VM were particularly disadvantaged compared with their male peers without VM status, because they strongly disagreed or disagreed that someone was available to talk to about experienced discrimination (7 [21.3%] vs 2 [2.9%]; P < .02) and that action would be taken to prevent a recurrence of reported incidences (16 [48.5%] vs 6 [8.8%]; P < .001).
Taken together, gender and racial/ethnic biases appeared to have implications for resident experiences during surgical training. Understanding the role of these biases in resident training, although complex and multifactorial, may help to guide the development of strategies for overcoming the perceived barriers that present during surgical training. For example, the inequalities in the perception of competence theme were predominantly associated with gender bias and thus disproportionately affected female residents. Overcoming this barrier would likely necessitate gender-focused strategies. All other identified themes (interpersonal relationships, resiliency, mentorship opportunities, training opportunities, and perceived support and discrimination) demonstrated the intersectionality between gender and race/ethnicity. Reported experiences were associated with both gender and VM status, and the perceived surgical training experiences were particularly unfavorable for women who self-identified as a VM. For these themes, simply addressing gender and racial/ethnic biases in isolation is unlikely to eliminate the current barriers without also implementing strategies focused on the intersectionality of gender and race/ethnicity.
Multifaceted strategies are required from training programs, accrediting bodies, and the medical community to overcome these barriers and improve the surgical training experience. The first, and most important, is the acknowledgment of the presence of these biases and barriers. They are prevalent across all resident training levels and training institutions, and they are not isolated to experiences within certain programs. Not acknowledging the existence of these barriers or the resident groups at risk may diminish efforts to optimize resident training. To address this knowledge gap, we have presented information at the national Canadian Surgery Forum, provincial research conferences, and local grand rounds.
Second, strong mentorship networks must be established and the available pool of role models must be enhanced. We advocate for implementing such strategies at the local, national, and international levels. One example includes developing a mentorship network designed to foster medical student diversity by connecting students with local resident mentors. Nationally, networking sessions intended to connect medical students with program directors across the country should also be implemented. Third, transparency must be demonstrated and emphasized through merit-based recruitment and hiring practices. Developing standardized applicant assessment tools along with a diverse panel of evaluators may promote a more transparent selection process. Fourth, all residents, regardless of gender or VM status, should be provided equal opportunities for parental leave, childcare, and career advancement. Together, these strategies underline the fundamental goal of promoting a culture of equality, a term used by a number of respondents. This culture should be one that is based on justice and transparency, and is free of discrimination from gender and racial/ethnic biases.
A paucity of literature exists regarding the implication of diversity for resident training experiences. In 2008, Wong et al36 first explored the association of race/ethnicity with training in a national surgical residency cohort. A cross-sectional national survey of 4339 surgical residents revealed that residents in racial/ethnic minority groups were less likely to feel like they fit in and had less positive relationships with both staff and peers.36 Most recently, Hu et al35 completed a cross-sectional national survey of general surgery resident burnout and mistreatment in the US. This survey involved 7409 residents and demonstrated that mistreatment was prevalent among general surgery residents, was more common among women, and was associated with burnout and suicidal thoughts. However, Hu et al35 did not collect information on the VM status or race/ethnicity of survey respondents, which limited their assessment of the factors associated with resident mistreatment. The findings of the present survey study are in keeping with the results of these previous studies, reinforcing that female residents encounter biases that have an adverse implication for their residency experience. In addition, this study builds on the literature by providing a comprehensive analysis of biases categorized by themes that identify a previously unexplored nuanced intersectionality between gender and VM status.
Systemic barriers associated with gender and racial/ethnic biases exist and appear to have implications for the training experience of general surgery residents in Canada. We identified the concept of intersectionality for 2 particularly at-risk groups, female residents and female residents who identified as a VM, for which the barriers appeared to be compounded. We believe that the results of this study could lead to understanding that the diversity among a surgical resident cohort is associated with different experiences during surgical training. Future studies are warranted to evaluate these experiences across other Canadian surgical cohorts and to evaluate their implications for patient care.
This study has a number of limitations. Although a reasonable response rate of 40.5% was achieved, nonresponse bias may be a substantial limitation. Because the survey platform used (Google Forms) is not encrypted, the confidentiality of respondents was not protected, which may have altered the results. The survey platform also did not allow the collection of partial interviews, refusals, and survey breakoff, restricting the assessment of refusal or cooperation rates. It is possible that nonresponders differed considerably from responders in demographics or in perception of training experience. However, when comparing the survey demographics with data from all matched surgical applicants from 2015 to 2019, we found that the proportion of female respondents in both groups was identical (53%). Although we did not have responses from all residents, we identified that gender and racial/ethnic biases were experienced by a substantial proportion of the resident cohort even if all nonresponders did not have these experiences.
In addition, the study was susceptible to voluntary response bias or overrepresentation of residents with particularly strong experiences with diversity during surgical training. Furthermore, although we collected a breadth of data from the resident cohort, we omitted sensitive data with fewer than 3 resident respondents to limit potential identifiable information. We acknowledge that including these data may provide for a more comprehensive understanding of resident diversity, but the overall data from such small samples would ultimately be difficult to interpret and would not be generalizable. Despite these limitations, the consistency and significance of the themes we identified from the national cohort suggest that these findings have broad implications.
Gender and racial/ethnic biases appeared to be associated with substantial barriers presented during the training of general surgery residents. We believe that this study serves as a call to training programs, accrediting bodies, and the medical community to develop strategies with a focus on the intersectionality of gender and race/ethnicity to improve the training experience of at-risk residents.
Accepted for Publication: April 17, 2020.
Corresponding Author: Valentin Mocanu, MD, Department of Surgery, Faculty of Medicine and Dentistry, University of Alberta Hospital, 8440 112 St NW, Edmonton, AB T6G 2B7, Canada (firstname.lastname@example.org).
Published Online: August 12, 2020. doi:10.1001/jamasurg.2020.2828
Author Contributions: Drs Mocanu and Kuper had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Mocanu, Kuper, Marini, Assane, DeGirolamo, Fathimani.
Acquisition, analysis, or interpretation of data: Mocanu, Kuper, DeGirolamo, Baxter.
Drafting of the manuscript: Mocanu, Kuper, Marini, Assane, DeGirolamo.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Mocanu, Kuper.
Administrative, technical, or material support: Marini, Assane, DeGirolamo.
Supervision: Marini, Fathimani, Baxter.
Conflict of Interest Disclosures: None reported.
Create a personal account or sign in to: