Abstract
This cohort study surveyed resident surgeons in the US to assess their experiences of racial/ethnic discrimination in residency programs.
Discrimination in medicine has been associated with decreased productivity, as well as increased alcohol use, depression, attrition, and suicidality among physicians.1,2 In surgical training, discrimination is common2 but has not been comprehensively evaluated among racial/ethnic minorities. The objectives of this study were to (1) determine the national prevalence and sources of discrimination based on race/ethnicity in US general surgery programs, (2) identify factors associated with discrimination, and (3) assess its association with resident wellness.
Methods
Resident physicians training in Accreditation Council for Graduate Medical Education–accredited general surgery programs were administered a survey following the 2019 American Board of Surgery In-Training Examination. Residents were asked about their experiences with various types of discriminatory behavior based on race/ethnicity or religion3,4 within that academic year. Burnout, thoughts of attrition, and suicidality were assessed with established instruments.2 The proportion of minority faculty members within each program was obtained from the Association of American Medical Colleges. This study was reviewed by the Northwestern University institutional review board office and was determined to not meet the definition of human-subjects research. As a result, this study was deemed exempt from full review and informed consent procedures.
Descriptive statistics were calculated. A multivariable regression model was developed to examine resident and program characteristics associated with discrimination. Adjusted analyses were repeated with stratification by sex to evaluate for potential interactions between race and sex. We performed χ2 tests to assess the associations of discrimination with burnout, thoughts of attrition, and suicidality. All tests were 2-sided with α = .05, using Stata version 15.1 (StataCorp). Data were collected in January 2019. The dates that data were analyzed include June 2019 to August 2019.
Results
A total of 6956 clinically active residents from 301 programs completed the survey (response rate, 85.6%). Of the 5679 who responded to the relevant questions, 1346 (23.7%) reported experiencing discrimination based on race/ethnicity or religion. Discrimination rates were higher in black respondents (171 of 242 [70.7%]), Asian respondents (442 of 963 [45.9%]), Hispanic respondents (122 of 482 [25.3%]), and other nonwhite respondents (175 of 526 [33.3%]) compared with white respondents (435 of 3455 [12.6%]). The most common discriminatory behavior was being mistaken for another person of the same race, experienced by 135 of 240 black residents (56.3%; 2 individuals did not respond to this question) and 361 of 963 Asian residents (37.6%; 4 individuals did not respond), with nurses and staff as the most common source (413 [43.8%]). Black residents frequently reported being mistaken for nonphysicians (151 of 242 [62.4%]; with 327 residents [73.2%] reporting this behavior), as well as experiencing different standards of evaluation (92 of 240 [38.3%]; with 243 residents [63.0%] reporting this behavior). Slurs/hurtful comments were experienced by 60 of 241 black residents (24.9%), most commonly from patients/families (126 [35.5%]; Table 1).
Table 1. Prevalence and Most Common Sources of Discrimination Based on Race/Ethnicity or Religiona.
| Characteristic | Respondents, No. (%) | Most common source of discriminationd | |||||||
|---|---|---|---|---|---|---|---|---|---|
| All (N = 5679)b | White (n = 3455) | Black (n = 242) | Hispanic (n = 482) | Asian (n = 963) | Other/prefer not to say (n = 526) | P valuec | Source | Respondents reporting this type of discrimination, No. (%) | |
| Overall prevalence | 1346 (23.7) | 435 (12.6) | 171 (70.7) | 122 (25.3) | 442 (45.9) | 175 (33.3) | <.001 | NA | NA |
| Discrimination components | |||||||||
| Different standards of evaluation | 468 (8.2) | 100 (2.9) | 92 (38.0) | 52 (10.8) | 137 (14.2) | 86 (16.3) | <.001 | Attending physicians | 243 (63.0) |
| Denied opportunities | 250 (4.4) | 69 (2.0) | 39 (16.1) | 27 (5.6) | 59 (6.1) | 55 (10.5) | <.001 | Attending physicians | 138 (67.3) |
| Mistaken for a nonphysician | 482 (8.5) | 51 (1.5) | 151 (62.4) | 66 (13.7) | 150 (15.6) | 63 (12.0) | <.001 | Patients and their families | 327 (73.2) |
| Slurs and/or hurtful comments | 416 (7.3) | 116 (3.4) | 60 (24.8) | 40 (8.3) | 129 (13.4) | 70 (13.3) | <.001 | Patients and their families | 126 (35.5) |
| Socially isolated | 208 (3.7) | 65 (1.9) | 28 (11.6) | 26 (5.4) | 37 (3.8) | 51 (9.7) | <.001 | Colleagues | 117 (70.1) |
| Mistaken for another person of the same race | 998 (17.6) | 300 (8.7) | 135 (55.8) | 74 (15.4) | 361 (37.5) | 127 (24.2) | <.001 | Nurses/staff | 413 (43.8) |
Abbreviation: NA, not applicable.
Eleven observations were missing data on race/ethnicity.
The denominator excludes those who did not respond to the discrimination question.
Two-tailed χ2 tests.
Based on those residents who reported experiencing discrimination based on race/ ethnicity or religion.
Residents were more likely to report discrimination if female (odds ratio [OR], 1.48 [95% CI, 1.27-1.74]; P < .001), nonwhite (Black: OR, 20.91 [95% CI, 14.39-30.38]; P < .001; Hispanic: OR, 2.62 [95% CI, 1.99-3.47]; P < .001; Asian: OR, 6.29 [95% CI, 5.18-7.63]; P < .001), more senior (program year 2 or 3: OR, 1.30 [95% CI, 1.08-1.57]; P = .005; program year 4 or 5: OR, 1.57 [95% CI, 1.28-1.92]; P < .001; compared with program year 1), and after experiencing violations of the 80-hour duty limit (>5 times in 6 months: OR, 2.26 [95% CI, 1.69-3.02]; P < .001; vs never; Table 2). Stratification by sex revealed higher odds in women for every nonwhite race/ethnicity (black men: OR, 18.6 [95% CI, 11.44-30.32] vs black women: OR, 23.93 [95% CI, 13.93-41.12]; Hispanic men: OR, 2.32 [95% CI, 1.57-3.42] vs Hispanic women: 3.09 [95% CI, 2.06-4.64]; Asian men: OR, 5.66 [95% CI, 4.36-7.34] vs Asian women: 7.35 [95% CI, 5.57-9.72]). Geographic location, program type (academic, community, or military), and the racial/ethnic compositions of the faculty and residency members were not associated with the likelihood of experiencing discrimination. Residents who experienced discrimination reported higher rates of burnout (51.6% vs 40.0%; P < .001), thoughts of attrition (16.2% vs 10.1%; P < .001), and suicidal thoughts (6.5% vs 3.8%; P < .001).
Table 2. Association of Resident and Program Characteristics With Racial/Ethnic and Religious Discriminationa.
| Characteristic | Respondents, No./total No. (%) | Odds ratio (95% CI) | P value |
|---|---|---|---|
| Resident characteristics | |||
| Sex | |||
| Male | 661/3241 (20.4) | 1 [Reference] | |
| Female | 648/2277 (28.5) | 1.48 (1.27-1.74) | <.001 |
| Race/ethnicity | |||
| Non-Hispanic white | 428/3380 (12.7) | 1 [Reference] | |
| Non-Hispanic black | 167/231 (72.3) | 20.91 (14.39-30.38) | <.001 |
| Hispanic | 115/460 (25.0) | 2.62 (1.99-3.47) | <.001 |
| Asian | 431/940 (45.9) | 6.29 (5.18-7.63) | <.001 |
| Other/prefer not to say | 168/507 (33.1) | 3.89 (3.09-4.91) | <.001 |
| Clinical postgraduate year | |||
| Intern, year 1 | 268/1263 (21.2) | 1 [Reference] | |
| Junior, year 2-3 | 521/2235 (23.3) | 1.30 (1.08-1.57) | .005 |
| Senior, year 4-5 | 520/2020 (25.7) | 1.57 (1.28-1.92) | <.001 |
| Marital status | |||
| Married | 525/2401 (21.9) | 1 [Reference] | |
| Relationship | 403/1742 (23.1) | 0.86 (0.73-1.02) | .09 |
| No relationship | 357/1270 (28.1) | 0.93 (0.78-1.11) | .42 |
| Divorced/widowed | 24/105 (22.9) | 0.92 (0.54-1.56) | .75 |
| American Board of Surgery In-Training Examination score quartiles | |||
| 1-3 | 1035/4144 (25.0) | 1 [Reference] | |
| 4 (highest) | 274/1374 (19.9) | 0.92 (0.73-1.10) | .34 |
| Violations of 80-h duty limit in last 6 mo | |||
| Never | 680/3428 (19.8) | 1 [Reference] | |
| 1-2 times | 372/1359 (27.4) | 1.51 (1.28-1.78) | <.001 |
| 3-4 times | 159/452 (35.2) | 2.48 (2.01-3.06) | <.001 |
| >5 times | 88/244 (36.1) | 2.26 (1.69-3.02) | <.001 |
| Program characteristics | |||
| Quartile of program size by No. of residents | |||
| 1 (6-25) | 317/1456 (21.8) | 1 [Reference] | |
| 2 (26-37) | 330/1408 (23.4) | 1.01 (0.80-1.30) | .91 |
| 3 (38-51) | 323/1397 (23.1) | 0.94 (0.71-1.23) | .64 |
| 4 (52-81) | 339/1257 (27.0) | 1.02 (0.76-1.37) | .88 |
| Program type | |||
| Academic | 774/3158 (24.5) | 1 [Reference] | |
| Community | 507/2199 (23.1) | 1.08 (0.94-1.53) | .39 |
| Military | 28/161 (17.4) | 1.56 (0.92-2.66) | .10 |
| Program location | |||
| Northeast | 475/1807 (26.3) | 1.20 (0.94-1.53) | .14 |
| Southeast | 244/1110 (22.0) | 1 [Reference] | |
| Midwest | 257/1233 (20.8) | 0.89 (0.72-1.19) | .55 |
| Southwest | 139/632 (22.0) | 1.03 (0.77-1.33) | .92 |
| West | 194/736 (26.4) | 1.23 (0.94-1.62) | .13 |
| Quartile of percentage of nonwhite faculty | |||
| 1 (0%-27%) | 296/1291 (22.9) | 1 [Reference] | |
| 2 (28%-33%) | 272/1136 (23.9) | 1.03 (0.82-1.30) | .77 |
| 3 (34%-41%) | 300/1247 (24.1) | 0.86 (0.68-1.06) | .16 |
| 4 (>41%) | 309/1096 (28.2) | 0.84 (0.65-1.08) | .18 |
| Quartile of percentage of nonwhite residents | |||
| 1 (6%-25%) | 263/1354 (19.4) | 1 [Reference] | |
| 2 (26%-37%) | 275/1242 (22.3) | 1.01 (0.79-1.27) | .97 |
| 3 (38%-46%) | 333/1237 (26.7) | 1.18 (0.91-1.53) | .20 |
| 4 (>46%) | 361/1198 (30.1) | 1.13 (0.88-1.44) | .32 |
A total of 161 observations with missing data were excluded from the multivariable logistic regression model: race (11), sex (131), and relationship status (19).
Discussion
Racial/ethnic discrimination is experienced by a large proportion of nonwhite residents in general surgery training and causes substantial distress. Because discrimination originates from multiple sources, strategies for mitigating it must be multifaceted and context specific, but they may include promoting a culture of zero tolerance, empowering trainees to report discriminatory behaviors, and training residents, faculty, and staff to recognize and respond appropriately to discrimination. We found no program characteristics significantly associated with discrimination, including the proportion of faculty members and/or residents who are nonwhite. It may be that few programs have reached the critical mass needed to overturn social conventions,5 adding to the concern that minority representation in surgery is decreasing.6 Limitations of this study include the inability to account for unmeasured trainee, hospital, and community factors (eg, patient population, social norms). These findings suggest that concerted efforts are needed to improve diversity, equity, and inclusion within surgical training programs.
References
- 1.Fnais N, Soobiah C, Chen MH, et al. . Harassment and discrimination in medical training: a systematic review and meta-analysis. Acad Med. 2014;89(5):817-827. doi: 10.1097/ACM.0000000000000200 [DOI] [PubMed] [Google Scholar]
- 2.Hu YY, Ellis RJ, Hewitt DB, et al. . Discrimination, abuse, harassment, and burnout in surgical residency training. N Engl J Med. 2019;381(18):1741-1752. doi: 10.1056/NEJMsa1903759 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Baldwin DC Jr, Daugherty SR, Rowley BD. Racial and ethnic discrimination during residency: results of a national survey. Acad Med. 1994;69(10)(suppl):S19-S21. doi: 10.1097/00001888-199410000-00029 [DOI] [PubMed] [Google Scholar]
- 4.Liebschutz JM, Darko GO, Finley EP, Cawse JM, Bharel M, Orlander JD. In the minority: black physicians in residency and their experiences. J Natl Med Assoc. 2006;98(9):1441-1448. [PMC free article] [PubMed] [Google Scholar]
- 5.Centola D, Becker J, Brackbill D, Baronchelli A. Experimental evidence for tipping points in social convention. Science. 2018;360(6393):1116-1119. doi: 10.1126/science.aas8827 [DOI] [PubMed] [Google Scholar]
- 6.Lett LA, Orji WU, Sebro R. Declining racial and ethnic representation in clinical academic medicine: A longitudinal study of 16 US medical specialties. PLoS One. 2018;13(11):e0207274. doi: 10.1371/journal.pone.0207274 [DOI] [PMC free article] [PubMed] [Google Scholar]