This study examines the association between medical student race/ethnicity and induction into the Alpha Omega Alpha honor society.
Key Points
Question
Are minority medical students less likely than white medical students to be members of the Alpha Omega Alpha honor society?
Findings
In this cohort study of 4655 US medical students, Alpha Omega Alpha membership for white students was nearly 6 times greater than that for black students and nearly 2 times greater than for Asian students, both significant differences.
Meaning
The selection process for Alpha Omega Alpha membership may be vulnerable to bias, which may affect future opportunities for minority medical students.
Abstract
Importance
Previous studies have found racial and ethnic inequities in the receipt of academic awards, such as promotions and National Institutes of Health research funding, among academic medical center faculty. Few data exist about similar racial/ethnic disparities at the level of undergraduate medical education.
Objective
To examine the association between medical student race/ethnicity and induction into the Alpha Omega Alpha (AΩA) honor society.
Design, Setting, and Participants
This study analyzed data from the Electronic Residency Application Service, the official service used by US medical students to apply to residency programs. A total of 4655 US medical students from 123 allopathic US medical schools who applied to 12 distinct residency programs associated with one academic health center in the 2014 to 2015 academic year were studied.
Main Outcomes and Measures
Membership in the AΩA society among black, white, Hispanic, and Asian medical students.
Results
A total of 4655 unique applications were analyzed in the study (median age, 26 years; 2133 women [45.8%]). Overall, self-reported race/ethnicity in our sample was 2605 (56.0%) white (691 [71.5%] of AΩA applicants were white), 276 (5.9%) black (7 [0.7%] AΩA), 186 (4.0%) Hispanic (27 [2.8%] AΩA), and 1170 (25.1%) Asian (168 [17.4%] AΩA). After controlling for US Medical Licensing Examination Step 1 scores, research productivity, community service, leadership activity, and Gold Humanism membership, the study found that black (adjusted odds ratio [aOR], 0.16; 95% CI, 0.07-0.37) and Asian (aOR, 0.52; 95% CI, 0.42-0.65) medical students remained less likely to be AΩA members than white medical students. No statistically significant difference was found in AΩA membership between white and Hispanic medical students (aOR, 0.79; 99% CI, 0.45-1.37) in the adjusted model.
Conclusions and Relevance
Black and Asian medical students were less likely than their white counterparts to be members of AΩA, which may reflect bias in selection. In turn, AΩA membership selection may affect future opportunities for minority medical students.
Introduction
Numerous studies have found evidence of racial/ethnic inequities in the receipt of academic rewards in medicine. Black and Hispanic faculty members are less likely to be promoted in academic health centers, and black and Asian faculty members are less likely to be awarded National Institutes of Health (NIH) research funding than are white faculty, suggesting that processes that require both subjective and objective measures of evaluation are vulnerable to bias. Although this phenomenon has been explored among faculty, there has been limited research regarding the association between race/ethnicity and the receipt of academic awards at the level of undergraduate medical education. Examining the Alpha Omega Alpha (AΩA) medical honor society membership by race/ethnicity presents an opportunity to explore whether a similar experience exists at the level of undergraduate medical education.
Induction into the AΩA honor society is associated with future success in academic medicine. The medical society boasts more than 150 000 inductees since its founding in 1902, including 11 of the 19 US Surgeons General and more than 50 Nobel laureates. Multiple studies have found that members of AΩA are more likely to match into the residency specialty of their choice, especially among the surgical subspecialties. Literature has also found an association between AΩA membership and the likelihood of a physician choosing a career in academic medicine and attaining the rank of full professor, dean, or departmental chair.
Although each AΩA chapter has specific criteria for the selection of members, the national AΩA society has 3 guidelines to which all chapters must adhere: (1) only students in the top quartile of their medical school judged by academic performance are eligible for AΩA membership, (2) each chapter can select only up to 16% of medical students to ultimately be members of AΩA, and (3) the students elected into AΩA are chosen “not just for their high academic standing, but as well for leadership among their peers, professionalism and a firm sense of ethics, promise of future success in medicine, and a commitment to service in the school and community.”(p 1) A committee, at the level of the individual medical school, ultimately chooses which medical students will be AΩA members.
Despite the importance placed on AΩA membership, little has been reported about medical student race/ethnicity and likelihood of induction into the honor society. In a national study of AΩA in 1989, Babbott et al reported that AΩA members were disproportionately white, but the investigators did not control for additional factors that may influence AΩA selection, such as leadership activity, community service, or academic achievement.
In this study, we examine the association between race/ethnicity and the likelihood of AΩA membership. We report on a range of medical student characteristics associated with election into AΩA by using data from medical students graduating from 123 US medical schools who applied to 12 distinct residency programs at 1 academic health center in the 2014 to 2015 academic year.
Methods
Study Design and Population
We conducted a retrospective cohort study of medical student Electronic Residency Application Service (ERAS) applications from 123 US allopathic medical schools submitted to 12 distinct categorical residency programs (anesthesia, internal medicine, neurosurgery, obstetrics and gynecology, orthopedic surgery, pediatrics, plastic surgery, psychiatry, radiology, general surgery, and thoracic and vascular surgery) at a single academic health center (Yale Medical Center, New Haven, Connecticut) during the 2014 to 2015 application cycle. The Yale Human Investigations Committee deemed this study exempt from review; therefore, no informed consent was required.
Study Protocol
We used Matlab (Mathworks) to extract data from the ERAS applications, including applicants’ self-reported race/ethnicity, sex, age, medical school, US Medical Licensing Examination (USMLE) Step 1 and Step 2 scores, total number of research publications, self-reported community service hours, self-reported time dedicated to leadership activities, additional degrees (such as master’s degree or PhD), Gold Humanism honor society membership, and AΩA membership. Duplicate ERAS applications from individuals applying to multiple residencies were excluded. Students from historically black medical schools were excluded because the likelihood of AΩA members being minorities is higher in these institutions than nonhistorically black medical schools. Applicants from schools without an AΩA chapter and applicants who did not self-identify race/ethnicity were excluded. Medical students can be elected to AΩA in their junior year or senior year of medical school. We did not make a distinction between junior and senior AΩA inductees in this study.
Statistical Analysis
Outcome data from the ERAS applications were transferred from the Matlab software to a separate electronic spreadsheet (Microsoft Excel, Microsoft Corporation). This spreadsheet was then imported into STATA, and all analyses were performed using STATA, version 14 (StataCorp).
We compared differences in demographic characteristics between AΩA and non-AΩA applicants using 2-tailed, unpaired t tests for continuous variables, Mann-Whitney tests for variables not normally distributed, and χ2 tests for categorical variables. We then used logistic regression to model the unadjusted effect of race/ethnicity on the likelihood of AΩA membership. To account for clustering within each medical school, we used generalized estimating equations with compound symmetry correlation to model the covariance structure. The national AΩA society advocates that professionalism and a firm sense of ethics, leadership, and community service be used as criteria for honor society selection. In our analysis, we added the variables self-reported total hours dedicated to community service, acceptance into Gold Humanism (representing applicants recognized for professionalism and a firm sense of ethics), and self-reported total hours of participation in leadership activities while in medical school as proxies for these criteria. We also included the applicant’s sex into this analysis to model the association of an applicant’s sex on AΩA membership.
After this analysis, we added predetermined variables that could potentially account for differences in AΩA selection, including USMLE Step 1 scores, and variables to designate whether the applicant authored an article in a medical journal or presented an oral abstract at a medical conference, had a master’s degree, or had a PhD. We refer to this analysis as the fully adjusted model. Odds ratios (ORs) and 95% CIs were calculated to measure the strength of observed associations. The ORs, derived from multivariable logistic regression modeling, were converted to estimated prevalence ratios (PRs) for the fully adjusted model given that the prevalence of AΩA membership is greater than 10% and to facilitate interpretation of results, using standard methods. Akaike information criterion analyses were performed to assess the goodness of fit for the 3 models. We also conducted an analysis to determine whether there was an interaction between an applicant’s race/ethnicity and USMLE Step 1 scores and the likelihood of AΩA membership.
Results
A total of 5239 unique ERAS applications were reviewed. Of these applications, 319 (6.1%) were excluded from schools without an AΩA chapter, 61 (1.2%) were excluded from historically black medical schools, and 204 (3.9%) were excluded because race/ethnicity was unknown. A total of 4655 unique applications were analyzed in the study. These applications came from 123 US medical schools or 96% of all US medical schools with an AΩA chapter. Applicants included in the study were similar to those excluded in terms of age, sex, and USMLE Step 1 and Step 2 scores with the exception of applicants from historically black medical schools who had mean lower USMLE Step 1 and Step 2 scores.
The median age of applicants was 26 years (range, 21-54 years), and 2133 (45.8%) of applicants were women (Table 1). Of all the applicants, 276 (5.9%) were black, 186 (4.0%) Hispanic, 1170 (25.1%) Asian, and 2605 (56.0%) white. The mean (SD) USMLE Step 1 score was 235 (18.5), and the mean (SD) USMLE Step 2 score was 245 (16.7). A total of 1626 applicants (34.9%) attended a school ranked in the top 40 by NIH funding. Among the sample, 966 applicants (20.8%) were elected into AΩA. Compared with all the medical students who applied to residency in the United States in 2014, the cohort of applicants reviewed was similar with respect to sex and race/ethnicity (Table 1). The cohort reviewed was more likely to have membership in AΩA and have a higher mean USMLE Step 1 score than the population of medical students applying to residency (Table 1).
Table 1. Characteristics of Overall US Senior Medical Student Population and the Study Cohorta.
| Characteristic | Overall US Seniorsb (n = 18 349) |
Study Cohort (n = 4655) |
|---|---|---|
| Female | 8725 (47.6) | 2133 (45.8) |
| Race/ethnicity | ||
| Black | 1061 (5.8) | 276 (5.9) |
| Hispanic | 865 (4.7) | 186 (4.0) |
| Asian | 3710 (20.2) | 1170 (25.1) |
| White | 11 012 (60.0) | 2605 (56.0) |
| USMLE Step 1 score, mean (SD)c,d | 229 (NA) | 235 (18.5) |
| USMLE Step 2 score, mean (SD)c,d | 242 (NA) | 245 (16.7) |
| AΩA membershipc,d | 2492 (15.2) | 966 (20.8) |
| Top 40 medical school (NIH funding)c,d | 5214 (31.8) | 1626 (34.9) |
| PhDd | 622 (3.9) | 166 (3.6) |
Abbreviations: AΩA, Alpha Omega Alpha; NA, not available; NIH, National Institutes of Health; USMLE, US Medical Licensing Examination.
Data are presented as number (percentage) of applicants unless otherwise indicated.
Data are from Association of American Medical Colleges.
Mean USMLE Step 1 and 2 scores and proportion of students in AΩA and from top 40 medical schools are statistically different for overall US seniors and the study cohort (P < .001).
Data are from National Resident Matching Progam.
Overall, self-reported race/ethnicity in our sample was 2605 (56.0%) white (691 [71.5%] of AΩA applicants were white), 276 (5.9%) black (7 [0.7%] AΩA), 186 (4.0%) Hispanic (27 [2.8%] AΩA), and 1170 (25.1%) Asian (168 [17.4%] AΩA). The USMLE Step1 and USMLE Step 2 scores were significantly higher for AΩA members (mean [SD], 251 [10.7] vs 230 [17.8] for USMLE Step 1 and 259 [10.8] vs 241 [16.0] for USMLE Step 2; P < .001). The AΩA members were more likely to be members of the Gold Humanism honor society (180 [18.6%] vs 313 [8.5%]; P < .001). No statistically significant differences were found in the median time dedicated to leadership activities and community service between AΩA and non-AΩA students (Table 2).
Table 2. Characteristics of AΩA and Non-AΩA Applicantsa.
| Characteristic | AΩA (n = 966) |
Non-AΩA (n = 3689) |
P Value |
|---|---|---|---|
| Race/ethnicityb | |||
| White | 691 (71.5) | 1914 (51.9) | <.001 |
| Black | 7 (0.7) | 269 (7.3) | <.001 |
| Hispanic | 27 (2.8) | 159 (4.3) | .03 |
| Asian or Pacific Islander | 168 (17.4) | 1002 (27.2) | <.001 |
| Multiracial | 58 (6.0) | 260 (7.0) | .25 |
| Other | 15 (1.6) | 85 (2.3) | .15 |
| Female | 404 (41.8) | 1729 (46.9) | .001 |
| Median age, y | 26 | 27 | <.001 |
| USMLE Step 1 | |||
| Mean | 251 | 230 | <.001 |
| Bottom quartilec | 19 (2.0) | 1159 (31.4) | |
| Second quartile | 77 (8.0) | 1135 (30.8) | |
| Third quartile | 274 (28.4) | 861 (23.3) | |
| Top quartile | 596 (61.7) | 534 (14.5) | |
| USMLE Step 2d | |||
| Mean | 259 | 241 | <.001 |
| Bottom quartilec | 1 (0.1) | 149 (5.2) | |
| Second quartile | 45 (6.0) | 1190 (41.3) | |
| Third quartile | 197 (26.1) | 983 (34.1) | |
| Top quartile | 513 (68.0) | 561 (19.5) | |
| Master’s degree | 125 (12.9) | 690 (18.7) | <.001 |
| PhD | 21 (2.2) | 145 (3.9) | .008 |
| Published article or presented abstract (yes/no) | 546 (56.5) | 1942 (52.6) | .03 |
| AΩA selection proxies | |||
| Gold Humanism member (yes/no) | 180 (18.6) | 313 (8.5) | <.001 |
| Leadership hours | |||
| Median (IQR) | 2 (0-287.5) | 0 (0-312) | .19 |
| <50th percentile | 465 (48.1) | 1931 (52.3) | |
| 50th-75th percentile | 264 (27.3) | 832 (22.6) | |
| >75th percentile | 237 (24.5) | 926 (25.1) | |
| Community service hours | |||
| Median (IQR) | 583 (0-1542) | 540 (0-1689) | .93 |
| <50th percentile | 474 (49.1) | 1854 (50.3) | |
| 50th-75th percentile | 266 (27.5) | 898 (24.3) | |
| >75th percentile | 226 (23.4) | 937 (25.4) |
Abbreviations: AΩA, Alpha Omega Alpha; IQR, interquartile range; USMLE, US Medical Licensing Examination.
Data are presented as number (percentage) of applicants unless otherwise indicated.
Overall χ2 for the correlation between race/ethnicity and AΩA membership is P < .001.
Overall χ2 for the correlation between AΩA membership and USMLE Step 1 and Step 2 score quartiles is P < .001.
At the time of application submission, USML step 2 was completed by 756 AΩA applicants and 2833 non-AΩA applicants.
Table 3 presents results for the unadjusted and adjusted models. In the unadjusted model, white medical students were more likely than black, Hispanic, and Asian medical students to be members of AΩA. This association remained when adding variables to account for AΩA selection criteria, including Gold Humanism membership as a proxy for professionalism, self-reported hours dedicated to leadership activities, and hours worked for community service. In addition, this association remained for black and Asian medical students after controlling for USMLE Step 1 scores and whether the applicant had an article published in a medical journal or presented an oral abstract in the fully adjusted model. However, no difference was found in AΩA membership between white and Hispanic medical students in the fully adjusted model, although the CI was wide (OR, 0.79; 95% CI, 0.45-1.37).
Table 3. Associations Between AΩA Status and Applicant Characteristics.
| Characteristic | OR (95% CI) | ||
|---|---|---|---|
| Unadjusted (n = 4655) |
Adjusted for AΩA Proxies (n = 4655) |
Fully Adjusted (n = 4655) | |
| Race/ethnicity | |||
| White | 1 [Reference] | 1 [Reference] | 1 [Reference] |
| Black | 0.07 (0.03-0.15) | 0.07 (0.03-0.16) | 0.16 (0.07-0.37) |
| Hispanic | 0.46 (0.30-0.71) | 0.47 (0.31-0.72) | 0.79 (0.45-1.37) |
| Asian or Pacific Islander | 0.46 (0.39-0.56) | 0.48 (0.40-0.59) | 0.52 (0.42-0.65) |
| Multiracial | 0.62 (0.46-0.83) | 0.65 (0.48-0.88) | 0.85 (0.59-1.21) |
| Other | 0.49 (0.28-0.85) | 0.50 (0.29-0.88) | 0.65 (0.33-1.27) |
| Sex | |||
| Male | NA | 1 [Reference] | 1 [Reference] |
| Female | NA | 0.82 (0.71-0.95) | 1.60 (1.33-1.93) |
| AΩA selection proxies | |||
| Gold Humanism | NA | 2.31 (1.89-2.86) | 2.93 (2.24-3.82) |
| Leadership hours | |||
| <50th percentile | NA | 1 [Reference] | 1 [Reference] |
| 50th-75th percentile | NA | 1.25 (1.04-1.51) | 1.13 (0.91-1.41) |
| >75th percentile | NA | 1.11 (0.89-1.39) | 1.03 (0.79-1.35) |
| Community service hours | |||
| <50th percentile | NA | 1 [Reference] | 1 [Reference] |
| 50th-75th percentile | NA | 1.08 (0.89-1.31) | 1.23 (0.97-1.55) |
| >75th percentile | NA | 0.87 (0.69-1.08) | 1.18 (0.91-1.53) |
| USMLE Step 1 | |||
| Bottom quartile | NA | NA | 1 [Reference] |
| Third quartile | NA | NA | 4.21 (2.51-7.08) |
| Second quartile | NA | NA | 23.26 (14.24-38.00) |
| Top quartile | NA | NA | 105.57 (63.95-174.28) |
| Master’s degree | NA | NA | 0.83 (0.64-1.07) |
| PhD | NA | NA | 0.89 (0.51-1.59) |
| Author-publication or abstract (yes/no) | NA | NA | 1.22 (1.02-1.47) |
Abbreviations: AΩA, Alpha Omega Alpha; NA, not applicable; OR, odds ratio; USMLE, US Medical Licensing Examination.
The USMLE Step 1 scores demonstrated a strong association with AΩA membership. The CI for the USMLE Step 1 scores was wide (top quartile: OR, 105.57; 95% CI, 63.95-174.28), and consequently, there remains uncertainty regarding the true effect size of USMLE Step 1 scores on AΩA membership. In addition, we found a statistically significant interaction between race/ethnicity and USMLE Step 1 scores when treating these scores as a binary variable (top quartile USMLE Step 1) (OR, 15.41; 95% CI, 1.68-141.74 for blacks; OR, 3.27; 95% CI, 1.08-9.94 for Hispanics; OR, 1.26; 95% CI, 0.82-1.92 for Asians; OR, 0.74; 95% CI, 0.38-1.45 for multiracial; OR, 1.25; 95% CI, 0.35-4.46 for other; P = .03). We found that black (adjusted OR, 0.02; 95% CI, 0.003-0.17), Hispanic (OR, 0.37; 95% CI 0.18-0.75), and Asian (OR, 0.43; 95% CI, 0.32-0.58) medical students were less likely than white medical students to be AΩA members in the model restricted to students not ranking in the top quartile of USMLE Step 1 scores. In the model restricted to students ranking in the top quartile of USMLE Step 1 scores, Asian students remained less likely than white students to be AΩA members (OR, 0.55; 95% CI, 0.41-0.76), and black applicants were also less likely than white applicants to be AΩA members; however, the association for black applicants did not reach statistical significance (OR, 0.38; 95% CI, 0.13-1.07).
No association was found between AΩA membership and the self-reported number of hours worked in leadership positions or the self-reported number of hours worked in community service. The presence of a master’s degree or PhD was not associated with AΩA membership. Publication in a medical journal or oral abstract presentation at a medical conference was associated with AΩA membership. Women were less likely to be AΩA members in the model adjusted for AΩA selection proxies; however, this trend reversed in the fully adjusted model after controlling for USMLE Step 1 scores (Table 3).
Akaike information criterion statistics were performed to assess the goodness of fit of each model, with the fully adjusted model having the smallest residual error (Akaike information criterion in the unadjusted model, 4593; adjusted for AΩA proxies, 4528; fully adjusted, 3388). Prevalence estimates for AΩA membership by race/ethnicity for the fully adjusted model were calculated (black applicants: PR, 0.19; 95% CI, 0.09-0.43; Hispanic: PR, 0.83; 95% CI, 0.51-1.28; Asian: PR, 0.57; 95% CI, 0.47-0.69; multiracial: PR, 0.88; 95% CI, 0.64-1.16; other: PR, 0.72; 95% CI, 0.38-1.17) and were similar to the calculated ORs.
Discussion
After controlling for numerous demographic and educational covariates, we found that the odds of AΩA membership for white students were nearly 6 times greater than those for black students and nearly 2 times greater than for Asian students. Although previous literature has reported that minority faculty are less likely to be promoted and to receive NIH funding in academic settings, this is the first study, to our knowledge, to describe a similar disparity in the receipt of academic awards at the level of undergraduate medical education.
No statistically significant difference in AΩA membership was found between white and Hispanic students. Nevertheless, the CI for the likelihood of Hispanic applicants being AΩA members is wide in the fully adjusted model, and it is possible that we lack statistical power, thus resulting in a type II error.
We identified a strong association between AΩA members and high USMLE Step 1 scores but found no difference in the self-reported number of hours dedicated to leadership activities or community service between AΩA members and nonmembers, suggesting that leadership activities and community service are not being strongly weighed in election to the society.
Limitations
Our study has several limitations. First, because of variation in the reporting in ERAS applications, it is difficult to know the class rank of every applicant. Some applications explicitly state the medical student’s rank, some give a range, and some make no statement at all. Consequently, we are unable to pinpoint the precise cause of the disparity making some racial/ethnic minorities less likely than white medical students to be AΩA members. Are minorities less likely to be in the top quartile of their class, or if in the top quartile, are minorities still less likely to be chosen as AΩA members? Although we are unable to adjust for academic performance, we adjusted for USMLE Step 1 scores, which is a reasonable surrogate.
A second limitation is the 1-year sampling frame. Although the racial disparity reported in this study is similar to the disparity reported more than 25 years ago (Table 4), it is possible that our cohort does not reflect trends in the racial/ethnic composition of AΩA members over time.
Table 4. Demographic Composition of US Medical Students and AΩA.
| Characteristic | No. (%) of Applicants | |||
|---|---|---|---|---|
| 1989 | 2015 | |||
| US Medical Students (n = 14 405) |
AΩA (n = 2246) |
US Medical Students (n = 18 349) |
AΩA (n = 966)a |
|
| Female | 3788 (26.3) | 602 (26.8) | 8725 (47.6) | 404 (41.8) |
| Race/ethnicity | ||||
| White | 12 319 (85.5) | 2083 (92.7) | 11 012 (60.0) | 691 (71.5) |
| Black | 769 (5.4) | 31 (1.4) | 1061 (5.8) | 7 (0.7) |
| Hispanic | 610 (4.2) | 52 (2.3) | 865 (4.7) | 27 (2.8) |
| Asian or Pacific Islander | 553 (3.8) | 67 (3.0) | 3701 (20.8) | 168 (17.4) |
Abbreviation: AΩA, Alpha Omega Alpha.
Proportions calculated from study cohort.
Third, all applications were collected at a single academic medical center. Because our institution draws applicants from more research-intensive institutions, it is possible that the disparity described in our cohort is attenuated among academic medical centers with less research focus. It is also possible that a disproportionately small number of black, Hispanic, and Asian medical students who are AΩA members applied to Yale residency programs, which could bias our results. Because there is no public database providing demographic data about AΩA members, information about the race/ethnicity of AΩA members nationally is limited. Nevertheless, the racial/ethnic composition of our cohort is similar to the national population of medical students with regard to race/ethnicity, and our cohort has a higher percentage of AΩA members than the national population of applicants.
Fourth, we used self-reported total hours dedicated to leadership activities and community service as proxies for AΩA membership selection criteria. It is possible that these proxies do not accurately reflect the depth and quality of the student’s engagement with the stated activities and/or that these measures may not reflect the data used by the individual AΩA chapter committees when evaluating medical students for society induction. Moreover, all information evaluated in the study was abstracted from ERAS applications, and it is possible that there are intangible characteristics or academic achievements not present in each application that contributed to a student being inducted into AΩA.
Implications
This study has potential implications for academic medicine leaders and the national AΩA society. Academic medicine leaders have consistently advocated for the need to strengthen the pipeline of minorities entering academic medicine. Although underrepresented minorities comprise 11% of all medical students, they only comprise approximately 5% of new academic medicine hires. Given the association between AΩA membership and a medical student choosing a career in academic medicine, the racial/ethnic disparity in AΩA membership described in this study may undermine the pipeline of minorities entering the academic health care workforce.
This study’s findings suggest that individual medical schools might benefit from internal review of their AΩA membership profiles with specific attention to differences by race/ethnicity. Medical schools with a racial/ethnic disparity in AΩA membership may wish to revisit the mission of their AΩA chapter and the guidance given to their honor society selection committee to ensure it is not vulnerable to this type of bias. In addition, medical schools should continue to think about creative ways to recognize a range of necessary skills and talents—beyond grades and USMLE Step 1 scores—that contribute to excellence in medical practice among physicians.
Given the study findings, residency program directors may want to pause and ensure that their interview triage process is multipronged. As the number of residency applications increases, program directors understandably look for filters to efficiently identify qualified applicants. Nevertheless, program directors using an interview triage system dominated by AΩA membership status could introduce systematic bias into their method of candidate selection. Residency program directors can influence the process substantially by requesting and relying on holistic student assessments. An applicant’s status as a first-generation college student, origin in a community that is medically underserved, and foreign language ability are examples of nonacademic criteria that some undergraduate medical school programs have successfully incorporated into their holistic review.
The racial/ethnic disparity demonstrated in this study may also be of interest to the national AΩA society. The society has been the hallmark of undergraduate medical excellence for more than a century. As medical schools further diversify along multiple dimensions, revisiting the standardization of the guidance provided to local chapters regarding the membership selection process may be appropriate.
The role of USMLE Step 1 scores in membership selection is an area where reexamination of the guidance given to local chapters may be appropriate. High USMLE Step 1 scores were highly correlated with AΩA membership in our study cohort despite the holistic review process prescribed by the national AΩA society. Many studies have found that, on average, underrepresented minorities and women score lower than white men on part 1 of the National Board of Medical Examiners.
Given that our findings suggest the current practice of membership selection might be vulnerable to bias, another important step for the national AΩA society could be the regular collection and dissemination of member characteristics across the country. Although publicly available data could be presented in aggregate, this database would provide AΩA with the ability to internally benchmark trends in member demographic data over time.
Conclusions
Overall, black and Asian medical students were less likely than their white counterparts to be members of AΩA, which may reflect bias in selection. In turn, AΩA membership selection may affect future opportunities for minority medical students. Promoting excellence among medical students is an admirable goal, but it must be accomplished free of bias and supportive of a diverse student membership.
Reference
- 1.Nunez-Smith M, Ciarleglio MM, Sandoval-Schaefer T, et al. Institutional variation in the promotion of racial/ethnic minority faculty at US medical schools. Am J Public Health. 2012;102(5):852-858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ginther DK, Haak LL, Schaffer WT, Kington R. Are race, ethnicity, and medical school affiliation associated with NIH R01 type 1 award probability for physician investigators? Acad Med. 2012;87(11):1516-1524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ginther DK, Schaffer WT, Schnell J, et al. Race, ethnicity, and NIH research awards. Science. 2011;333(6045):1015-1019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Fang D, Moy E, Colburn L, Hurley J. Racial and ethnic disparities in faculty promotion in academic medicine. JAMA. 2000;284(9):1085-1092. [DOI] [PubMed] [Google Scholar]
- 5.Palepu A, Carr PL, Friedman RH, Amos H, Ash AS, Moskowitz MA. Minority faculty and academic rank in medicine. JAMA. 1998;280(9):767-771. [DOI] [PubMed] [Google Scholar]
- 6.Alpha Omega Honor Society [website]. 2016. https://alphaomegaalpha.org. Accessed January 19, 2016.
- 7.LaGrasso JR, Kennedy DA, Hoehn JG, Ashruf S, Przybyla AM. Selection criteria for the integrated model of plastic surgery residency. Plast Reconstr Surg. 2008;121(3):121e-125e. [DOI] [PubMed] [Google Scholar]
- 8.Rinard JR, Mahabir RC. Successfully matching into surgical specialties: an analysis of national resident matching program data. J Grad Med Educ. 2010;2(3):316-321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.DeCroff CM, Mahabir RC, Zamboni WA. The impact of alpha omega alpha membership on successfully matching to residency. Plast Reconstr Surg. 2010;126(2):113e-115e. [DOI] [PubMed] [Google Scholar]
- 10.National Resident Matching Program. Charting Outcomes in the Match. Characteristics of Applicants Who Matched to Their Preferred Specialty in the 2014 Main Residency Match. 2014. http://www.nrmp.org/wp-content/uploads/2014/09/Charting-Outcomes-2014-Final.pdf. Accessed April 25, 2016.
- 11.Brancati FL, Mead LA, Levine DM, Martin D, Margolis S, Klag MJ. Early predictors of career achievement in academic medicine. JAMA. 1992;267(10):1372-1376. [PubMed] [Google Scholar]
- 12.How Members Are Chosen. Alpha Omega Alpha Honor Society [website]. 2016. https://alphaomegaalpha.org/how.html. Accessed February 28, 2016.
- 13.Babbott D, Weaver SO, Baldwin DC Jr. Personal characteristics, career plans, and specialty choices of medical students elected to Alpha Omega Alpha. Arch Intern Med. 1989;149(3):576-580. [PubMed] [Google Scholar]
- 14.Zhang J, Yu KF. What’s the relative risk? a method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280(19):1690-1691. [DOI] [PubMed] [Google Scholar]
- 15.Association of American Medical Colleges Data and Analysis, Total Graduates by US Medical School and Race and Ethnicity, 2014-2015. https://www.aamc.org/download/321536/data/factstableb4.pdf. Accessed February 9, 2017.
- 16.Cohen JJ. Time to shatter the glass ceiling for minority faculty. JAMA. 1998;280(9):821-822. [DOI] [PubMed] [Google Scholar]
- 17.Nivet MA, Taylor VS, Butts GC, et al. Diversity in academic medicine no. 1 case for minority faculty development today. Mt Sinai J Med. 2008;75(6):491-498. [DOI] [PubMed] [Google Scholar]
- 18.Petersdorf RG, Turner KS, Nickens HW, Ready T. Minorities in medicine: past, present, and future. Acad Med. 1990;65(11):663-670. [DOI] [PubMed] [Google Scholar]
- 19.Sullivan L. Missing Persons: Minorities in the Health Professions: A Report of the Sullivan Commission on Diversity in the Healthcare Workforce. 2004. http://www.aacn.nche.edu/media-relations/SullivanReport.pdf Accessed February 8, 2017, 2016.
- 20.Drees B, Nairn R, Nivet M, Danek J. Holistic Admissions in the Health Professions: Findings From a National Survey. 2014. http://urbanuniversitiesforhealth.org/media/documents/Holistic_Admissions_in_the_Health_Professions.pdf. Accessed September 25, 2016.
- 21.Case SM, Becker DF, Swanson DB. Performances of men and women on NBME Part I and Part II: the more things change.... Acad Med. 1993;68(10)(suppl):S25-S27. [DOI] [PubMed] [Google Scholar]
- 22.Veloski JJ, Callahan CA, Xu G, Hojat M, Nash DB. Prediction of students’ performances on licensing examinations using age, race, sex, undergraduate GPAs, and MCAT scores. Acad Med. 2000;75(10)(suppl):S28-S30. [DOI] [PubMed] [Google Scholar]
- 23.Dawson B, Iwamoto CK, Ross LP, Nungester RJ, Swanson DB, Volle RL. Performance on the National Board of Medical Examiners: Part I examination by men and women of different race and ethnicity. JAMA. 1994;272(9):674-679. [PubMed] [Google Scholar]