Study Discussion – Gender Disparities in Internal Medicine Residency Awards

In the March 2021 issue of The American Journal of Medicine, seven physicians, whose first names suggest that they are all female, wrote “Investigating Gender Disparities in Internal Medicine Residency Awards.”[1] The authors began by recounting gender disparities in salary, academic rank, grant funding, and awards. They performed a multi-institutional study based on survey data from academic internal medicine residency programs starting in 2009 and extending through 2019. These physicians’ initial findings are in Table 1:[2]

Table 1

Program Male Residents

%

Male Awardees

%

Male Chiefs

%[3]

Years Male Program Director %[4] Years Male Chair % Male Awardees/Male Residents % Male Chiefs/Male Residents % Hours of implicit bias training
A 54.2 85 58.3 36.4 100 1.57 1.08 5
B 48 65 38.8 100 100 1.35 0.81 4
C 46.6 60 57.1 18.2 100 1.29 1.23 0
D 63.7 68.6 64.3 90.9 0 1.08 1.01 4
E 65.3 67.5 67.5 20 100 1.03 1.03 2
F 59.3 61.7 50 36.4 72.7 1.04 0.84 2
G 49.8 48.4 44.2 9.1 63.6 0.97 0.89 4
H 68.5 63.8 73.3 100 100 0.93 1.07 ?
Avg 56.9 65 56.7 51.4 79.5 1.16 0.99 3

 

These data suggest that males received awards (Awardee/Resident ratio of 1.16) at a higher rate than females, but the difference was not statistically significant. It also shows that women gained prestigious positions (such as chief resident and program director) at the same rate or a higher rate than men. In this study, men held the position of department chair more commonly than women. Such data also shows no relationship between receiving anti-implicit bias training and a higher female representation in earning awards or more prestigious positions. Why are these results what they are? There are three primary possibilities:

  1. Males did not deserve more awards and prestigious positions in this sample. The overrepresentation in awards and “years male chair” noted herein is a function of gender bias. Similarly, females did not deserve to have more years as program director, which also suggests gender bias.
  2. Males deserved more awards and prestigious positions in this sample because they, on the whole, performed better than females. Likewise, females performed better as program director and this fact justified their representation.
  3. Logistic, administrative, financial, or other factors influenced the process of selection for awards and prestigious positions that have nothing to do with performance or gender bias. If a physician withdrew from consideration to handle personal matters, and a person of a different gender was selected instead, that fact might communicate gender bias in the selection when in reality the result was due to unforeseen circumstances. If such events occurred more commonly in one gender than the other, it could skew the results.

The information presented does not allow readers to reach any conclusions, though the authors strongly suggest that males have been unjustly overrepresented in the receipt of awards and prestigious positions. Table 2 indicates the gender of award winners by award category.

Table 2

Award Category Male Winners Female Winners Female Winners % P Value
Ambulatory and Community 4 12 75.0 0.012
Compassion and Humanism 53 98 64.9 <0.001
Excellence 35 31 47.0 0.255
Research 14 10 41.7 0.828
Resident of the Year 32 19 37.3 0.342
Teaching 166 56 25.2 <0.001

 

Notably, awards for “ambulatory”, “community”, “compassion”, and “humanism” went overwhelmingly to females. The authors imply that such awards are rooted in a gender stereotype of women as caring. Awards for “excellence”, “research”, and “resident of the year” show no statistically significant gender predilection. “Teaching” awards, however, went overwhelmingly to males. As noted above, the data reveal “what” but not “why,” and the options are similar.

  1. Females do not deserve more awards in these categories in this sample. The overrepresentation noted herein is a function of gender bias.
  2. Females deserve more awards in these categories in this sample because they, on the whole, performed better than males in compassion, humanism, ambulatory, and community areas.
  3. Other factors skewed the results.

Similarly,

  1. Males do not deserve more teaching awards in this sample. The overrepresentation noted herein is a function of gender bias.
  2. Males deserve more teaching awards in this sample because they, on the whole, perform better than females.
  3. Other factors skewed the results.

If one assumes that men and women, as it pertains to medical practice, are exactly the same, then gender bias is the leading possibility to explain disparities in medical awards. If male physicians and female physicians have no pertinent differences, then the only reason for any disparity is poor selection methods. The authors provided some data on winners by selection method in Table 3.

Table 3

Selection Mechanism Male Winners Female Winners Female Winners % P Value
Selection committee 5 12 70.6 0.027
Medical student vote 10 9 47.4 0.759
Program director selection 42 36 46.2 0.686
Chief selection 24 20 45.5 0.833
Faculty vote 8 5 38.5 0.694
Resident vote 276 167 37.7 0.011

 

Table 3 demonstrates no statistically significant difference in award receipt between male and female internal medicine residents when the selection mechanism was medical students, program directors, chief residents, or faculty. However,

  1. Selection committees were far more likely to give awards to women. Selection committees are composed of leaders in the field, are usually gender balanced, try to incorporate a variety of stakeholders, and must be acceptable to the institution as a whole. Composed of 6-12 people, such committees’ decisions are tracked to minimize unacceptable bias. The authors of this study called for “diverse selection committees, particularly those aware of the impact of implicit bias,” to mitigate gender disparities in awards in internal medicine residencies. Using the data in Table 3, if selection committee was the only approved mechanism for awards, women would get 70% of all awards. Such an outcome would clearly be a gender disparity, but perhaps an acceptable one to these authors.
  2. Resident votes were far more likely to favor men. The average program size of internal medicine residency programs in this study was 106 residents. The authors noted that “We had hypothesized that based on age residents might hold less intrinsic gender bias than senior faculty and that therefore awards voted on by residents might have less gender disparity; however, our data show that gender bias in voting appears to be deeply rooted among the residents studied.”[5] This statement proves that the authors believed that male predominance in teaching awards was caused by bias, not by differences in teaching ability.

“Investigating Gender Disparities in Internal Medicine Residency Awards,” demonstrates areas in which males and females receive awards out of proportion to their numbers. The authors recommended gender balanced selection committees, which we have seen above heavily favor female awardees. If voting must be used, which favors men in their data, the authors recommended having a statement on the ballot reminding voters to be aware of their implicit bias. These female physician authors also recommended using gender-balanced language in award selection.

Conclusion

The study at hand shows a remarkable lack of gender disparity overall, and undoubtedly less gender disparity than would have been the case 40 years ago. As noted in the article, implicit bias training does nothing to improve gender disparities, and sometimes merely reinforces gender stereotypes. One wonders if some characteristics that traditionally characterize women, such as compassion, are stereotypes, defined as “widely held but fixed and oversimplified images or ideas of a particular type of person or thing,” or archetypes, defined as “the original pattern or model of which all things of the same type are representations or copies.” Restated, are women perceived as more compassionate, on average, than men, when in reality they are not? Or are women truly biologically and socially more compassionate than men, on average?

Bias, whether sexual, racial, or some other type, is a major focus of American life today. How should internal medicine residency programs select who to award and who to promote? More importantly, how should other organizations, from churches to political parties, make such choices? Using only a selection committee disenfranchises many stakeholders, but allows organizations to award and promote those who they prefer. In the dataset above, selection committees favored women. Voting enfranchises many more people, but it limits organizations’ ability to heavily influence or even pre-select the winners. In the dataset above, voting favored men. Organizations will determine their own goals and will develop policies and procedures to achieve their goals. Will they honor the will of the majority, or the will of the few?

[1] Michelle Hannon MD, Katherine Duffey MD, Sonia Bharei MD, Rachel Redfield MD, Alison Greidinger MD, Emily Stewart MD, Gretchen Diemer MD, Investigating Gender Disparities in Internal Medicine Residency Awards, The American Journal of Medicine, Vol 134, No 3, March 2021, pp 405-409.

[2] Overall residency class population was 43.9% female.

[3] “Chief” refers to chief resident

[4] “Years Male” communicates what percentage of the ten-year span of the study a male physician held the position of chief resident, program director, or department chairman.

[5] The population of internal medicine residents was 56% male and 44% female. Residents voted for 276 male awardees and 167 female awardees, which is a 62 to 38 ratio. This fact suggests that more than a few female residents demonstrated the “gender bias” that the authors of this study lament.

2 thoughts on “Study Discussion – Gender Disparities in Internal Medicine Residency Awards”

  1. Have any of these studies been looking at whether possibly the men are actually better at the job than the women ? I know that is not a politically correct concept, but it IS a possible concept !

  2. I have not seen studies addressing the question you pose. They would certainly be difficult to perform, and controversial at best.

We love constructive feedback! Please leave a reply.