Racial differences in cardiothoracic surgery letters of recommendation.

Abstract

Objective: Cardiothoracic surgery lacks racial diversity. Attracting the best candidates requires valuing diversity and advancing a recruitment process that minimizes racial bias. Implicit bias in recommendation letters has been identified as an influential mechanism for lack of diversity in cardiothoracic surgery. We aim to investigate how applicant race impacts language used in cardiothoracic surgery fellowship recommendation letters.

Methods: We examined applications to an accredited cardiothoracic surgery fellowship program between 2017 and 2021. Recommendation letter authors were categorized as cardiothoracic and noncardiothoracic surgeons. Applicant self-reported race was dichotomized as White and non-White. Applicant characteristics were described using standard descriptive statistics and compared between applicant race. Recommendation letters were examined with linguistic analysis software and higher-level analysis to assess language associations between applicant race and author specialty.

Results: A total of 597 recommendation letters were analyzed from 158 individual applications. Cardiothoracic surgeons composed 334 (55.9%) recommendation letters. A total of 95 applicants (60.1%) were categorized as White. White applicants were described as more driven (P = .03), whereas letters for non-White applicants were more authentic (P = .02) and highlighted past experiences (P = .05). Cardiothoracic surgeons wrote longer letters (P = .04) with standout adjectives (P = .03) more frequently for White applicants and emphasized perceptiveness (P = .02) more often when writing for non-White applicants.

Conclusions: We demonstrate that authors' recommendation letters are influenced by applicant race. Letters written for White candidates are significantly longer and more descriptive of drive and outstanding characteristics than letters written for non-White candidates, particularly when written by cardiothoracic surgeons. Addressing this bias is crucial for recruiting the best candidates for cardiothoracic surgery.