Background: Letters of recommendation (LORs) are crucial for fellowship applications, but implicit biases can affect their quality. Objective: The aim of this study was to examine sex-based differences among multiinstitutional pulmonary and critical care medicine (PCCM) fellowship LORs. Methods: All 2021 PCCM fellowship applications submitted to five regional U.S. programs (South, Midwest, East, Northeast, and West) were deidentified, and applicant demographics and accomplishments were abstracted. The letter writer's gender was identified through an online search of professional websites. Linguistic Inquiry and Word Count was used to identify the frequency of adjective types (e.g., grindstone, communal). We measured the associations between applicant sex and the outcomes of total word count and degree of support using multivariable linear regression models adjusting for applicant demographics, accomplishments, and letter writer's gender for each program. Results: In total, 9,153 LORs were included in the analysis. The majority of applicants (64.1% [n = 1,703 of 2,658]) were male, and most letter writers (72.1% [n = 6,603 of 9,152]) identified as men. When adjusting for applicant demographics and accomplishments and letter writer's gender, female applicants' LORs had significantly more supportive words compared with those of male applicants in three of the five programs (Midwest, 2.4 more words [95% confidence interval (CI), 0.5-4.3; P = 0.01]; South, 1.9 more words [95% CI, 0.1-3.6; P = 0.03]; Northeast, 2.5 more words [95% CI, 0.2-4.9; P = 0.04]) and longer letters in two programs (South, 40.1 more words [95% CI, 12.3-68.0; P = 0.005]; Midwest, 29.8 more words [95% CI, 0.7-58.9; P = 0.045]). Conclusion: The length and supportiveness of PCCM fellowship applicants' LORs varied by applicant sex and region. Further work should explore why these differences exist and how the information garnered from them is viewed by fellowship selection committees.
Keywords: fellowships; letters of recommendation; pulmonary and critical care medicine; sex/gender disparities.