Representation of Women on National Institutes of Health Study Sections

This cross-sectional study examined participants on chartered and special emphasis NIH study sections during one review cycle from May 15 to July 15, 2019. The University of Chicago institutional review board deemed this study exempt because it used publicly available data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. Data were extracted about study section reviewers (name, degree, academic rank, temporary membership, chair), scientific review officers (SROs; name), and institutes, centers, or offices (hereafter referred to as institutes; chair name, total funding, number of awards, award amount, and success rate for research grants and projects in 2018). Individuals’ gender presentation (man or woman) was determined through name-based internet searches for pictures and pronouns. If unconfirmed, gender was assigned using software (Genderize.io [Demografix ApS]; threshold = 60%). χ2 tests were used to compare gender distributions between institutes. Univariate and multivariate logistic regressions were applied to determine the likelihood of having women on study sections based on institute and section characteristics. Significance was defined by P < .05 using 2-sided tests. Data analyses were performed using SAS version 9.4 (SAS Institute) in August 2020.


Introduction
Disparities in gender representation among researchers are well described throughout the career trajectory, including grant awards. [1][2][3] This study aimed to evaluate whether differences exist in gender representation on National Institutes of Health (NIH) study sections, which help determine funding.

Methods
This cross-sectional study examined participants on chartered and special emphasis NIH study sections during one review cycle from May 15 to July 15, 2019. The University of Chicago institutional review board deemed this study exempt because it used publicly available data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Data were extracted about study section reviewers (name, degree, academic rank, temporary membership, chair), scientific review officers (SROs; name), and institutes, centers, or offices (hereafter referred to as institutes; chair name, total funding, number of awards, award amount, and success rate for research grants and projects in 2018).
Individuals' gender presentation (man or woman) was determined through name-based internet searches for pictures and pronouns. If unconfirmed, gender was assigned using software (Genderize.io [Demografix ApS]; threshold = 60%). χ 2 tests were used to compare gender distributions between institutes. Univariate and multivariate logistic regressions were applied to determine the likelihood of having women on study sections based on institute and section characteristics. Significance was defined by P < .05 using 2-sided tests. Data analyses were performed using SAS version 9.4 (SAS Institute) in August 2020.   convened; integrated review groups and study sections for business and fellowship grants were excluded from this analysis. There were a total of 8820 total participants (unique = 8346). Gender was identified using name-based internet searches to find pictures and/or pronouns. For 23 participants, gender could not be confirmed. Using software (Genderize.io, probability threshold = 60%), gender was assigned for 20 of these participants. Gender could not be determined for 3 participants (unique = 2). Thus, the analysis included 8817 participants. Data about study section participants was obtained from publicly available meeting rosters at the NIH Scientific Review Group Roster Index. Data about NIH institutes, centers, and offices was obtained from the National Institutes of Health Data Book.

Results
A total of 367 study sections convened with 8817 participants analyzed ( Table 1). Of these, 3432 were female (38.9%). Among the 25 institutes, the mean (SD) proportion of women reviewers was 39.0% (15.1%). Women constituted more than half of reviewers in 4 institutes (16% Overall, reviewers were more likely to be men than women (61.1% vs 38.9%, P < .001). Six institutes (24%) were more likely to have men as reviewers and one was more likely to have women.
Across all institutes ( were not associated with the proportion of women reviewers on study sections.

Discussion
To our knowledge, this study is the first to describe differences in gender representation among NIH study sections. Men were more likely to be reviewers and chairs. Women were more likely to have temporary affiliations and serve on study sections with lower total funding and research grants awarded, suggesting less influential opportunities to impact the nation's research agenda. These findings may partially explain gender bias in peer review and differences in funding and promotion [4][5][6] with potential ramifications for gender representation in academia.