Author Diversity on Clinical Practice Guideline Committees

Abstract

Reports have demonstrated that women overall and women physicians, in particular, are underrepresented as authors of clinical practice guidelines. This analysis used publicly available information to explore the diversity of rehabilitation-related clinical practice guideline authors by gender, race, and ethnicity. Primary analysis identified authors’ gender, race, ethnicity, and visible minority status. Two sets were analyzed: (1) clinical practice guidelines by Department of Veterans Affairs (VA) categorized as “Rehabilitation” or “Pain” (n = 7; VA clinical practice guidelines) and (2) a set (n = 10) published in the United States (US) from 2019 to 2021 that were selected because of low numbers of inclusion at less than 20% women authors. Key findings include that among physician authors, both the VA and US clinical practice guidelines underrepresented women (15 [24.2%] and 27 [16.7%], respectively) and those coded as a racial or ethnic minority were particularly underrepresented. Notably, women authors overall were equally represented (92 [50.0%]) in the VA clinical practice guidelines. The US clinical practice guidelines had women authors who were underrepresented (36 [19.0%]). Secondary analysis of the entire set of VA clinical practice guidelines (n = 21) found gaps in diversity-related content. Clinical practice guidelines have far-reaching health and economic impacts, and addressing disparities in the diversity of author teams and/or gaps in diversity-related content is of paramount importance.

Gender Representation Among Physician Authors of Practice Guidelines Developed, Endorsed, or Affirmed by the American Academy of Neurology

“…Disparities in authorship raise concerns about the content included or absent, ease of implementation, and potentially patient outcomes. 32…”

Analysis of physician recipients of recognition awards from the American Academy of Physical Medicine and Rehabilitation by Race and Ethnicity

“…For our methodology, we have carefully drawn from previous studies that similarly assess race and ethnicity, and we have developed a model that has already been successfully utilized in a recent study. 24 However, we recognize that our approach was influenced by a relatively small data set that included award recipients known to one or more of our authors. As such, our approach may face considerable challenges with larger data sets.…”

“…Given the aforementioned methodological challenges combined with a pressing need to provide more support to medical students, residents, fellows, and physicians who identify with racial and/or ethnic minority groups, we used the approach of analyzing publicly available data, a two‐rater system, and standardized race and ethnicity categories. For our methodology, we have carefully drawn from previous studies that similarly assess race and ethnicity, and we have developed a model that has already been successfully utilized in a recent study 24 . However, we recognize that our approach was influenced by a relatively small data set that included award recipients known to one or more of our authors.…”

“…The methodology of coding gender, race, and ethnicity is not novel, and the process used in this study is similar to methods in the published literature for gender equity studies 2,5,7,9,21 and the evolving literature on race/ethnicity. [22][23][24][25] Importantly, our data represent the perceived race and ethnicity of recipients, and we do not claim to know anyone's true race or ethnicity.…”

“…Therefore, the only disagreement was related to ethnicity, and the two senior authors each did an independent search of the four individuals (five awards) and agreed on the coding of the ethnicity for all four individuals using both their own knowledge of the individuals as well as publicly available information (eg, voter registration, where an individual grew up and went to school, http://ancestry.com). The methodology of coding gender, race, and ethnicity is not novel, and the process used in this study is similar to methods in the published literature for gender equity studies 2,5,7,9,21 and the evolving literature on race/ethnicity 22‐25 . Importantly, our data represent the perceived race and ethnicity of recipients, and we do not claim to know anyone's true race or ethnicity.…”