The representation of women on Australian clinical practice guideline panels, 2010-2020.

Objectives: To assess the composition by gender of Australian clinical practice guideline development panels; to explore guideline development-related factors that influence the composition of panels.

Design, Setting, Participants: Survey of clinical guidelines published in Australia during 2010-2020 that observed the 2016 NHMRC Standards for Guidelines, identified (June 2021) in the NHMRC Clinical Practice Guideline Portal or by searching the Guideline International Network guidelines library, the Trip medical database, and PubMed. The gender of contributors to guideline development was inferred from gendered titles (guideline documents) or pronouns (online biographies).

Main Outcome Measures: The overall proportion of guideline panel members - the guideline contributors who formally considered evidence and formulated recommendations (ie, guideline panel chairs and members) - who were women.

Results: Of 406 eligible guidelines, 335 listed the names of people who contributed to their development (82%). Of 7472 named contributors (including 511 guideline panel chairs [6.8%] and 5039 guideline panel members [67.4%]), 3514 were men (47.0%), 3345 were women (44.8%), and gender could not be determined for 612 (8.2%). A total of 215 guideline panel chairs were women (42.1%), 280 were men (54.8%); 2566 guideline panel members were men (50.9%), 2071 were women (41.1%). The proportion of female guideline panel members was smaller than 40% for 179 guidelines (53%) and larger than 60% for 71 guidelines (21%). The median guideline proportion of female panel members was smaller than 50% for all but two years (2017, 2018).

Conclusions: The representation of women in health leadership roles in Australia does not reflect their level of participation in the health care workforce. In particular, clinical guideline development bodies should develop transparent policies for increasing the participation of women in guideline development panels.