Abstract
OBJECTIVES: We examined the evolution and influencing factors of women's authorship in randomised controlled trials (RCTs) published in rheumatology. METHODS: We analysed all RCTs published in rheumatology from 2009 to 2023 determining authors' gender using the Gender API service. The percentage of women as authors in published RCTs and its association with potential factors was assessed using generalised estimating equations. We considered women's gender as the primary outcome and included the continent of the RCT, the status of international collaboration, industrial funding, intervention type, sample size, the journal's adherence to International Committee of Medical Journal Editors recommendations, impact factor, publication year, authors' non-academic affiliations and author positions as exposures of interest. RESULTS: We included 1092 RCTs with 10 794 authors; in the overall non-adjusted estimated analysis, we found that women accounted for 39.8% (95% CI 38.4% to 41.2%) of all authors. Women authorship was higher in African-based RCTs, among pharmaceutical-affiliated authors, and when the last author was a woman (OR 2.34 (95% CI 1.02 to 5.38), +19.46 pp). It was lower in Asian and European RCTs and industry-funded RCTs (OR 0.79 (95% CI 0.66 to 0.93), -5.85 pp). Women were less often last (OR 0.63 (95% CI 0.54 to 0.74), -10.2 pp) or second-to-last authors (OR 0.73 (95% CI 0.62 to 0.85), -10.19 pp), with no differences by international status or publication year. CONCLUSION: The persistent under-representation of women in RCTs and their lower chances of being senior authors highlight the need for better strategies to close the gender gap. RCTs with a woman last author were more likely to have a woman first author, suggesting a potential role model effect.