Factors associated with women authorship over 25 years in high-impact critical care randomized controlled trials: the pipeline is still leaking

Abstract

Studies have shown an underrepresentation of women in clinical research, but little is known about trends over time and factors associated with this underrepresentation. This study aimed to investigate the evolution and the independent factors associated with the proportion of women in the authorship of high-impact critical care randomized controlled trials (RCTs) over 25 years. This meta-epidemiological study screened adult critical care RCTs published between 1999 and 2023 in the six highest-impact general and critical care journals. Sex was assessed using a combination of authors’ biographies, available photographs, the gender R package, and a native Chinese speaker’s assistance to ensure cultural accuracy. The primary outcome was the proportion of women among the authors. Unadjusted and adjusted generalized linear mixed models were performed. Of 1,203 RCTs, the sex of all authors was determined in 1,199 (99.7%). Overall, 4,335 out of 16,057 authors (27%) were women. Women were less frequently first (247/1,199 [21%], p < 0.001) or senior authors (174/1,199 [15%], p < 0.001) compared to other positions. The proportion of women among authors increased by a change rate of 0.7% per year ([0.5%—0.9%]) from 18% in 1999 to 32% in 2023. In multivariable analysis, the proportion of women increased significantly per year of publication (odds ratio [OR] = 1.05, 95% confidence interval [1.02–1.09], p < 0.01) and sample size (OR = 1.007 [1.003–1.012] per 100 patients increase, p = 0.01), and decreased significantly in European RCTs (OR = 0.53 [0.33–0.85], p < 0.01), RCTs on ventilation (OR = 0.50 [0.25–0.97], p = 0.04) but not sepsis (OR = 0.74 [0.37–1.46], p = 0.39), mortality as primary outcome (OR = 0.36 [0.14–0.92], p = 0.03), and with endorsement by a study group (OR = 0.36 [0.18–0.69], p < 0.01). Although the proportion of women in authorship has risen over 25 years, women are still widely underrepresented in the authors of high-impact RCTs, especially as first and senior authors. This underrepresentation is exacerbated in Europe, in trials with mortality as primary outcome, on ventilation, or endorsed by a study group. This illustrates the “leaky pipeline” framework: women are still excluded from the highest-level of critical care research.