Sex and Gender in Randomized Controlled Trials of Adults Receiving Maintenance Dialysis: A Meta-epidemiologic Study.
Collister D., Pyne L., Bhasin AA., Ahmed SB., Smyth B., Herrington W., Jardine M., Walsh M.
RATIONALE AND OBJECTIVE: How sex and gender concepts are incorporated into randomized controlled trials (RCTs) in adults with kidney failure receiving maintenance dialysis is largely unknown. We sought to describe these practices in published journal articles as well as the proportion of females and women participants in these studies. STUDY DESIGN: Meta-epidemiologic study. SETTING & STUDY POPULATIONS: RCTs in maintenance dialysis. SELECTION CRITERIA FOR STUDIES: Trials published in high-impact journals from 2000 to 2020. DATA EXTRACTION: Implemented in duplicate with conflicts resolved by a third reviewer. ANALYTICAL APPROACH: Meta-regression was performed to identify trial characteristics independently associated with the proportion of female and women participants. RESULTS: Among 561 included RCTs, 69.7% were parallel and 28.0% were crossover in design. 80.6% were conducted in the hemodialysis population. 25% of trials compared the treatment of interest to a placebo arm, 25% to a usual care treatment arm, and 50% to an active alternative therapy arm. 37.6% of RCTs were blinded. The median (IQR) size was 60 participants (26, 151) and the median (IQR) follow-up was 154 days (42, 365). The mean (SD) proportion of female or women participants was 0.40 (0.13). 39.0% of trials reported sex and 26.6% reported the gender of participants. 56.2% referred to participants as females, 25.3% referred to participants as women, and 15.5% referred to both females and women. No trial characteristic other than region (Asia, ß 0.062 95% CI 0.007-0.117) was associated with the proportion of female or women participants. Considering trial design and conduct, 2.7% of trials used sex and/or gender as an inclusion criterion, 26.6% as an exclusion criterion, 4.5% for randomization, 4.8% for subgroup analyses, and 15.7% for covariate adjustment. LIMITATIONS: Only high-impact journal articles were studied, the lack of pediatric trials and of those addressing chronic kidney disease or kidney transplantation, the absence of any trials from Africa and underrepresentation of other regions, and missing data. CONCLUSIONS: RCTs in dialysis are representative of the general dialysis population with regard to sex and gender but uncommonly report both, and often do not include either in their reporting or analysis.