(1029) Principal investigator sex and trial success: An analysis of early discontinuation, results reporting, and publication

To characterize United States (US) obstetric clinical trials by principal investigator (PI) sex and analyze the association between PI sex and features of trial success.

Study Design:

This is a cross-sectional study of all US-based obstetric clinical trials registered on ClinicalTrials.gov (2007-2020). We examined associations between PI sex and three primary outcomes: early discontinuation, results reporting of complete trials to ClinicalTrials.gov, and publication on PubMed.gov, controlling for number of sites, number of PIs, funding source, and year. We conducted exploratory analyses of other trial features.

Results: Of 1287 US-based obstetric trials, 1046 (81.2%) listed PI information. Trials included all female (F:627 trials, 59.9%), all male (M:375, 35.9%), or both male and female PIs (B:44, 4.2%). Most trials (946, 90.4%) reported one PI. The distribution of trials by PI sex was largely unchanged over the study period (Figure 1A). A lower proportion of female trials were discontinued early (F:5.7%, M:12.0%, B:9.1%; p=0.002), whereas results reporting and publication frequencies did not differ significantly by PI sex (p >0.05, Figure 1B). In multivariable analysis, female PI trials had lower odds of early discontinuation than male PI trials (adjusted odds ratio 0.43, 95% confidence interval 0.27-0.69), but demonstrated no differences in results reporting or publication (Figure 2). Female and male PI trials were similar in randomization, blinding, and use of an active comparator (Figure 1B). A lower proportion of female PI trials were industry-funded (F:1.0%, M:2.4%, B:4.5%; p=0.05), included oversight by a data monitoring committee (F:35.1%, M:41.6%, B:54.5%; p=0.05), or had multiple sites (F:16.7%, M:22.1.6%, B:36.4%; p=0.002).

Conclusion: In recent years, the majority of obstetric trials were led by female PIs, although the proportion of female PIs remained lower than the proportion of female obstetricians. Female-led trials were less likely to discontinue early but were similar with regard to other characteristics of trial success, such as publication.