Sex Bias in Interventional Clinical Trials

Abstract

Introduction: Prior studies have shown that sex bias exists with subject enrollment in clinical trials, with more men being enrolled than women. The objective of this study was to identify if sex bias continues to exist in present day clinical trials entered into ClinicalTrials.gov . We hypothesize that males and females are not equally represented in recent clinical trials.

Materials and methods: Data were abstracted from all interventional Phase I, II, and III clinical trials with adult subjects entered into ClinicalTrials.gov from January 1, 2011 to December 31, 2013 and completed by November 30, 2015. Number and sex of subjects, funding source, allocation, end point classification, interventional model, and purpose were recorded. Studies pertaining to diseases that were sex specific were excluded.

Results: Of 1,668 studies included in the initial search, 167 were excluded due to sex-specific study topic. Of the remaining 1,501 studies, 177,656 (51.1%) male and 170,331 (49.0%) female subjects were included. There was a significant difference in the sex of the subjects included in Phase I (64.1% male/35.9% female), Phase II (48.4% male/51.6% female), and Phase III (51.0% male/49.1% female) clinical trials (p < 0.05). Similarly, there was a significant difference in the sex of the subjects included in industry (50.7% male/49.3% female), National Institutes of Health (NIH) (56.6% male/43.4% female), "Other US Federal" (62.5% male/37.5% female), and "Other" funded (53.4% male/46.6% female) clinical trials (p < 0.0001), as well as between randomized (50.5% male/49.6% female) and nonrandomized (54.8% male/45.2% female) clinical trials (p < 0.0001). Upon evaluating if the sex of the subjects included in the individual clinical trials was equally matched, we found that only 4.1% of clinical trials had 100% sex matching, 22.2% had 80% sex matching, and 56.5% had 50% sex matching. Using a liberal 50% sex-matching criteria, Phase III and II clinical trials matched the sex of the subjects more frequently compared to Phase I trials (60.8%, 57.8%, and 45.5%, respectively, p = 0.003).

Conclusion: These data reveal that sex bias is present in current day clinical trials. Despite legislation requiring NIH-funded clinical trials to include women, NIH-funded trials were not better than industry-funded trials at matching the inclusion of both sexes.

Keywords: clinical trials; clinicaltrials.gov; sex bias.