Does Evidence-Based Medicine Serve Everyone? Representation, Reward and Risk In Phase I Clinical Trials

Abstract

Evidence-based medicine is a vast industry created around supporting the testing of new therapies before they are prescribed in the general population. Previous researchers have been concerned with the representativeness of these trials, citing concerns of external validity for female patients and racial minorities. However, Phase I trials, where the therapy is tested on a small group of healthy volunteers, have not been studied as extensively. Recent work suggests this group of participants may be disproportionately Black and Hispanic in the United States, reflecting a phenomenon of serial participation. These participants are often unfairly disadvantaged in the labor market, often facing unemployment, and are attracted to Phase I participation for the payment. These findings have not been replicated with a systematic review, as the original researchers relied on self report data, recruiting at trial centers themselves. This methodology is prone to selection bias, as it relies on permission. The present study was a systematic review of publicly available Phase I trials with the goal of clarifying factors which affect female enrollment as well as determining if enrollment by race is affected by whether or not the participants receive therapeutic benefit. A total of 300 articles reporting on 384 trials were included in this analysis. I found overall low female enrollment, except for in trials which were conducted in more than one global region (p < .001). Pharmacokinetic studies, pharmacodynamic studies, and studies conducted in Africa were also less likely to enroll female participants (p < .001). Within the United States, Black and Hispanic participants were found to be overrepresented as compared to their share of the population according to Census data, however this was not predicted by whether or not they participated as healthy volunteers as opposed to patients (p > .05). Most of the trials in the sample were conducted on healthy volunteers, however (76%). Future researchers may consider investigating a larger sample of trials with participant benefit for a more fair comparison.