Project Summary
Clinical trial participation in the United States does not reflect the diversity of disease populations. While about 43% of the US population is not white, less than 10% of the population in clinical trials over the last 25 years were people of color. The imbalance in clinical research inclusion leads to limitations in applying clinical data, therapeutic indexes, drug safety, and toxicity, resulting in poorer health outcomes for excluded populations. This problem raises the concern that underrepresented minorities and marginalized populations may not have equitable access to life-saving treatments.
We propose AIquitas, a two-pronged machine learning approach embedded in a scalable, modular technology system that addresses the problem from both the investigator and patient perspective. We aim to develop a system to parse clinical trial inclusion and exclusion criteria to identify the specific requirement(s) that impacts diversity in the recruitment process. Furthermore, our model and platform will allow easier recruitment of more diverse patients, integrated into a tool that patients can use to screen trials they may be eligible for. Such a tool can be easily used by patients to identify trials they can participate in; help clinical trial investigators access a larger and more diverse pool of participants; and support government regulators, such as the FDA, in making policy changes that incentivize better clinical research practices.