Clinical research has historically failed to include representative levels of historically underrepresented populations and these inequities continue to persist. Ensuring representativeness in clinical trials is crucial for patients to receive clinically appropriate treatment and have equitable access to novel therapies
enhancing the generalizability of study results
and reducing the need for post-marketing commitments focused on underrepresented groups. As demonstrated by recent legislation and guidance documents, regulatory agencies have shown an increased interest in understanding how novel therapies will impact the patient population that will receive them. Despite these efforts, a systematic approach to measure and optimize representativeness remains underdeveloped. Here, we introduce the novel Population Optimization, Representativeness Evaluation, and Fine-tuning Framework, designed to quantify and enhance representativeness. Our framework includes methods for evaluating overall and subgroup representativeness, identifying drivers of non-representativeness, and optimizing eligibility criteria to achieve representative populations. We demonstrate our framework by selecting patients who met the eligibility criteria for nine oncology clinical trials from a nationwide electronic health record-derived de-identified database and quantifying the representativeness of each trial's eligible population. This framework addresses gaps in current literature by providing a comprehensive, data-driven approach to enhance the representativeness of clinical trials, thereby supporting regulatory and internal decision-making processes. This framework is adaptable to various disease indications and can be extended to evaluate enrolled study samples, ensuring that clinical trials are representative.