P123 Data driven clinical trial recruitment – From the common to the rare

<h3>Background</h3> Approximately 85% of clinical trials fail to reach recruitment goals and recruitment in the UK has fallen by 44% between 2017 and 2022. This has significant consequences for patients suffering from diseases with limited treatment options. Hepatology trials are particularly affected by this decline exacerbated by high levels of screen failure. Hepatology studies often have complex eligibility criteria including co-morbidities and laboratory abnormalities which limits patient recruitment. In addition, more deprived communities are less likely to participate in clinical trials. Common conditions such as Metabolic dysfunction Associated Steatotic Liver Disease (MASLD), as well as rarer conditions such as Primary Biliary Cholangitis (PBC) are the focus of drug trials currently in the recruitment phase. In Somerset, we developed a case-finding search engine which allows rapid analysis of patterns in blood tests to identify cohorts of patients. Here we evaluated the potential of an approach using this database for clinical trial recruitment. <h3>Methods</h3> We identified clinical trials currently recruiting PBC and MASLD patients using the NIH database (at www.clinicaltrials.gov). We assessed the recruitment targets, number of centres recruiting for each study and eligibility criteria (EC). We then assessed an alternative approach for recruitment using our database as an analytical tool to identify patients for these studies. It is anticipated the database will soon cover 10 million population. We assessed the potential power of recruitment by extrapolating from our current to our proposed footprint. <h3>Results</h3> 4 studies in MASLD and 4 in PBC were selected for analysis (figure 1). Those EC which were based upon biomarkers were rapidly assessed to identify large numbers of potential recruits using the database. Conventional approaches aimed to recruit on average 2.42 patients per research site. In contrast, our approach (based on 10M footprint) would identify on average 1,819 patients per trial. With 10% enrolment of eligible patients, our approach would yield the equivalent of 70 (S.D. 44) conventional sites per 10M footprint. <h3>Conclusion</h3> Our approach led to the rapid identification of a group of patients eligible for clinical trials in MASLD and PBC compared to traditional approaches. Specifically, this approach does not rely on waiting for a clinical encounter and does not rely on historic coded data. Far reaching implications of undertaking this approach include widening access to and increasing participation in clinical trials leading to more rapid availability of therapeutics.