PS34: Patient Reported Outcomes in Small-Size Hematological Clinical Trials: Statistical Challenges and Considerations

Clinical outcome assessments (COAs), which measure how patients feel, function, or survive, are increasingly recognized as essential components of drug development and regulatory decision-making. COAs such as patient-reported outcomes (PROs) can capture direct patient input and help to inform assessments of treatment effects in clinical trials. While PROs are now commonly incorporated into clinical trials, there is variability by disease area in terms of endpoint definitions, statistical considerations, and analytical approaches. This variability complicates PRO endpoint interpretation and reporting in rare disease clinical trials, where limited sample sizes and patient heterogeneity create additional challenges. While conducting studies with fit-for-purpose PROs in their intended context of use is essential for capturing patient experience, suitable study designs, valid endpoints, and appropriate statistical analyses are crucial. Critical considerations for PRO endpoints in small-scale studies include but not limited to: 1) clinical relevance and statistical impact of individual PRO items used for the target population or subpopulations, 2) evaluation strategies for PROs as binary versus continuous endpoints, 3) handling of missing data at both item or form levels and intercurrent events, 4) ordering and multiplicity control for PRO endpoint(s), and 5) interpretability of PRO endpoint results in single-arm and/or open-label study designs. Inadequate handling of these statistical challenges can bias estimates, compromise result reliability, undermine conclusion robustness, and may ultimately lead to study failure. This session will illustrate these challenges through pertinent case examples of PRO-related efficacy endpoints in rare hematological clinical trials to foster extensive discussion.