Influence of Adaptive Analysis on Unnecessary Patient Recruitment: Reanalysis of the RATPAC Trial

Abstract

Recruitment to clinical trials is a challenging but essential activity in emergency medicine. Conventional fixed-sample trials may continue to recruit patients after efficacy has been demonstrated or when further recruitment is futile. Adaptive trials make use of emerging information to modify aspects of a trial or terminate it prematurely, potentially resulting in savings in terms of sample size, time, and cost. We aim to use sequential testing procedures to reanalyze data from a fixed-sample trial, the Randomised Assessment of Treatment Using Panel Assay of Cardiac Markers (RATPAC) trial, and investigate the potential for adaptive designs to reduce unnecessary recruitment. The trial was reanalyzed with a triangular group sequential design, with interim analyses planned every 3 months. Patients were analyzed in the order in which they entered the original trial. We found that the RATPAC trial could potentially have stopped 1 year earlier, with 722 patients enrolled compared with 2,243 patients in the original trial, making a potential saving of approximately $390,000. Estimates of effect were similar, and the qualitative conclusions of the original and group sequential RATPAC trials were in agreement. However, the group sequential approach is not without limitations and would have resulted in less precise estimates of effect and less information available for the subsequent evaluation of secondary endpoints. Sequential designs are well suited in emergency medicine because of the rapidly obtained outcomes and the need to avoid unnecessary recruitment. We recommend that group sequential designs be considered for clinical trials in emergency medicine.