Practical guidance for conducting high-quality and rapid interim analyses in adaptive clinical trials.

BACKGROUND: Adaptive designs are increasingly being used in clinical trials within diverse clinical areas. They can offer advantages over traditional non-adaptive approaches, including improved efficiency and patient benefit. The level of improvement observed in practice depends to a large degree on conducting interim analyses (at which adaptations can be made to the trial based on collected data) rapidly and to a high standard. METHODS: The ROBust INterims for adaptive designs (ROBIN) project aimed to identify best practice for conducting high-quality and rapid interim analyses. This was done through evidence synthesis of published work, qualitative research with trial stakeholders working at public sector clinical trials units, engagement with patients and the public, and a meeting of trial stakeholders to discuss findings and agree recommendations. RESULTS: This paper provides recommendations for teams that conduct adaptive trials about how to ensure interim analyses are done rapidly and to a high standard. We break down recommendations by stage of the trial. We also identify a lack of methodology on how best to involve patients in adaptive trials and related decision-making. A limitation of our recommendations is that the research was mostly focused on UK academic settings, although we believe much of the recommendations are relevant in other countries and to industry-sponsored trials. CONCLUSIONS: When following the recommendations outlined in this paper, the process of planning and executing interim analyses will be smoother; in turn, this will lead to more benefits from using adaptive designs.