Staff from the MRC Clinical Trials Unit at UCL have been busy again at day two of the joint 4th International Clinical Trials Methodology Conference - Society for Clinical Trials 38th Annual Meeting in Liverpool. We have had oral presentations and posters on a wide variety of topics, including implementing adaptive trial designs, participant involvement, practical trial conduct challenges and solutions, and sensitivity analysis for missing data.
One of the major themes of the Unit's presentations today was adaptive trials. The Unit Director, Max Parmar, started the day talking about the statistical benefits and practical challenges of implementing adaptive designs for Phase III trials. He argued that new approaches are needed, as the traditional two-arm trial is too slow, too costly, often unsuccessful, and inefficient when there are many therapies to test. Adaptive designs are becoming more popular for early and middle stage trials, but have been used less for late stage trials, where perhaps the biggest gains can be made.
Max Parmar outlined the Multi-Arm Multi-Stage (MAMS) trial design, which addresses these challenges through comparing several new therapies to a common control arm, and dropping insufficiently active arms at interim analyses. He used the example of the STAMPEDE trial to show how this approach can speed up evaluation of new therapies, and how new arms can be added as the trial progresses.
Later in the day, Francesca Schiavone and Riya Bathia presented a trial management perspective on the operational aspects of adaptive trials, using STAMPEDE and FOCUS4 as examples. This was complemented by a poster by Dominic Hague, Stephen Townsend and Lindsey Masters exploring the data management perspective.
Kenneth Babigumira also presented on an operational issue, describing how a drug supply management system was developed for the Add-Aspirin trial. Suzie Cro presented on how to deal statistically with a common problem for many trials: missing data. She discussed sensitivity analysis for missing data using the Delta method.
Claire Vale presented a challenge to the current INVOLVE definition of patient and public involvement, which states that for most studies it is not appropriate for people involved in the research to also be participants in the research. She used two interesting case studies of how trial participants can become actively involved in clinical trials, from the MDP301 and PROUD trials, to show that participant involvement can make a unique and valuable contribution to clinical trials.
There were also a number of posters presented, including a framework for adaptive meta-analysis, and database development.