Sequential Trials Symposium
Wednesday 26 September 2018
London School of Hygiene & Tropical Medicine
The Centre for Statistical Methodology and the Tropical Epidemiology Group of the London School of Hygiene and Tropical Medicine (LSHTM) held a half day symposium on sequential trials on the afternoon of 26 September 2018. Presentations highlighted recent research and application of these methods, and speakers included Sue Todd (University of Reading), Dominic Magirr (AstraZeneca), Babak Choodari-Oskooei (MRC Clinical Trials Unit at UCL), and Neal Alexander (LSHTM).
Sequential designs are intended to rapidly assess new interventions, particularly when the outcome is rare, and have arguably been under-used in medical research. For example, sequential trials have been proposed as a way of ‘getting an answer sooner and cheaper’ for neglected tropical diseases. This meeting concentrated on ‘fully sequential’ trials, but group sequential designs were also addressed. In the former, the data are analysed after each patient’s results become available, and the trial is stopped after a pre-specified objective is reached. Hence, the sample size is not fixed in advance. In the group-sequential approach, the sample size is determined conventionally, but the trial may be stopped at one of a small number of interim analyses.
Sue Todd gave a review of designs involving selection, based on practical examples, and including different approaches to sequential monitoring, and reporting standards for sequential trials. Dominic Magirr presented methods for generalizing boundaries for triangular designs, e.g. for asymmetric stopping, motivated by a trial of treatments for visceral leishmaniasis. Babak Choodari-Oskooei examined the impact of adding new experimental arms to multi-arm multi-stage (MAMS) platform trials. Finally, Neal Alexander reflected on experience with sequential trials against visceral leishmaniasis, and the pros and cons of such designs as compared to classical designs.