Event

An increasing number of trials are using adaptive design methodology, seeking to improve study efficiency or flexibility. To date though, all of the adaptive trials that have been held up as gold-standard practical examples have had individually randomised designs. Whilst there is a growing literature on methodology for the adaptive design of cluster randomised trials (CRTs), these papers do little to indicate when an adaptive CRT might be a good idea in practice. Here, I will describe work that seeks to resolve this issue. Examining the case of a parallel-group CRT, I will first detail methodology that facilitates interim efficacy/futility monitoring, sample size re-estimation, or response adaptive design of such trials, focusing on the limitations of these procedures. I will then present a simulation study that looks to clarify when the idealised theoretical advantages of adapting a CRT may hold-up in practice; principal considerations in this will be the rate of cluster recruitment and outcome accrual. Several CRTs that have used an adaptive design will also be discussed to consider what they’ve taught us so far about adapting a CRT. In all, it will be argued that the range of settings in which an adaptive design is a logical approach for a CRT is substantially smaller than that for an individually randomised trial, but that it could still be a good idea for many trials.

Michael Grayling has been a Research Fellow in Biostatistics at Newcastle University since 2018. Prior to this he was a Statistician at the MRC Biostatistics Unit, University of Cambridge. His current position involves a mix of research on the development of efficient methods of designing and analysing clinical trials and collaborating on applying such methods to real trials in practice. His interests include adaptive trial design, early phase drug development, and longitudinal study designs. He holds a PhD in Biostatistics, an MSci in Systems Biology, and a BA in Mathematics from the University of Cambridge.

This seminar will be held via Zoom from 6-7pm (AEDT)

Please click here to join.
Or, go to https://monash.zoom.us/join and enter meeting ID: 879 4335 9757 and passcode: 221873