2017/10/07 by Michael J. Grayling, Grayling, Michael, Adrian Mander +3
Decision Sciences · Mathematics · #Advanced Statistical Process Monitoring #FOS: Computer and information sciences #Methodology (stat.ME) #Statistical Methods and Bayesian Inference #Statistical Methods in Clinical Trials
paper · pdf · doi:10.48550/arxiv.1710.02683
openalex publication_date 2017/10/07 · openalex created_date 2022/10/04 · openalex updated_date 2026/07/28
The ability to accurately estimate the sample size required by a\nstepped-wedge (SW) cluster randomized trial (CRT) routinely depends upon the\nspecification of several nuisance parameters. If these parameters are\nmis-specified, the trial could be over-powered, leading to increased cost, or\nunder-powered, enhancing the likelihood of a false negative. We address this\nissue here for cross-sectional SW-CRTs, analyzed with a particular linear mixed\nmodel, by proposing methods for blinded and unblinded sample size re-estimation\n(SSRE). Blinded estimators for the variance parameters of a SW-CRT analyzed\nusing the Hussey and Hughes model are derived. Then, procedures for blinded and\nunblinded SSRE after any time period in a SW-CRT are detailed. The performance\nof these procedures is then examined and contrasted using two example trial\ndesign scenarios. We find that if the two key variance parameters were\nunder-specified by 50%, the SSRE procedures were able to increase power over\nthe conventional SW-CRT design by up to 29%, resulting in an empirical power\nabove the desired level. Moreover, the performance of the re-estimation\nprocedures was relatively insensitive to the timing of the interim assessment.\nThus, the considered SSRE procedures can bring substantial gains in power when\nthe underlying variance parameters are mis-specified. Though there are\npractical issues to consider, the procedure's performance means researchers\nshould consider incorporating SSRE in to future SW-CRTs.\n