Response to Letter from Wong on determining the target difference in sample size calculations for randomised controlled trials

Excerpt

We would like to thank Wong for their response to our article on the relevance of clinical importance when determining the target difference in sample size calculations and clarification on their position [1, 2]. It is with the hope of making clearer the differences in our positions that we make this further response, though we note we do agree on a number of practical points such as the challenges in estimation and the impact of the trial context [1]. As before, we consider the setting of designing a phase III or “definite” randomised controlled trial (RCT). The target difference refers to the treatment effect specified in the sample size calculation conducted when designing the study to detect a difference between treatments (i.e. a superiority contrast). We refer to Wong’s proposal as the “true benefit” approach to specification of the target difference in the sample size calculation, based on our understanding of the method [1, 3]. Unless otherwise stated, we assume a conventional (frequentist) Neyman-Pearson approach is being implemented which is currently the most common sample size determination approach used for RCTs. …