Structuring simulation to determine optimal sample size re-estimation strategy at trial design
Featuring Emily Hammarstrom-Wickens, Statistician II at PHASTAR
Abstract: 'What would happen if...? Structuring simulation to determine an optimal sample size re-estimation strategy at the trial design stage'.
When proving something mathematically is difficult or limited to specific situations, simulation studies may be an effective solution. In a clinical trial setting, simulations can represent the results of simultaneously performing many real trials, creating an empirical distribution for measures of interest, such as the Type I error. Anticipated parameters, such as the expected treatment effect, and the sensitivity in changes to such parameters can be assessed via simulating plausible scenarios for a clinical trial. Transparent and well-planned simulation studies can be modified to suit particular requirements, making them both a useful and informative tool when designing a clinical trial.
Assessing the benefit and functionality of a new adaptive design would be an appropriate occasion for utilising simulation studies. For example, assessing the effect of using Bayesian sample size reestimation methods (SSR) on measures such as the Type I error rate. This example is motivated by the RO LARR trial Uayne et al., 2017), which failed to show a statistically significant treatment effect, potentially as a result from the misspecification of the sample size assumptions. SSR methods use accumulated study data to adjust the initial assumptions and re-estimate the sample size. Assessment and comparison of properties associated with these methods provides an example of where simulation studies can be used at the design stage of a clinical trial.
This talk provides guidance on the use of simulation studies to assess the benefit of using a Bayesian SSR design and the selection of an optimal strategy.