Abstracts Category : Other

Relevant Accessible Sensitivity Analysisfor Clinical Trials with Missing Data

by S Cro

The statistical analysis of longitudinal randomised controlled trials is frequently complicated by theoccurrence of protocol deviations which result in incomplete datasets for analysis. However analysisis approached, an unverifiable assumption about the distribution of the unobserved post-deviationdata must be made. In such circumstances it is consequently important to assess the robustnessof the primary analysis of the trial to different credible assumptions about the distribution of themissing data.Reference based multiple imputation procedures have been proposed for contextually relevantsensitivity analysis of longitudinal trials. Differences between the mean and variance of observedand missing data are specified with qualitative reference to trial arms and multiple imputation isused for estimation and inference. The primary analysis model is retained in the sensitivity analysisto assess the impact of alternative sampling behaviour on the original planned analysis. Rubin'srules are used to combine the treatment effect and variance estimates across imputed datasets,however it is unclear precisely what an appropriate measure of variance is in this setting and howRubin's variance formula relates to this.We begin by defining a lower bound for variance estimation in the reference based settings asthe variance estimate we would obtain were we able to observe the deviation data under thepostulated post-deviation data assumption. We show Rubin's variance estimate always exceedsthis and moreover it approximately preserves the loss of information in the primary analysis. Wealso explore Rubin's variance estimate in the -adjusted sensitivity analysis setting and show thatRubin's variance formula preserves the loss of information in this context.Alongside, we develop a new Stata command mimix" for implementation of reference based sensitivityanalyses. We illustrate the relevance and accessibility of the proposed methods of sensitivityanalysis using data from a chronic asthma trial and a study of peer review.