Handling Missing Outcome Data in Cluster Randomized Trials With Both Individual- and Cluster-Level Dropout

Missing outcome data are common in cluster randomized trials (CRTs), which can complicate inference. Further, the missingness can occur due to dropout of individuals, termed sporadically missing data, or dropout of clusters, termed systematically missing data, and these two types of missingness could have potentially different missing data mechanisms. We aimed to develop a well-performing and practical approach to handle inference in CRTs when outcome data may be both sporadically and systematically missing. To this end, we first examined the performance of several multilevel multiple imputation (MI) methods to handle sporadically and systematically missing CRT outcome data via a simulation study. Specifically, we examined performance under a multilevel covariate-dependent missingness assumption. Our findings indicated that several full conditional specification (FCS) methods designed for missingness in linear mixed models performed well under various scenarios, while an FCS approach using a two-stage estimator often performed poorly. We then developed methods for conducting sensitivity analysis to test the robustness of inferences under different missing at random (MAR) and missing not at random (MNAR) assumptions. The methods allow for different MNAR assumptions for cluster dropout and individual dropout to reflect that they may arise from different missing data mechanisms. We used graphical displays to visualize sensitivity analysis results. Our methods are illustrated using a real data application.