Random effect restricted mean survival time model

The restricted mean survival time (RMST) model has gained attention as a clinically intuitive measure of mean survival time. RMST models, which apply methods based on pseudo time-to-event values and inverse probability censoring weighting, can adjust covariates. However, no existing approach incorporates random effects to account for cluster-level variability. In this paper, we propose a new random-effects RMST model. We present two analytical methods that account for variable effects: (i) a generalized mixed model with pseudo-values and (ii) an approach that integrates the estimated results from inverse probability censoring weighting estimating equations applied to each cluster. We evaluate the performance of our proposed methods through computer simulations. Additionally, we demonstrate their practical application by analyzing the effect of maternal age at birth on under-five mortality in India, using states as clusters.