Conventional longitudinal data analysis methods assume that outcomes are independent of the data-collection schedule. However, the independence assumption may be violated, for example, when adverse events trigger additional visits in between prescheduled follow-ups. Outcome-dependent observation times may introduce bias when estimating the effect of covariates on outcomes using a standard longitudinal regression model. Recent research has focused on the development of joint modeling approaches that combine a semi-parametric regression model for a continuous outcome with a recurrent event model for the observation times. One such approach is based on estimating observation-level visit-intensity weights, which are used to reweight the model for longitudinal outcomes. We extend this approach to the analysis of binary outcomes. Computational limitations motivate us to consider a parametric, yet flexible, adjustment for temporal trends in the outcome model. We examine the performance of our methods via simulation studies, and illustrate their application to a randomized controlled trial regarding interventions to improve anti-coagulation control among patients treated with warfarin.
This program is for the research community.