|
Abstract:
|
When analyzing successive or recurrent times-to-event, close attention is oftentimes paid to patient health-related quality-of-life (HRQOL). Assessing patient health status is a highly complex process, requiring that numerous HRQOL scores be collected and analyzed. HRQOL-adjusted times-to-event are useful and simple ways to summarize such complex data structures. As an application, treatment strategies that favor improved survival adjusted for one or more, possibly patient-selected, HRQOL scores, might be envisioned. Therefore, one needs to understand the joint distribution of several times-to-event adjusted by multiple HRQOL scores. Hence, using inverse probability-of-censoring weighting techniques, we propose nonparametric estimators for such joint distributions. Their performance is evaluated in simulations and their practical usefulness is illustrated by a clinical trial application.
|
