Restricted mean models for transplant benefit and urgency.

The US lung allocation policy estimates each individual's urgency and transplant benefit in defining a lung allocation score (LAS). Transplant benefit, as defined by the Organ Procurement and Transplantation Network Thoracic Committee, is the days of life gained over the following year if transplanted versus not transplanted. Urgency is measured by days of life during the next year without transplant. In both definitions, accurate estimation of wait list days lived, or a wait list restricted mean lifetime, is required. Risk factors are available to estimate patient urgency when listed, with more urgent patients removed from the wait list upon death or transplant. As a patient progresses, priority for transplant (censoring) changes accordingly. Therefore, it is crucial to adjust for dependent censoring in modeling days of life. We develop a model for the restricted mean as a function of covariates, by using pseudo-observations that account for dependent censoring linked to a series of longitudinal measures (LAS). Simulation results show that our method performs well in situations comparable with the LAS setting. Applying wait list and post-transplant model results that account for dependent censoring to wait list patients, we obtain estimates of transplant benefit that are larger for many of the more urgent patients in need of transplant. The difference in LAS for an individual, when properly accounting for dependent censoring, has high impact on the priority and timing of an organ offer for these patients.