A2040
Title: Covariate hypothesis tests for the cure rate in mixture cure models
Authors: Ana Lopez-Cheda - University of A Coruna (Spain) [presenting]
Maria Amalia Jacome Pumar - Universidade da Coruna (Spain)
Antia Dorado - University of A Coruna (Spain)
Abstract: Survival analysis studies the time to an event in the presence of censoring. In many biomedical studies, a group of individuals does not experience the event, which motivates the use of cure models. Mixture cure models (MCM) allow for the separation of the incidence (probability of being susceptible) and the latency (survival function of susceptible individuals). A key aspect is evaluating the influence of covariates on both parts of the MCM using hypothesis tests. Two approaches of nonparametric methods, implemented in the R packages npcure and MDcure, are reviewed. Furthermore, two approaches that combine both methods are proposed. The behavior of the four methods is analyzed through an extensive simulation study. Finally, the methodology is applied to a dataset of breast cancer patients.
