Description:

Join guest speaker Lauren C. Balmert from the University of Pittsburgh for a seminar entitled "Nonparametric Inference and Regression on Quantile Lost Lifespan."

Time-to-event data may be analyzed based on the cumulative information up to a specific time point or based on the residual information beyond that time point. Here, a new summary measure, termed lost lifespan or life lost, is introduced to infer the distribution of time lost due to experiencing an event before some specified time point. The proposed approach reveals several advantages over existing methods for analyzing time-to-event data. A two-¬sample nonparametric inference procedure is first proposed for comparing distributions of lost lifespans between groups. The method is then extended to a regression setting to evaluate covariate effects on the lost lifespan. Simulation results are presented to assess the finite properties of the proposed methods, which are illustrated with a breast cancer dataset from a Phase 111 clinical trial conducted by the National Surgical Breast and Bowel Project (NSABP).

Location: Stamler Conference Room, 680 N Lake Shore Drive Ste. 1400