Events and Seminars : Upcoming Seminars

Modelling the Heterogeneity of Disease Recurrence Patterns in Early-Staged Breast Cancer

Department of Biostatistics and Epidemiology
University of Pennsylvania Perelman School of Medicine
11:00 a.m.–12:00 p.m., CRB 692

Despite early detection and active treatment, early-staged breast cancer patients still experience recurrences or deaths after the initial BCT treatment. Common single endpoint analysis does not fully describe the passage of disease progression and often ignores the possibility that the population under study is a mixture of patients that with and without the potential for recurrence, and/or spread into other distant locations. We propose a mover-stayer mixture of two multi-state Markov processes to model such heterogeneity. Covariates are assumed to have multiplicative effects on transition intensities. The method is applied to a prospective cohort of 1892 women (median follow-up: 7.6 yrs) seen in the Hospital of University of Pennsylvania between 1977 and 2011.