Semi-Markov Processes for Estimating Pathways in the Diagnostic Evaluation of Suspected Breast Cancer

• Activity Number: 34
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Rebecca Hubbard and Jane Lange and Lurdes Inoue and Yue Zhang*+
• Companies: Group Health Research Institute and University of Washington and University of Washington and University of Utah
• Keywords: multi-state models ; Bayesian estimation ; cancer ; semi-markov process

Abstract

Semi-Markov processes (SMPs) provide a flexible approach to modeling multi-state processes with repeated events and multiple failure types because they do not assume exponential sojourn times. Despite this flexibility, SMPs have been underutilized, particularly in biomedical applications, due to the complexity of estimation. In this study we develop novel Bayesian estimation methods for SMPs. We apply this approach to a study of the process of diagnostic evaluation with mammography, ultrasound, and biopsy following a positive screening mammogram in terms of the timing and types of tests received. We also investigate risk factors associated with the sojourn time between diagnostic tests. By utilizing SMPs we expand on prior work which described the timing of the first test received by providing additional information such as the expected number of tests and length of time required to reach a definitive cancer diagnosis.