A Markov Multistate Analysis of the Relationship Between Performance Status and Death Among a Population of Cancer Patients

• Activity Number: 348
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Rinku Sutradhar*+ and Lisa Barbera
• Companies: University of Toronto and University of Toronto
• Keywords: interval-censored event times ; Markov process ; multistate model ; palliative performance scale

Abstract

In observational studies on cancer patients, progression of performance status over time can be described by using a multistate model in which state-to-state transitions represent changes in a patient's health condition. Although a patient experiences transitions in continuous time, assessments on the patient are often made at irregularly spaced time points. This paper formulates a Markov model for examining longitudinal data on performance status collected under intermittent observation. The main aims are to understand the association between performance status and death, and to determine the probability of deterioration in performance status over time. The authors also compare this multistate model with extended time-to-death Cox models in which performance status is a time-varying measure. Methods for multistate analyses should be used by epidemiologists, as it proves particularly useful for examining the complexities of disease processes.