Median Cost Analysis for Recurrent Event Data

• Activity Number: 169
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Alexander McLain*+ and Raji Sundaram and Subhashis Ghoshal
• Companies: Eunice Kennedy Shriver National Institute of Child Health and Human Development and National Institute of Child Health and Human Development and North Carolina State University
• Keywords: Cost analysis ; Intensity function ; Proportional rate model ; Recurrent events

Abstract

In biomedical studies subjects may experience event of interest repeatedly. Such recurrent event data have been much studied in the statistical literature with application to cancer tumors, small bowel motility, schizophrenia, serious AIDS infections and many others. In many of these examples the time to such episodic events are associated with medical costs in treating them. To understand the accumulated cost for treating such a subject an understanding of both the recurrent event process and the associated medical costs process is needed. Furthermore, the distributions of the number of recurrent events and the cost of each episode are commonly right skewed. For this reason the mean cost process may not be a representative estimate. Here we use nonparametric recurrent event survival analysis techniques and a Gamma cost model to estimate, and predict, aspects of the median cost distribution. We derive the asymptotic distribution of the mean and median cost, and apply our techniques to the SEER-Medicare data set. This is based on joint work with S. Ghoshal (NCSU) and R. Sundaram (NICHD/NIH).