The Peters-Belson Method for Assessing Health Care Disparities, Extended to Survival Analysis

• Activity Number: 572
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Health Policy Statistics
• Author(s): Lynn Eberly and James S. Hodges*+ and Donna Z. Bliss and Kay Savik and Olga Gurvich and Susan L. Harms and Christine A. Mueller
• Companies: University of Minnesota and University of Minnesota and University of Minnesota and University of Minnesota and University of Minnesota and University of Minnesota and University of Minnesota
• Keywords: disparities ; survival analysis ; Peters-Belson method

Abstract

The Peters-Belson method was developed for quantifying and testing disparities between groups, using linear regression to compute group-specific observed and expected outcomes. It has since been extended to logistic regression for binary outcomes. For an NIH-funded project assessing racial/ethnic disparities in nursing home care (the REDSKIN Study), we extended the Peters-Belson approach to survival analysis, including stratified analyses. The extension uses the theory and methods of expected survival based on Cox regression in a reference population, as developed in Therneau & Grambsch (Modeling Survival Data, 2000; Springer; Chapter 10). We used the survival package in the R system to do computations; Therneau & Grambsch give a SAS macro that could be adapted readily. We describe the extension, show how we applied it in the REDSKIN Study, and discuss some issues in implementing it.