Evaluating Center-Specific Long-Term Outcomes Through Differences in Mean Survival Time

• Activity Number: 618
• Type: Contributed
• Date/Time: Thursday, August 2, 2012 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Zhi He*+ and Douglas E Schaubel
• Companies: University of Michigan and University of Michigan
• Keywords: Center effect ; Mean survival time ; Piece-wise Exponential ; Shared frailty model ; Survival analysis

Abstract

The majority of existing methods in the evaluation of center-specific survival involve contrasting death rates. However, the difference in mean lifetime is arguably a more meaningful measure than the ratio of death rates. Existing methods based on restricted mean lifetime tend to focus on short- and medium-term clinical outcomes. Such measures may fail to capture the long-term effects that relate to the quality of follow-up care. In this report, we propose a method that combines a log-normal frailty model and piece-wise exponential baseline rates to compare the mean survival time across centers. The proposed methods allow for consistent estimation of mean survival time as opposed to restricted mean lifetime and, within this context, more accurate profiling of long-term centerspecific outcomes. Maximum likelihood based estimation is carried out using a Laplace approximation for integration. Direct standardization methods are applied to contrast mean survival time by center. Asymptotic properties of the proposed estimators are derived, and finite-sample properties are examined through simulation. The proposed methods are applied to national kidney transplant data.