|
Activity Number:
|
374
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Health Policy Statistics
|
| Abstract - #309904 |
|
Title:
|
Analyzing Health Care Cost Data in the Presence of Heterogeneous Patients: Wouldn't Three Pieces Be Better Than One?
|
|
Author(s):
|
James F. Burgess, Jr.* and Kevin L. Sloan*+
|
|
Companies:
|
Boston University and University of Washington
|
|
Address:
|
Department of Psychiatry and Behavioral Sciences, Seattle, WA, 98195-6560,
|
|
Keywords:
|
Risk Adjustment ; Sub-Classification Models ; Cost Estimation
|
|
Abstract:
|
Effective prediction of health care costs for policy purposes requires balancing transparency and unbiasedness in choosing methods. Case mix information first can classify individual patients into spending type groups, then predict health care costs in each group. This is a special case of sub-classification models in statistics, but computationally much simpler to implement. We use split sample validation on FY2001 VA data to compare six models to our proposed three piece OLS model using residual mean square error (RMSE), mean absolute prediction error (MAPE), and predictive ratios by deciles for evaluation across the distribution of costs. The approach we propose performs much better on these evaluation criteria. We take a simple approach to subclassification, as opposed to more complex methods previously suggested (i.e., mixture models) to achieve better cost estimation performance.
|
