|
Activity Number:
|
397
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Health Policy Statistics
|
| Abstract - #307979 |
|
Title:
|
Propensity Scores and Multiple Treatments: An Application to Beneficiary Evaluations from the Medicare CAHPS Surveys
|
|
Author(s):
|
Marc N. Elliott*+ and Amelia Haviland
|
|
Companies:
|
RAND Corporation and RAND Corporation
|
|
Address:
|
1776 Main St., Santa Monica, CA, 90407-2138,
|
|
Keywords:
|
Multinomial ; Prescription Drugs ; Insurance ; Consumer Evaluation of Healthcare ; Observational Data ; Propensity Score
|
|
Abstract:
|
The Centers for Medicare and Medicaid Services provides comparisons of beneficiary experiences with original and managed care Medicare. In this paper, we use the 2004 Medicare CAHPS surveys to compare beneficiary perceptions of care across five different types of Medicare health coverage that include original Medicare and managed care plans with and without prescription drug coverage and other supplementary coverage. Enrollees are not randomly assigned to types of health coverage, and there are well documented differences between the enrollees across these five options. This paper compares estimates of beneficiary experience using regression methods to 'control' for differences in enrollee characteristics to estimates obtained from a propensity score analysis for multiple groups. We also use related models to demonstrate that the "optimal" coverage varies by enrollee characteristics.
|
