|
Abstract:
|
Linking administrative and survey records is a cost-effective method to enhance survey data. The Medicare Current Beneficiary Survey (MCBS), a nationally representative survey of Medicare beneficiaries, is unique because its sampling frame is the Medicare enrollment data making the survey data inherently linked to administrative data. For Original Medicare Beneficiaries, the administrative data contain information on cost and utilization. The survey questionnaire also includes these measures, along with information that is not available in the administrative data. The overlap between the two sources provides a unique opportunity to examine their agreement. Beyond the valid instances where a claim is not linked to the event reported in the survey (e.g. due to Medicare Advantage enrollment) this examination can help highlight the potential for measurement error in the survey. In this paper we will compare beneficiary characteristics for those with survey only, claims only, and linked survey and claims data across certain health care event types. The results from this research will inform ways to use Medicare administrative data to improve the MCBS design and data collection.
|
