Designing Data Linkage Systems for the National Health and Aging Trends Study

The new National Survey of Aging and Health (NSAH) is a longitudinal survey of aged Medicare Beneficiaries that seeks to better understand the disablement process, its precursors and consequences. This is one of the first national data sets being designed with explicit personal and geographical data linkage in mind from the outset, since these contextual factors are increasingly recognized as highly influential in shaping elders' response to the disablement process. We plan linkages to: 1) Medicare enrollment records and all Part A (institutional), Part B (physician) and Part D (drug) claims, and 2) information about providers used by survey respondents based upon the Medicare Provider of Service files and other sources (e.g. the increasingly detailed information available about provider quality of care from external sources). Geo-coding the respondents' home address, provides the potential for measures of distance from various health care providers as well as socio-economic status of the census block group, zip code and/or county where the respondent resides. Creating restricted files that allow users to link respondents' survey responses to these rich health care utilization and local area characteristics should be relatively straight-forward using established data security protocols for data use agreements with individual researchers. Creating public use files which gives users access to summary data on utilization, providers and other contextual and neighborhood data will be more challenging since the threshold for determining the identifiability of individual respondents is complex.