Although this is a time of constrained resources, there are exponentially expanding needs for data that will assist policy makers and program administrators at the federal, state, and local levels ensure coverage of eligible populations, manage programs efficiently, and evaluate the effectiveness of program options and approaches.
One consistent theme of the workshop was the emerging power to link survey to survey, survey to administrative record data, and administrative to administrative record data systems. In addition to hearing about pioneering record matching projects, such as those involving State Health Access Data Assistance Center (SHADAC), the National Center for Health Statistics (NCHS), the Agency for Healthcare Research and Quality (AHRQ), the U.S. Department of Health and Human Services Assistant Secretary for Planning and Evaluation (ASPE), the Centers for Medicare & Medicaid Services (CMS), and the U.S. Census Bureau, the workshop was informed that the CMS has released a task order to produce person-level data files on Medicaid eligibility, service utilization, and payment information to support comparative effectiveness research funded by the ARRA.
As was pointed out in the workshop, as a component of ARRA, CMS has been approved by the U.S. Department of Health and Human Services to produce and enhance the Medicaid Analytic eXtract file (MAX). The agency has released a task order relating to the production of MAX, along with the development of an accelerated version of MAX to produce more timely data for research use. The range of activities that are optional under this task order include verification of Social Security numbers, production of a master file of Medicaid/CHIP enrollment, and linkage of MAX to federal surveys, identified in the solicitation as the National Health Information Survey, the National Health and Nutrition Examination Survey, the Longitudinal Study on Aging, the National Nursing Home Survey, the Medicare Current Beneficiary Survey, and a prototype for the Census Bureau’s ACS.
Many of the opportunities for matching exist at the state level. It was suggested that state-level matching could involve using other data to get a better match of the potentially eligible population with enrollment files. These links between administrative data systems would permit improvements in the modeling of state impacts and in the development of measurement error models. Because insurance coverage is dynamic, useful links would be not only cross-sectional but also longitudinal—creating records on individuals over time. One benefit of this matching activity, it was suggested at the workshop, would be to improve understanding of the existing data, their limitations, and how they could be better used for policy analysis.