The environmental scan outlined several sources of claims and administrative data for Medicaid enrollees, which were discussed by the TEP. The TEP focused primarily on MSIS and the enhancements and extensions under development in the Transformed Medicaid Statistical Information System (T-MSIS). (Additional detail on these data sources is included in Appendix B.) Utilization of HCUP data sources was not considered in detail because not all states participate and because access to the state data must be purchased through a central distributor. MSIS emerged as the recommended dataset for measuring realized access, with the caveat that there is a need for more complete understanding of the causes and implications of state-level variations in the data.
MSIS was identified as the main data source for monitoring realized access. However, there are several limitations and caveats to using MSIS for cross-state comparisons. Variation in benefit packages, program design, percent of the Medicaid population with fee-for-service (FFS) coverage, and data coding and reporting make cross-state comparisons challenging. Additionally, because it is expected that most newly eligible Medicaid beneficiaries will be enrolled in managed care plans, the lack of data on managed care Medicaid beneficiaries is a significant issue. In the short term, the use of MSIS for monitoring access will hinge on ASPE's and CMS's ability to engage with MSIS data experts to understand and control for state-level effects in the measures. Until these issues are identified and better understood, other types of comparisons may improve understanding and monitoring of access issues at the state-level, including: 1) comparing access for different eligibility populations across states; 2) comparing access within a given state across different eligibility groups over time; and 3) using a normative benchmark for care (e.g., are diabetic Medicaid enrollees meeting guidelines for visits and services?). As the data are utilized for monitoring access and the issues hindering state-level comparisons are identified, long-term solutions can be established to improve standardized reporting and to develop a fuller understanding of the factors that account for differences among states.
A recent report published by Mathematica Policy Research raises concerns about the completeness, reliability, and usability of encounter data in MAX, which is a set of annual person-level data files derived from MSIS. The report's findings indicate that MSIS may not be able to provide consistent measures on Medicaid beneficiaries' service utilization in the near future. The TEP concluded that improving this measure should be a priority area for the T-MSIS system.
The T-MSIS was identified as a potential data source when more states report data after 2013. The T-MSIS offers several advantages over the MSIS, such as the inclusion of the National Provider Identifier (NPI); more timely access to the data (within sixty days); integrated databases (including provider and managed care files),; and a more robust infrastructure, including automated data validation and analytic reporting. Currently, twelve states are using T-MSIS as part of Phase I. Phase II will add another four states by the end of 2012 and Phase III will add the remaining thirty four by the end of 2013.