Recommendations for Monitoring Access to Care among Medicaid Beneficiaries at the State-level. Long-Term Recommendations to Improve the Data Infrastructure for State-level Assessments of Medicaid Access


The data sources currently available for measuring state-level access to care for Medicaid beneficiaries reveal several limitations for developing a robust system for monitoring access. There are several high-priority measures which the TEP identified that lack state-based estimates, and each of the existing survey sources for measuring consumer perceptions at the state-level have limitations. However, even if these state measures existed, there is no population household survey that has access measures for both adults and children which can be used to derive estimates in all fifty states and the District of Columbia. Current surveys provide access data that are robust only for healthy adults and for children in the thirty-two states made possible by 2011 NHIS sample augmentation (selected states), and only when the effective sample size of the Medicaid population is sufficient to provide precise estimates for policy analysis. There are many causes for concern, including access issues for subpopulations (such as people with disabilities and racial and ethnic minorities) and the inability to perform sub-state analysis. The ability of the data products to produce reliably comparable state estimates will be a challenge for surveys (due to sample design and sample size) and administrative data collected by the states (due to differences in data collection practices and definitions across states). Funding and infrastructure for data linkage work is insufficient to support current needs (e.g., the ability to link survey data with administrative records data—, such as linking provider surveys with provider characteristics information from administrative data, or linking household surveys with corresponding administrative data). Finally, T-MSIS implementation is not currently taking place in all states, and many states still contribute to MSIS only.

To improve the quality of the recommended indicators, several enhancements to existing data sources can be pursued. Additional funding for BRFSS will be required to include the question on type of health insurance coverage beyond 2013. Additionally, working with states to better understand what the data represent, given differences in sampling, survey administration, and periodicity of some questions, will provide additional insight into understanding variability in access to care issues across the states. The potential exists to add access questions to the American Community Survey (ACS), and some TEP members and their organizations were already considering suggestions. The TEP did not seriously consider the ACS as a data source because it was not perceived to have access measures, although it does have health insurance and income data with large annual state sample sizes. NCHS should be supported in its efforts to enhance the sample design of the NHIS to improve the survey's ability to make state estimates by 2016. Work could also be conducted with NCHS to explore the possibility of long-term support for increasing the sample size of NAMCS.

Continued support for the development, deployment, and adoption of T-MSIS will facilitate the measurement of Medicaid access in the future. Furthermore, CMS is developing additional datasets and systems which will complement T-MSIS. For example, the MACPro system is a new online system designed to facilitate submission of state applications to amend existing state plans or waivers, propose new options under the Medicaid and CHIP programs, and submit key administrative information. MACPro will modernize the CMS and state information exchange by offering an electronic workflow to CMS reviewers as well as our state partners. MACPro will be the official system of record for these submissions and changes. MACPro will be implemented in phases with the first release in early 2013. Additional Medicaid authorities will be included in future releases with all authorities incorporated into MACPro in 2014. MACPro will provide important contextual data to better utilize T-MSIS data by:

■        Enabling states to share program information with each other for benchmarking purposes

■        Providing structured data about a state's program which can be used to compare state programs and can be integrated and analyzed with expenditure data and granular operational data

Additional, long-term improvements could take the form of data linkage opportunities, sample size expansion, and the creation of surveys that specifically examine Medicaid access. Additional ways to work with the Census Bureau should be explored in order to improve researcher access to linked administrative and survey data within a secure working environment and ultimately to increase the utility of these linked data for health policy analysis. There is general agreement that many of the data sources contain the necessary access variables, but there is a need to expand the sample size of the NHIS and NAMCS. To provide a more robust understanding of consumer perceptions of access, a survey targeting Medicaid beneficiaries could be developed to facilitate state comparisons and attain the power to compare beneficiaries by key sub-populations (e.g., racial/ethnic minorities, disability status). The development of a state-based Medicaid Consumer Assessment of Healthcare Providers and Systems (CAHPS) could also be considered to support measurement of consumer assessments of Medicaid.

Additionally, TEP members and stakeholders recommend that ASPE and CMS continue working closely with states, not only in the development of the measures, but also once measures are in place. In an ideal system, HHS would share a report with each state that shows its data compared to national averages. When indicators are far outside the mean or moving in the wrong direction, HHS could engage with the states to determine the potential causes (e.g., data analysis issues, data interpretation issues, reporting issues, or programmatic/policy issues). This would not only improve understanding of the access issues, but also have the potential to improve claims data. As noted by one TEP member, "Data used is data improved."

View full report


"rpt_MedicaidAccessStudy.pdf" (pdf, 310.04Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®