The Effect of Network Adequacy Regulation on MA Networks

• In Event: Expanding Coverage, Ensuring Access: Challenges in Public Insurance

Abstract

Medicare, the primary public insurance program for elderly adults in the United States, serves over 65 million beneficiaries. Enrollment in Medicare Advantage (MA), which is administered  by federally-contracted private plans, has surged in the past decade, with more than half of Medicare beneficiaries participating in MA plans. A key distinction between MA and traditional Medicare is the use of restricted provider networks, which limit beneficiaries to a defined subset of providers.

To ensure sufficient access to providers given provider networks, the Centers for Medicare & Medicaid Services (CMS) dedicates significant resources to network adequacy regulation. However, empirical evidence on the effectiveness of network adequacy standards remains sparse, particularly within MA. This study addresses this gap by leveraging regulatory discontinuities across county borders to generate the first causal estimates of the impact of CMS’s network adequacy standards on MA provider networks.

We use publicly available data on county population and land area, the 2019 CMS network adequacy files, and 2019 Ideon Medicare Advantage provider directory data. CMS has five classifications of county urbanicity that are defined based on a combination of population levels and population density metrics. Each county classification has different requirements for the number of providers in each specialty and the allowable travel time and distance. More urban counties require more providers who are closer to residents than more rural counties. This creates four discontinuities in the restrictiveness of network standards.

We calculate the percent change in population that would increase or decrease each county’s classification. We identify counties within a 10 percent population change of a classification to each of the four discontinuities and use a regression discontinuity approach to estimate the causal effect of regulations on provider networks. Our key outcome measure is the number of contracted providers, across 23 regulated specialties. The regression discontinuity design includes fixed effects for specialties and county designation comparisons. We evaluate the overall effect across all specialties, explore heterogeneity across specialties, and test the sensitivity of our findings to multiple regression discontinuity specifications.

 

Our analytic sample includes 22,823 plan-counties-specialties across 300 counties. In the first stage, we find that an increase in a county’s designation leads to a 0.2 higher minimum provider ratio and a 19.2 mile narrower radius for providers to be included to meet the criteria. However, we find no relationship between increases in county designation and the number of contracted providers (-0.11, p-value = 0.94 in primary specification). Results are consistent and precise across all specialties and specifications.

Network adequacy is the primary tool that CMS uses to ensure sufficient access to providers in MA, which covers more than 50% of Medicare beneficiaries and a disproportionate share of Black and Hispanic beneficiaries. Our findings indicate that network adequacy standards do not affect network design, suggesting that CMS should consider alternative tools to improve access.

Authors

• Kelly E Anderson, University of Colorado Denver
  Non-Presenting Co-Author

• Matthew Lavallee, Johns Hopkins University
  Presenting Author

• David M Anderson, University of South Carolina
  Non-Presenting Co-Author

• Sih-Ting Cai, Indiana University
  Non-Presenting Co-Author

• Mark Katz Meiselbach, Johns Hopkins University
  Non-Presenting Co-Author