Main Library
Background: The Center for Medicare & Medicaid Services’ Center for Medicare & Medicaid Innovation tests service delivery models designed to improve care quality and patient outcomes while reducing or maintaining program costs. Historically, Innovation Center models were not explicitly designed to serve underserved populations, and most evaluations have not focused on health equity-related outcomes. In its 2021 strategy refresh, the Innovation Center prioritized health equity in all of its operations. To support the new goals, we conducted a retrospective analysis of model evaluations to determine the reach of Innovation Center models and assess the degree of and impact of health equity incorporation in model designs and evaluations.
Methods: We reviewed model evaluations underway or recently completed (date range: January 2018 – June 2022) and identified 17 with sufficient information to include in a health-equity focused synthesis. Models were separated into two groups: Group 1: models designed to address the needs of underserved enrollees, and Group 2: models that include underserved enrollees but are designed to address the population at large. Group 2 models were then separated into four categories, based upon the model’s health area of focus or payment intervention. Model evaluation reports were reviewed and evaluation leads consulted to determine the reach, impact (i.e. ability to assess and identify results), and experiences of the included models with respect to underserved populations. The most rigorous evaluations with respect to health equity-related outcomes employed a range of methods addressing enrollment, impacts, and experiences, specifically for non-white or lower income populations. Quantitative analyses used sources such as Medicare and Medicaid claims, other administrative data, health-related social needs screening results, and surveys. Many of these evaluations also collected primary data, including perspectives from individuals enrolled in Medicare or Medicaid, that illuminated challenges they faced.
Results: Our analysis identified means for evaluating health equity impacts and revealed gaps. Models designed to address the needs of underserved enrollees (Group 1) reached a higher proportion of racial and ethnic minorities, included the highest proportion of Medicaid enrollees, and all screened for health-related social needs. While Group 2 models were designed to address the population at large, some show potential for improving care and outcomes for underserved populations. Overall, many of the models encountered data challenges to addressing health equity including small population sizes, missing data on key elements (e.g., race) and categorization that does not capture critical subpopulations. Data opportunities include linking data sets to fill information gaps, targeted sampling in primary data collection, and identifying or developing instruments and protocols that account for health equity. Other promising practices include incorporating health equity priorities in model design, requiring participating entities to collect specific measures or beneficiary-level information, and employing multiple research methods in each evaluation to assess the impacts on and experiences of underserved groups.
Conclusions: The Innovation Center’s diverse portfolio ranges across populations, demographic subgroups, health care delivery systems, payment structures, and health conditions. Our study provides a foundation for assessing equity in prior health care models and for designing and promoting equitable quality improvement and evaluation.