4
COLLECTING DATA TO ENSURE EQUITY IN PAYMENT POLICY DISCUSSION HIGHLIGHTS

During a meeting with experts on collecting data to support equity in payment policy, various stakeholders, including the Center for Medicare & Medicaid Innovation (CMMI), shared their perspectives and engaged in a discussion on the critical steps necessary for transformation. Below is a summary of the multi-stakeholder perspectives and the discussion. For more information on the meeting’s content, please see the full Discussion Proceedings of this meeting in Appendix C. The meeting agenda and invited attendees are featured in Appendixes G and H.

• The landscape of data collection to support equity. Cara James, Grantmakers In Health, defined health equity as the fair and just opportunity to be as healthy as possible as measured by the reduction and elimination of health disparities and their determinants that affect marginalized groups. According to the Affordable Care Act (ACA), race, ethnicity, primary language, sex, and disability data are important to address health equity. However, data available to understand these aspects are severely lacking across U.S. government departments and agencies, except for data collected by Federally Qualified Health Centers. For example, the Centers for Medicare & Medicaid Services (CMS) reported that race and ethnicity data in Medicaid in 22 states are unusable. Data if often missing at the local level for people of color, people who have disabilities, older people, and those living with mental and behavioral health issues. James raised several considerations that could serve as areas for CMMI to target, incorporate, or consider as it embarks on its efforts to

• accomplish goals. These considerations include adding race and ethnicity data to the Medicare Part C and D application, incorporating available data toward payment options to inform value-based payment efforts and drive health care value and better outcomes, enhancing CMMI’s ability to collect, analyze, and report the demographic data necessary to monitor and evaluate the impact of programs and policies, providing technical assistance and resources to ensure seamless implementation of CMMI models at the state level, and engaging meaningfully the communities whose voices could be incorporated into these efforts, and conceiving and implementing a plan to build a usable data infrastructure. Finally, James suggested several policy actions to facilitate CMMI’s measurement related efforts. James suggested that CMMI use the multidimensional and comprehensive Health Equity Summary Score (HESS) data set as a template to incorporate population health factors across timeframes, populations, and benchmarks.
• Race and ethnicity. It remains uncertain how health systems can transform without substantial financial support to build electronic health record systems and train staff to collect race and ethnicity data. If data to support equity remain siloed or restricted to well-resourced health systems, many providers will struggle to utilize race and ethnicity data to inform treatment and interventions at the point of care. CMMI could include race and ethnicity data for billing for Medicare patients hosted on an interoperable electronic health database to build trust and proactively collect data to inform care delivery. It could also collect presently available data through private payers, existing collaboratives and registries, and health systems with advanced data-sharing capabilities. Ultimately, these efforts to standardize and collect race and ethnicity data can be leveraged to calculate patient risk, improve outcomes, and prevent emergency hospitalizations.
• Preferred language. Another dimension intersecting with race and ethnicity is preferred language. Collecting language data that is actionable at the point of care is essential to improving the cultural competence, experience, and qual-

• ity of care. By collecting and applying available data in this area, providers could work to enhance access to interpreter services and systems navigation. Furthermore, beneficiaries from diverse backgrounds would benefit through obtaining, understanding, and utilizing health information to make well-informed decisions for themselves or their next of kin. To support these advancements at a systems level, CMMI could provide guidance on collecting patient demographic and background questions. CMMI could also create unified expectations around the purpose and use of these data, especially related to its applications in population health and quality improvement.
• Disabilities. The conversation then focused on another overlooked area, the lack of disability status data and accommodations for disabled people. Additionally, optional standards for accessible medical equipment under the ACA make it challenging to provide high-quality care across the care continuum for people with disabilities. For example, people with disabilities might require different levels of support and treatment, such as physical accommodations for health screenings for people with different kinds of disabilities. CMMI could benefit from having qualitative and quantitative data illustrating the experiences of people with disabilities within health and health care systems. Additionally, CMMI could take an intersectional approach to understand health outcomes in various subpopulations who are more likely to experience worse health and screen positive for a social driver of health—for example, people of color from the LGBTQ+ community who also have disabilities.
• Sexual orientation and gender identity. Discreetly collecting sexual orientation and gender identity (SOGI) data is critical to protect the privacy, discretion, and trust of beneficiaries and improve care quality. To ensure both aims are achieved, CMMI could partner with other organizations and use existing government agency structures to collect SOGI data. To incorporate SOGI measures in existing data and diagnostic tools, CMMI could partner with the Bureau of Primary Health Care to amend the uniform data set to include SOGI

• data fields and eventually expand this to programs such as the State Innovation Models; use ICD-10 Clinical Modification measures for gender dysphoria, endocrine disorders, and hormone therapies to identify transgender beneficiaries; provide technical assistance and demonstrations via the National LGBTQIA+ Health Education Center on best SOGI data collection practices; leverage collaboratives and networks such as the CMS Quality Innovation Network-Quality Improvement Organizations Learning Action Networks to support and use SOGI data; and support clinical demonstrations and pilots in SOGI data collection.
• Social drivers of health. Another cross-cutting issue raised was the incorporation of upstream social drivers of health into CMMI models. Many CMMI models have included social drivers of health screening or navigation for factors such as housing, food, and income. However, while the CMMI Accountable Health Communities (AHC) model’s Year 1 Evaluation reported that 33% of beneficiaries screen positive for one or more drivers of health, only 14% of beneficiaries had a social need resolved after one year of support, with the main barrier being inadequate community resources. Within this sample, racial and ethnic minorities were overrepresented. The heightened attention to the drivers of health presents an opportunity for CMMI to include these measures in CMS quality and payment programs and models. CMMI could also continue its efforts to facilitate cross-referencing hierarchical condition category risk scores and health screening data to better understand the impact of social risk on cost. Ultimately, incorporating drivers of health measures in CMMI programs and models could help CMMI demonstrate the possibility of efficiently and effectively scaling up drivers of health services and programs and signal their wide-reaching impacts on the health and well-being of beneficiaries.

OPEN DISCUSSION

• CMMI’s Liz Fowler, Dora Hughes, and Kathryn Davidson began by affirming President Biden’s Executive Order on ad-

• vancing equity in payment policy and identified the challenges of limited resources and staffing as data collection challenges. CMMI is currently focused on launching successor models to Primary Care First and Comprehensive Primary Care Plus with considerations for data collection, geographic penetration, impact on underserved communities, specialty care, and bias in eligibility criteria and payment algorithms. Additionally, they are studying the impact of data attribution methodology, application criteria, and eligibility as barriers to CMMI’s 2030 goal of ensuring 100% of Medicare beneficiaries and that the vast majority of Medicaid beneficiaries are in an accountable provider relationship by 2030. Finally, they are clarifying the legal basis upon which they could mandate race, ethnicity, and language data collection and their ability to share these data with providers.
• Despite these efforts from CMMI, the agency’s resource constraints to coordinate multi-stakeholder efforts to collect population-data to support equity is a core concern. CMMI could issue a call to action with designated roles and responsibilities for field stakeholders to overcome this barrier. While designing new models and payment reforms, CMMI could publicly discuss available data and employ creative mechanisms such as secret shopper programs to evaluate provider quality of care. Additionally, CMMI could employ multi-stakeholder networks to aggregate complaints data, identify systemic problems, and use qualitative evaluations and analyses.
• Additionally, attendees highlighted that CMMI is best served by more actively involving stakeholders throughout its data collection and implementation efforts. For example, while designing new models and value-based payment innovations, CMMI could publicly discuss available data and employ creative mechanisms such as secret shopper programs to evaluate providers on the quality of care given to patients. Multi-stakeholder networks could be engaged and coordinated by CMMI to aggregate complaints data, identify systemic problems, and use qualitative evaluations and analyses.
• Attendees also encouraged CMMI to provide guidance, stan-

• dards, and training for providers to collect data and discuss identifying questions for critical factors such as disabilities, race, ethnicity, language, SOGI, and immigration status. These standards and guidelines could be based on the Institute of Medicine 2009 report on standardizing race/ethnicity and language data collection and disseminated across health systems, states, and providers. This effort would help increase provider confidence and build the patient trust needed to obtain critical data points.
• Beyond CMMI, attendees also raised that CMMI could help guide providers with more operational clarity and incentives to share, use, and apply data to support equity. By reducing administrative burden for providers, more capacity could be directed toward collecting and using data at the point of care. CMMI could also explore incentives, technical assistance, and other strategies to help providers overcome the high costs of creating electronically-enabled data collection systems and the staff and resources necessary to maintain them.
• In the process of data collection efforts to support equity, it is critical to engage behavioral health and developmental disability providers. Involving these providers could galvanize the field more comprehensively toward population-level health data collection and documentation efforts. However, these providers may require substantially more resources, support, and attention to improve their data infrastructure and technology to achieve progress in data collection. Some attendees noted that they have been omitted from previous data interoperability initiatives such as the Health Information Technology for Economic and Clinical Health Act of 2009. These considerations require attention to the pressing need for a strong and sustainable workforce, especially in the face of substantial health care worker burnout and shortages due to the COVID-19 pandemic, an increase in chronic disease and mental health needs, and the associated high rates of burnout, stress, and trauma in the U.S. population.
• With a supportive federal administration, new CMS leadership, and the fresh urgency of the nation’s reckoning with structural racism and inequities amidst the COVID-19 pan-

• demic, CMMI has an unprecedented opportunity to unite stakeholders by collecting population-level data to embed equity in payment policy. CMMI also has the opportunity to advocate for, measure, and pay for systems transformation and capacity-building efforts, strengthen financial incentives, and relax eligibility requirements.