CMS, CMMI, NCQA: The Outcomes Data AI Needs for Value-Based Care

The promise of artificial intelligence in healthcare is vast, offering unprecedented opportunities for efficiency gains, improved patient outcomes, and, crucially for payers, significant cost reductions. However, for AI health platforms to truly integrate into value-based care (VBC) arrangements, they must navigate a complex landscape of regulatory expectations and demonstrate tangible, measurable value. The analytical question confronting health plan executives and policymakers today is clear: what level of outcomes data do CMS, CMMI, and NCQA genuinely expect before contracting, and how can AI solutions meet these rigorous demands?

The Imperative for Peer-Reviewed Outcomes in Value-Based Care AI

In the evolving ecosystem of value-based care, the shift from fee-for-service models necessitates a fundamental change in how healthcare technologies are evaluated and reimbursed. For AI health platforms, this means moving beyond mere technological prowess to proving quantifiable impact on patient health and financial performance. The Centers for Medicare & Medicaid Services (CMS), through initiatives like the Medicare Shared Savings Program, and its innovation center, the Center for Medicare and Medicaid Innovation (CMMI), are increasingly emphasizing outcomes-based metrics. Similarly, the National Committee for Quality Assurance (NCQA) sets standards that demand robust evidence of effectiveness for health plans and the solutions they employ.

The core expectation from these influential bodies is clear: AI health platforms seeking to participate in VBC contracts must provide peer-reviewed evidence, standardized metrics, risk adjustment, and population-level reporting. This rigorous evidentiary bar ensures that investments in AI translate into real-world benefits for patients and sustainable cost savings for the healthcare system. Without such data, AI tools risk being relegated to the periphery, unable to secure the trust and contracts necessary for broad adoption within value-based frameworks. As Karen DeSalvo, a prominent voice in health policy, has consistently highlighted, the emphasis must be on solutions that demonstrably improve health and healthcare value. Hemant Taneja, a leading venture capitalist, has also underscored the necessity of AI solutions proving their worth through concrete outcomes, particularly in areas ripe for disruption and efficiency gains.

Companies like Omada Health and Hinge Health have been at the forefront of demonstrating this commitment. Both have invested significantly in clinical validation, publishing peer-reviewed studies that detail their impact on chronic disease management and musculoskeletal conditions, respectively. This dedication to evidence aligns directly with the requirements set forth by CMS and CMMI, making them attractive partners for health plans like Cigna and UnitedHealth Group, who are actively seeking proven solutions to integrate into their VBC strategies. The ability to present robust, independently verified data on reduced hospitalizations, improved clinical markers, and lower overall healthcare costs is paramount. Such evidence directly addresses the concerns of payers looking to mitigate financial risk and improve population health outcomes, as outlined in CMS VBC Rules.

Navigating Regulatory Expectations: CMS, CMMI, and NCQA Standards

For health plan executives and policymakers, understanding the specific data requirements from key regulatory and accreditation bodies is critical. CMS VBC Rules and the Medicare Shared Savings Program are foundational. These regulations stipulate that any technology or intervention seeking to reduce costs or improve quality must demonstrate its efficacy through measurable outcomes. This often translates to a demand for data that shows reductions in utilization of high-cost services, improvements in preventive care, and better management of chronic conditions, all adjusted for population risk. CMS guidance on VBC data reporting

CMMI, through its various demonstration projects and payment models, further refines these expectations, often piloting innovative approaches to value-based care. Their focus is on scalable solutions that can generate significant savings while maintaining or improving quality. AI platforms must therefore be able to integrate with existing healthcare infrastructure, provide transparent data on their performance, and demonstrate a clear return on investment. This includes not just clinical outcomes but also financial performance metrics, such as reductions in claims costs or improved efficiency in care delivery.

The NCQA, as an independent non-profit organization that accredits and certifies a wide range of healthcare organizations, plays a crucial role in defining quality standards. NCQA Standards for health plans often include requirements for evidence-based interventions. For AI health platforms, this means demonstrating that their algorithms are built on sound clinical principles, that their interventions lead to measurable improvements in HEDIS (Healthcare Effectiveness Data and Information Set) measures, and that they contribute positively to patient experience and safety. The rigor of NCQA accreditation pushes health plans to adopt only those AI solutions that have proven their value through robust data. NCQA standards for digital health tools

The Role of Standardized Metrics, Risk Adjustment, and Population-Level Reporting

The demand for peer-reviewed evidence is intrinsically linked to the necessity of standardized metrics, appropriate risk adjustment, and population-level reporting. Without standardized metrics, comparing the efficacy of different AI platforms or even different interventions becomes nearly impossible. Organizations like the American Health Insurance Plans (AHIP) and the American College of Cardiology (ACC) advocate for common data elements and reporting frameworks to facilitate meaningful comparisons and ensure accountability. This ensures that the outcomes reported are consistent and interpretable across diverse settings and patient populations. For AI health platforms, this means aligning their data collection and reporting with established industry standards, allowing for seamless integration into existing VBC contracts and performance measurement systems.

Risk adjustment is equally vital. Healthcare populations are inherently diverse, with varying levels of chronic illness, socioeconomic factors, and access to care. To accurately assess the impact of an AI intervention, its outcomes must be adjusted for these baseline differences. This ensures that platforms serving sicker or more complex populations are not unfairly penalized, and conversely, that platforms serving healthier populations are not credited with outcomes that are simply a reflection of their patient mix. CMS/CMMI/NCQA require peer-reviewed evidence that utilizes robust risk adjustment methodologies. Finally, population-level reporting moves beyond individual patient anecdotes to demonstrate systemic impact. This is where AI health platforms can truly shine, showing how their interventions can improve health outcomes and reduce costs across entire cohorts, rather than just in isolated cases. The ability to aggregate and analyze data at scale is a core strength of AI, and its application in demonstrating population-level financial performance is a key differentiator for VBC contracting. (CW6-DP-15) AHIP white paper on AI in VBC

Key Takeaways for Health Plan Executives and Policymakers

The message for health plan executives and policymakers is unambiguous: the future of AI in value-based care is inextricably tied to verifiable outcomes. Investing in AI health platforms that lack peer-reviewed evidence, standardized metrics, robust risk adjustment, and comprehensive population-level reporting is a gamble that CMS, CMMI, and NCQA are increasingly unwilling to endorse. The regulatory landscape is clear: only those AI solutions that can demonstrably prove their value through rigorous, transparent data will secure a meaningful place in value-based contracts. As the healthcare industry continues its pivot towards outcomes-based models, the onus is on AI innovators to meet these elevated expectations, ensuring that technological advancement translates directly into improved health and sustainable financial performance.