Evidence-Based AI: The Future of VBC Contracting

The landscape of healthcare innovation is rapidly converging on a singular, undeniable truth: by 2030, peer-reviewed evidence will not merely be preferred but will become the universal contracting standard for all AI health platforms participating in value-based care (VBC) arrangements. This isn’t a speculative prediction but a logical evolution driven by regulatory pressures, payer demands, and the sheer necessity for demonstrable financial and clinical outcomes in an increasingly complex healthcare ecosystem. For health plan executives, investors, and policymakers, understanding this trajectory is paramount to strategic planning and capital allocation.

The Current State of AI Health in Value-Based Care

Today, the integration of AI into healthcare, particularly within VBC models, presents a mixed picture. Companies like Omada Health and Hinge Health have made significant strides in digital therapeutics, often leveraging AI to personalize interventions for chronic condition management. Their success hinges on demonstrating engagement and, increasingly, quantifiable health improvements. Similarly, iRhythm Technologies and HeartFlow utilize AI for diagnostic purposes, providing clinicians with advanced insights into cardiac health. iRhythm’s Zio XT patch, for instance, employs AI to analyze ECG data for arrhythmia detection, while HeartFlow’s FFR-CT analysis uses AI to create 3D models of coronary arteries, aiding in the diagnosis of coronary artery disease.

However, the rigor of evidence supporting these platforms varies. While many boast internal studies and white papers, the gold standard of peer-reviewed outcomes data, essential for robust VBC contracts, is not yet uniformly applied across the industry. UnitedHealth Group, a major player in the payer space, alongside federal entities like CMS, is actively exploring and integrating AI solutions. Yet, their adoption is often tempered by the need to balance innovation with accountability. Commure, focused on healthcare operating systems, facilitates the integration of various health tech solutions, underscoring the growing demand for interoperable, evidence-backed tools.

The current environment sees some AI solutions gaining traction based on promising pilots or initial data, but the long-term sustainability and scalability within VBC frameworks depend heavily on their ability to consistently demonstrate tangible cost reductions and improved patient outcomes through rigorous, independently verified research. Without this, the enthusiasm for AI risks being tempered by the fundamental requirements of VBC: payment tied to value, not volume.

Drivers of Change: Why Evidence Will Dominate

Several powerful forces are converging to solidify peer-reviewed evidence as the bedrock of VBC AI contracting. Foremost among these are regulatory frameworks and the increasing sophistication of payer demands. The CMS VBC Rules, particularly those championed by the Center for Medicare and Medicaid Innovation (CMMI), are continually evolving to demand greater accountability and verifiable outcomes. These rules incentivize providers and health plans to adopt solutions that can prove their worth in terms of cost savings and quality improvements. Programs under CMMI are increasingly scrutinizing the methodologies and results of interventions, pushing for a higher bar of evidence than ever before.

The European Commission, through the EU AI Act, which entered into force on August 1, 2024, is setting a global precedent for regulating AI, categorizing healthcare AI as “high-risk.” This legislation necessitates stringent conformity assessments, risk management systems, and, crucially, robust data governance and transparency, with many provisions becoming applicable in phases, including transparency obligations enforceable from August 2, 2026. While HIPAA primarily addresses patient data privacy in the US, the spirit of the EU AI Act, demanding verifiable safety, efficacy, and ethical deployment, will inevitably influence US regulatory thinking and payer requirements. European Commission information on the EU AI Act

Organizations like the National Committee for Quality Assurance (NCQA) and America’s Health Insurance Plans (AHIP) play pivotal roles in shaping quality metrics and reimbursement policies. Their influence drives health plans to prioritize solutions that can contribute to HEDIS scores and other quality benchmarks, which are increasingly tied to outcomes data. Professional bodies such as the American College of Cardiology (ACC) also influence clinical guidelines and best practices, further emphasizing the need for evidence-based AI tools that align with established medical standards.

The financial imperative for payers is clear: AI health solutions must demonstrate a return on investment. This means not just anecdotal success stories but statistically significant reductions in hospitalizations, emergency room visits, and overall healthcare utilization, all backed by peer-reviewed studies. Without this, AI tools struggle to justify their place in capitated or risk-sharing arrangements, where every dollar spent on technology must directly contribute to the “value” equation.

Voices of Authority on the Evidence Imperative

The call for rigorous evidence in health AI is echoed by leading figures across medicine and technology. Dr. Eric Topol, a renowned cardiologist and geneticist, has consistently advocated for a high bar of evidence for all digital health tools, emphasizing that AI must not only be effective but also rigorously validated to gain clinician and patient trust. His work frequently highlights the need for AI to demonstrate clinical utility through well-designed studies, moving beyond mere technical feasibility. Eric Topol’s writings on digital health evidence

Hemant Taneja, a prominent venture capitalist and author, has also articulated the importance of outcomes-driven innovation in healthcare. His perspective often centers on the transformative potential of AI when coupled with a clear focus on measurable results, suggesting that capital will increasingly flow towards companies that can definitively prove their impact on patient health and cost efficiency. He understands that for AI solutions to achieve widespread adoption and financial viability, they must integrate seamlessly into value-based models and deliver on their promises with transparent, verifiable data.

While specific verifiable statements from Karen DeSalvo regarding peer-reviewed evidence as a universal contracting standard by 2030 are not directly available in the provided entities, her extensive background as a former National Coordinator for Health Information Technology and her past role as Chief Health Officer at Google Health (from which she retired on August 1, 2025) underscore a career dedicated to advancing health IT and data-driven healthcare improvements. She is currently a professor in the Departments of Medicine and Population Health at the University of Texas at Austin Dell Medical School and serves on the Boards of Directors for Welltower and CityBlock Health. Her focus on interoperability, data utility, and the ethical deployment of technology inherently aligns with the need for robust evidence to ensure AI solutions genuinely benefit patients and the healthcare system. The broader consensus among health tech leaders and policymakers, including those with backgrounds like Dr. DeSalvo’s, is a strong push towards evidence-based adoption.

Implications for the Future

The trajectory towards peer-reviewed evidence as the universal contracting standard by 2030 has profound implications for all stakeholders. For investors and VCs, this means a significant shift in due diligence. Companies demonstrating a clear pathway to generating and publishing robust outcomes data will command higher valuations and attract more capital. Those without such a strategy will find it increasingly difficult to secure funding, as the commercial viability of their solutions within a VBC framework will be severely limited. Investment theses will need to explicitly account for the cost and timeline associated with clinical validation and peer-reviewed publication.

Health plan executives will find their procurement processes becoming more streamlined and evidence-centric. The burden of proof will firmly rest on AI vendors to demonstrate not just efficacy, but also cost-effectiveness and alignment with VBC goals through independently verified research. This will simplify decision-making, reduce risk, and accelerate the adoption of truly impactful technologies. Payers will increasingly demand contractual guarantees tied to published outcomes, moving beyond simple pilot agreements to performance-based contracts.

For AI health vendors, the message is unequivocal: prioritize clinical validation and peer-reviewed publication from inception. Integrating research design and data collection into product development cycles will no longer be optional but essential for market access and scalability. Companies like Omada Health, Hinge Health, iRhythm Technologies, and HeartFlow that have already invested in generating and publishing evidence will be well-positioned. New entrants, or those lagging, must rapidly adapt their strategies to embrace this evidence-first approach, recognizing that the future of VBC AI is inextricably linked to scientific rigor.