AI in VBC: Why Long-Term Outcomes Drive Contract Success

The promise of AI in healthcare is vast, yet its true impact on value-based care (VBC) hinges not on ephemeral pilot results, but on demonstrable, multi-year outcomes data. For health plans and clinicians navigating the complex landscape of AI-driven solutions, distinguishing lasting value from market hype requires a critical eye toward the duration and rigor of clinical evidence. This distinction is paramount for VBC contracts, where financial performance is inextricably linked to sustained improvements in patient health and cost reduction.

The Imperative of Long-Term Evidence in Value-Based Care AI

The Centers for Medicare & Medicaid Services (CMS) and its innovation center, CMMI, increasingly emphasize long-term accountability in VBC models. This regulatory context, alongside guidelines from organizations like the National Committee for Quality Assurance (NCQA), the American College of Cardiology (ACC), and the American Heart Association’s Journal (JAHA), frames the essential requirement for AI health platforms to provide evidence of sustained efficacy and financial performance. Many AI health companies, particularly those in their nascent stages, operate with 2-3 year histories, offering limited evidence that often stems from short-term pilots. However, VBC demands multi-year data, reflecting the chronic nature of many conditions and the long-game strategy of cost management. Consider the challenge of algorithmic drift, where an AI model’s performance degrades over time as real-world data distributions shift away from its training data. A cardiac AI trained on 2018 to 2020 data, for instance, may exhibit decreased accuracy by 2026 due to demographic shifts or evolving clinical practices. research on algorithmic drift in medical AI Payers need assurance that an AI solution will not only perform well initially but will maintain its accuracy and impact over the lifespan of a VBC contract. This necessitates ongoing validation and, ideally, a Predetermined Change Control Plan (PCCP) with the FDA, allowing AI/ML devices to make predefined modifications without new premarket submissions, thus ensuring continuous relevance and safety.

Benchmarking Leaders: iRhythm Technologies and HeartFlow

When evaluating digital health platforms, the duration and quality of evidence become a critical differentiator. Two companies in the cardiac space provide an excellent contrast in their approach to evidence generation and market penetration. iRhythm Technologies, with its Zio patch, commands over 70% of the US Long-Term Cardiac Monitoring (LTCM) market share, generating $825.34 million in trailing twelve-month revenue as of Q2 2026. Their success is deeply rooted in a substantial “data moat”, millions of labeled ECG recordings that make it nearly impossible for new entrants to match their diagnostic accuracy. This vast dataset allows for continuous model refinement and robust real-world evidence (RWE) generation, which is crucial for demonstrating long-term clinical utility and cost-effectiveness. The Zio patch’s ability to provide continuous, extended monitoring for arrhythmias has allowed for earlier detection and intervention, contributing to reduced downstream costs associated with stroke and other cardiac events. In contrast, HeartFlow, which achieved a $364 million IPO and $212 million in trailing twelve-month revenue as of June 30, 2026, has built a formidable patent thicket around CT-FFR technology. While HeartFlow boasts over 600 publications in cardiac CT diagnostics, the focus here is often on diagnostic accuracy and initial clinical utility. The challenge for VBC contracts lies in translating this diagnostic precision into demonstrable, sustained reductions in long-term cardiac treatment costs and improved patient outcomes over several years. While a precise diagnosis is invaluable, payers require evidence that its application leads to measurable changes in care pathways, fewer invasive procedures, and ultimately, a lower total cost of care over the contract’s duration. HeartFlow clinical evidence publications

Beyond Cardiac: The Broader Landscape of Outcomes-Driven AI Health

The need for multi-year outcomes data extends across the digital health spectrum. Companies like Omada Health, Hinge Health, Noom, and Pear Therapeutics, while operating in different therapeutic areas, all face the same imperative to prove long-term value. Omada Health, with a $150 million IPO and a broad digital chronic care platform, offers programs for diabetes prevention, hypertension, and mental health. For VBC contracts, Omada must demonstrate not just initial engagement or short-term biometric improvements, but sustained weight loss, blood pressure control, and mental health stability over multiple years. This requires robust data collection on adherence, clinical markers, and healthcare utilization over extended periods. Hinge Health, a digital musculoskeletal (MSK) health leader that currently has a market capitalization of $6.91 billion and achieved a $437 million IPO, claims a 2.4x ROI in MSK digital health. While impressive, payers will scrutinize the duration over which this ROI is realized and whether it translates into sustained reductions in surgeries, pain medication use, and improved functional outcomes for patients over several years. Short-term pain relief is one thing; preventing chronic pain and avoiding costly interventions years down the line is another. Noom, known for its behavioral change programs, and Pear Therapeutics, a pioneer in prescription digital therapeutics (PDT) with FDA-cleared solutions for substance use disorder (SUD) and insomnia, exemplify the varying levels of regulatory rigor and evidence expectations. Pear Therapeutics, as a SaMD (Software as a Medical Device), underwent a stringent FDA review process, including clinical trials to support its efficacy claims. However, Pear Therapeutics filed for Chapter 11 bankruptcy in April 2023 and ceased operations, highlighting the challenges of commercialization and reimbursement in the PDT space despite initial FDA clearances. The challenge for all these platforms is to move beyond pilot successes to consistently demonstrate durable clinical benefits and financial savings that align with the multi-year horizons of VBC contracts.

The Payer’s Perspective: What VBC Contracts Demand

For health plan executives, the decision to integrate AI health platforms into VBC models is fundamentally an economic one, intertwined with clinical efficacy. The question isn’t merely “Does it work?” but “Does it work consistently and cost-effectively over the long term?” This economic impact analysis directly addresses the investor prompts regarding lowering long-term cardiac treatment costs, reducing cardiovascular risk, and managing hypertension costs. CMS VBC rules, often interpreted through the lens of organizations like the ACC and JAHA for cardiac care, demand evidence of sustained outcomes. This includes:

• Reduced Hospitalizations and ER Visits: AI interventions should demonstrably decrease acute care utilization for chronic conditions over multiple years.
• Improved Clinical Markers: Sustained improvements in metrics like HbA1c, blood pressure, and cholesterol levels, validated over extended periods.
• Medication Adherence and Optimization: Evidence that AI tools improve adherence and help optimize medication regimens, leading to better control and fewer complications.
• Patient Engagement and Retention: High, sustained engagement rates are critical for long-term impact.
• Total Cost of Care Reduction: The ultimate metric, requiring comprehensive claims data analysis over the contract term. As prominent cardiologists like Eric Topol and Valentin Fuster have consistently argued, the integration of AI into clinical practice must be underpinned by robust, peer-reviewed evidence that withstands the test of time. Without this, AI risks becoming another fleeting trend rather than a transformative force in healthcare.
  

  Conclusion

  The healthcare AI market rewards companies that combine regulatory clarity, published outcomes, and revenue durability. This pattern is evident in the enterprises that have successfully navigated the complexities of VBC. For health plans and clinicians, the message is clear: prioritize AI health solutions that demonstrate long-term clinical performance through multi-year outcomes data, not just promising pilot results. This rigorous approach ensures that investments in AI genuinely contribute to the goals of value-based care: better patient outcomes at a lower cost, sustained over time.