The Shift to Outcome-Based AI: How the CMS ACCESS Program is Redefining Healthcare Reimbursement
For decades, the American healthcare system has operated on a simple, if flawed, premise: providers are paid for the time they spend with patients. This “fee-for-service” model creates a ceiling for innovation, as there is little financial incentive to implement tools that improve health outcomes between office visits. That is about to change.
The Centers for Medicare & Medicaid Services (CMS) has launched ACCESS (Advancing Chronic Care with Effective, Scalable Solutions), a 10-year program designed to test the viability of AI-driven medical care at a federal scale. By shifting the payment structure from activity-based to outcome-based, the government is effectively creating “swim lanes” for AI innovation in one of the most heavily regulated industries in the world.
- Outcome-Based Payments: ACCESS rewards providers when patients meet measurable goals (e.g., lower blood pressure) rather than for the number of check-ins performed.
- AI Integration: The program provides a reimbursement mechanism for AI agents that monitor patients, coordinate referrals, and ensure medication adherence.
- Targeted Care: The initiative covers critical chronic conditions including diabetes, hypertension, obesity, depression, anxiety, and chronic kidney disease.
- Scalability: Companies like Pair Team are using AI to reach vulnerable populations who face social determinants of health, such as housing and food insecurity.
Pair Team: Bridging the Gap Between Medicine and Social Stability
Among the 150 participants chosen for the first cohort of ACCESS is Pair Team. Founded in 2019 by Neil Batlivala, the company focuses on a patient population often ignored by Silicon Valley: those managing chronic conditions while simultaneously battling unstable housing, food insecurity, or a lack of transportation.
Pair Team’s core philosophy is that health outcomes cannot improve without addressing the full context of a patient’s life. This community-integrated model blends medical, behavioral, and social care. The effectiveness of this approach is backed by peer-reviewed evidence; a study co-authored by Pair Team researchers and published in the Journal of General Internal Medicine found that this model led to strong patient engagement and significant reductions in avoidable emergency room and inpatient hospital utilization.
According to Batlivala, the impact is stark: one in four hospital visits and one in two ER visits are avoided when a patient is under Pair Team’s care. To date, the company has raised approximately $30 million from investors including Kleiner Perkins, Kraft Ventures, and Next Ventures, and has grown its revenue to above nine figures.
Flora: The AI Agent as a Clinical Intervention
While human teams are essential for complex care, they are difficult to scale. To solve this, Pair Team deployed Flora, a voice AI agent that serves as the primary patient-facing interface. Flora operates 24/7, handling intake, coordinating referrals, and performing the frequent check-ins necessary to keep patients engaged between clinical visits.
Beyond the administrative efficiency, Flora provides a critical psychological component: companionship. Batlivala notes that for isolated patients—such as those living in their cars—hour-long conversations with Flora can be the only social interaction they have for weeks. In these cases, the AI isn’t just a tool for efficiency; it becomes a clinical intervention in its own right.
A New Philosophy at CMS
The design of the ACCESS program reflects a shift in leadership at CMS. The program was architected by Abe Sutton, Director of the CMS Innovation Center, and Jacob Shiff, Chief AI and Technology Officer of the CMS Innovation Center. Both bring startup backgrounds to the federal government—Sutton as a former venture capitalist and Shiff as a former healthcare founder.
Their influence is evident in the program’s structure, which emphasizes:
- Competition: A push for the “best solution” to win based on performance.
- Direct-to-Consumer Enrollment: Streamlining how patients enter the program.
- Lean Operations: Reimbursement rates are intentionally set low to ensure that only organizations running efficient, AI-first operations can remain financially viable.
Risks: Data Privacy and Financial Sustainability
Despite the potential, the intersection of AI and federal healthcare infrastructure carries significant risks. The most pressing is data security. Participants are feeding sensitive patient data—including intimate details about mental illness and housing status—into federal systems that have a documented history of breaches, including the exposure of Social Security numbers.
There are also systemic financial concerns. The track record of the CMS Innovation Center has been inconsistent. A 2023 analysis by the Congressional Budget Office (CBO) revealed that the center increased federal spending by $5.4 billion during its first decade, failing to produce the projected savings.
The Road Ahead
The ACCESS program goes live on July 5, marking a pivotal moment in the digitization of Medicare. Pair Team currently has partnerships providing access to roughly 500,000 potential patients, with a goal of reaching one million within three years.
If successful, ACCESS will prove that AI can do more than just automate paperwork—it can manage the complex, messy reality of chronic care for the most vulnerable members of society. By aligning payment with actual health improvements, the government is finally incentivizing the “lean, AI-first” operations necessary to make healthcare scalable and sustainable.
Frequently Asked Questions
What is the main difference between traditional Medicare and the ACCESS program?
Traditional Medicare typically reimburses based on the time a clinician spends with a patient. ACCESS uses an outcome-based model, paying providers when patients meet specific health goals, such as reduced pain or lower blood pressure.
Which health conditions does the ACCESS program cover?
The program focuses on chronic conditions including diabetes, hypertension, obesity, depression, anxiety, and chronic kidney disease.
How does Pair Team’s AI, Flora, assist patients?
Flora is a 24/7 voice AI agent that handles patient intake, coordinates referrals, and performs check-ins to maintain engagement between clinical visits, providing both medical coordination and companionship.