Addressing Social Needs Saves Money and Improves Health in Medicaid. Now Scale It.
New evaluation results from CMMI reinforce findings from North Carolina that screening for and addressing upstream drivers of health reduces costs and improves outcomes — without cutting benefits.
Authors: Mandy Cohen and Melinda Dutton
Editors: Patti Boozang and Amanda Eisenberg
tl;dr
The final evaluation report of the Center for Medicare and Medicaid Innovation (CMMI)’s Accountable Health Communities (AHC) Model, released in late February, found that screening and navigating Medicaid and Medicare beneficiaries to address social needs like food insecurity, housing instability, violence, utility instability and transportation problems reduced emergency department (ED) visits, inpatient admissions, and total health care expenditures across 28 communities and more than one million individuals.
This comes on the heels of a 2025 peer-reviewed study published in JAMA finding that North Carolina’s Healthy Opportunities Pilots (HOP), the nation’s first comprehensive Medicaid program to pay for services addressing upstream drivers of health, saved $85 per participant per month and significantly reduced emergency department visits.
Together, these two bodies of evidence represent the strongest case yet for making services that address upstream drivers of health a permanent, scalable part of Medicaid.
Scaling requires action on multiple fronts: new federal Medicaid policy, state-level accountability, strategic use of the $50 billion Rural Health Transformation Program, and ultimately, Congressional action to make services addressing upstream drivers of health a permanent Medicaid benefit.
The 80 Million Impact
At a moment when the national conversation about Medicaid is dominated by cuts, coverage losses, and the mechanics of work reporting requirements, something important is at risk of getting lost: We now have rigorous evidence that Medicaid can spend smarter, improving health and reducing costs, by addressing what happens outside the “clinic walls.” Not by cutting benefits. Not by reducing provider rates. Not by kicking people off coverage. By connecting people to food, housing and transportation.
This is the story of two landmark evaluations — one federal, one state — that together provide the strongest evidence base we’ve had for making services that address upstream drivers of health a standard part of Medicaid. And it’s a call to action for policymakers, states, managed care organizations (MCOs), provider systems and Congress to stop admiring the evidence and start building the infrastructure to scale it nationally.
What the AHC Model Proved
The AHC Model was one of CMMI’s most ambitious experiments: a test of whether systematically screening Medicare and Medicaid beneficiaries for social needs and deploying community health workers to navigate them to services could reduce health care utilization and costs. The model ran across 28 communities in 21 states for five years, reaching more than 1 million individuals.
The final evaluation report, released in late February, confirmed what advocates and practitioners have long argued. Medicaid expenditures decreased by approximately 3% and Medicare expenditures by 4% for beneficiaries who received navigation services. The model was associated with reduced ED visits, fewer inpatient admissions and lower total health care spending. Notably, beneficiaries with multiple unmet needs saw the largest reductions in costs and the most pronounced improvements in quality.
Perhaps the most striking finding: These outcomes occurred even though the model achieved only moderate rates of social need resolution. Fewer than half of individuals receiving services reported that at least one need had been resolved, and around one in four reported that all their needs had been resolved. Navigation itself, or the act of helping someone understand and access available resources, appears to have independent value. As one evaluator put it, the very process of engaging a person in their own care and social circumstances changes how they interact with the health care system.
For populations that have historically been underserved, including racial and ethnic minorities, the model’s effects were even more pronounced, providing important evidence that addressing social needs can be a tool for reducing health disparities, not just costs.
What North Carolina’s Healthy Opportunities Pilots Proved
Full disclosure: We helped launch this program during Cohen’s tenure as secretary of the North Carolina Department of Health and Human Services. But the evidence speaks for itself.
When we designed HOP, the goal was straightforward: Test whether Medicaid could buy health, not just health care. The first Trump administration authorized up to $650 million in Medicaid funding under an 1115 Waiver to pay community-based organizations (CBOs) to deliver evidence-based services in four domains: food, housing, transportation and interpersonal safety. It was the first program of its kind in the country to do this at scale within Medicaid.
In February 2025, JAMA published the independent evaluation results confirming what the interim data had been showing: The program saved money and improved health. Specifically, health care spending declined by $85 per participant per month relative to a comparison group, with the program paying for itself by month eight after enrollment. ED visits fell significantly, by six per 1,000 person-months. Among those enrolled for 12 months or more, the reduction in ED visits was even more dramatic: 22 fewer visits per 1,000 participants per month. Eighty-nine percent of enrollees received at least one service, with food services (food boxes, produce prescriptions) comprising 85% of utilization.
The program ultimately served more than 38,000 Medicaid beneficiaries across 33 predominantly rural counties and delivered more than 819,000 services.
These are not marginal results. This is rigorous, peer-reviewed evidence that addressing social needs in Medicaid saves money while improving quality at meaningful scale, in some of the most economically challenged communities in the country.
Two Evaluations, One Conclusion
Read side by side, the AHC evaluation and the HOP evaluation tell a consistent and compelling story: When health care systems invest in screening people for social needs and connecting them to community resources, costs go down and quality improves. The AHC model proved this across a national footprint in both Medicare and Medicaid. North Carolina proved it in a real-world Medicaid managed care environment with actual service payments flowing to CBOs.
Together, these results should move the conversation from “does it work?” to “how do we scale it?”
The Scale-Up Agenda
Getting from demonstration to national standard requires action on several fronts simultaneously. Here’s what needs to happen:
Federal Medicaid Policy. While rescinding Biden-era Medicaid guidance on covering upstream drivers of health, CMS has indicated it will continue to consider proposals on a case-by-case basis. Existing 1115 approvals remain intact. More than 20 states and the District of Columbia received or were pursuing waivers to address upstream drivers of health under the previous framework. CMS should use the AHC and HOP evidence to develop updated, bipartisan guidance that creates a durable pathway for states to cover services addressing upstream drivers of health in Medicaid and can survive changes in administration. The evidence now supports moving beyond the demonstration model toward a scalable benefit design. The AHC model is one of the few CMMI demonstrations that generated net savings: a powerful case for expansion, not sunset. In addition, a few weeks ago the Medicaid and CHIP Child and Adult Core Sets Annual Review Workgroup voted to recommend the SNS-E quality measure (a HEDIS measure of screening and intervention for food, housing and transportation needs) for inclusion in the 2028 Core Sets. CMS now has the opportunity to follow that recommendation, providing important information on how effectively states and MCOs are promoting screening and intervention.
The Rural Health Transformation Program. Through H.R. 1, Congress allocated $50 billion to all 50 states for initiatives aimed at improving rural health care outcomes and access, with $10 billion available annually from 2026 through 2030. One of the program’s five strategic goals is to support evidence-based interventions to improve prevention and address root causes of disease, including food-as-medicine initiatives and chronic disease prevention. States should make the case to CMS that the infrastructure needed to address upstream drivers of health, including the screening platforms, referral networks, community health worker training and CBO capacity that programs like HOP depend on, qualifies as startup infrastructure that these funds can support. Rural communities, where social needs are often most acute and CBOs are least resourced, have the most to gain. North Carolina’s HOP results come disproportionately from rural counties. The Rural Health Transformation Fund is an opportunity to extend that kind of infrastructure to rural communities in every state.
State Accountability Through Managed Care Contracts and/or Value Based Care. More than 70% of Medicaid beneficiaries are in managed care delivery models. States have significant latitude to require MCOs to screen for social needs, invest in community health worker capacity, and partner with CBOs, all through managed care contract terms and quality metrics. States that are negotiating new MCO contracts or reprocuring plans should build screening, navigation, and service delivery addressing upstream drivers of health into the baseline expectations, not treat them as optional add-ons. North Carolina’s model embedded HOP within its managed care delivery system from the outset; that integration was a key factor in its success. States without managed care can build these features into value-based care arrangements. Duke-Margolis Institute for Health Policy researchers who have evaluated the program’s implementation underscored that sustainable financing models, CBO capacity building and managed care accountability are essential to making these programs work long-term.
Congressional Action. Ultimately, scaling services that address upstream drivers of health nationally requires Congress to act. Congress should authorize CMS to make social needs screening and navigation and targeted, evidence-based services covered Medicaid benefits, not as a waiver demonstration but as a permanent, optional state plan benefit. This would provide states with a stable, administratively simpler pathway to cover these services without the complexity and time limitations of 1115 demonstrations. It would also send a clear signal that the federal government views upstream drivers of health as a legitimate part of the health care continuum, not a policy experiment.
The Bottom Line
We are in a moment of extraordinary pressure on Medicaid. The program faces nearly $1 trillion in federal spending reductions over the next decade. States are scrambling to implement work requirements. Coverage losses are looming. Against that backdrop, it would be easy to treat services addressing upstream drivers of health as a “nice to have”: something to revisit when the fiscal picture improves.
That would be a mistake. The AHC and HOP evidence demonstrate that addressing upstream needs isn’t an add-on to Medicaid. It’s a strategy for making Medicaid work better: reducing avoidable utilization, improving quality, generating savings, and doing so in a way that disproportionately benefits the populations and communities that need Medicaid most. At a time when every dollar in the program matters, this is one of the few proven approaches to spending smarter.
The demonstrations worked. The evidence is peer-reviewed and published. The question now is whether we have the policy will to scale what been proven.