Social Needs Integration

Social determinants of health have become healthcare’s most popular policy concept. Research consistently demonstrates that up to 80% of health outcomes derive from social and environmental factors rather than clinical care. This finding has launched a thousand initiatives: SDOH screening requirements, health-related social needs navigation programs, community information exchange platforms, and billions of dollars in investment to address the non-medical factors shaping patient health. The enthusiasm is palpable. The evidence is more complicated.

The distinction between social determinants and health-related social needs matters for understanding what healthcare can actually accomplish. Social determinants operate at the population level: income inequality, educational opportunity, neighborhood investment, environmental quality, structural racism. These require policy interventions far beyond healthcare’s scope. Health-related social needs operate at the individual level: a specific patient lacks food, cannot reach appointments, faces eviction, lives with domestic violence. Healthcare organizations have focused their attention here, where clinical workflows can identify needs and referral systems can attempt to address them.

RHTP applications universally embrace social care integration. Every state includes SDOH screening, HRSN navigation, or community information exchange platforms in some form. States committed substantial portions of their requested funding to these components. The premise is straightforward: identify patients with unmet social needs, connect them to community resources, resolve those needs, improve health outcomes.

The premise contains assumptions that evidence does not uniformly support. Most studies of SDOH interventions measure process rather than outcomes: screening rates completed, referrals generated, navigation contacts made. Fewer studies measure whether referrals result in services received. Fewer still measure whether services received improve health. And the rural evidence, where community resources are scarce and distances are vast, remains particularly thin. States implementing RHTP social care initiatives face a fundamental question the research has not definitively answered: does screening for social needs and referring to resources actually improve patient health, or does it merely document problems the healthcare system cannot solve?

The Rural Context

#

Social care integration developed primarily in urban safety-net settings where the model’s assumptions hold: dense populations enabling efficient navigation, robust networks of community-based organizations, proximity between patients and resources, organizational capacity for case management. Rural America systematically violates these assumptions.

Resource scarcity defines the rural social care landscape. The navigation model assumes a destination. Urban CHWs can connect food-insecure patients to food banks, housing assistance programs, transportation services, and social service agencies within their communities. Rural counties often lack these resources entirely or offer them at distances that create barriers as significant as the original need. A 2024 analysis of community-based organization density found that rural counties average 60% fewer social service organizations per capita than urban counties. Screening patients for food insecurity in a county without a food bank does not address food insecurity.

Geographic dispersion multiplies the challenge. A patient identified as needing housing assistance in frontier Wyoming may find the nearest housing counseling agency 90 miles away. Transportation-insecure patients, by definition, cannot reach services requiring travel. The closed-loop referral model that works in Philadelphia, where a community health worker can walk a patient to a food pantry, breaks down when the food pantry requires a 45-minute drive.

Agricultural paradox creates rural-specific food insecurity patterns. Counties producing abundant food often contain food-insecure residents who cannot access what grows around them. Farm workers experience the highest rates of food insecurity while harvesting crops they cannot afford. Produce prescription programs that work in urban food deserts may find no participating retailers in agricultural communities where the grocery store closed years ago.

Social isolation functions as a rural-specific social determinant largely absent from standardized screening tools. Depression, cognitive decline, and chronic disease management all worsen with isolation, yet rural residents routinely live miles from neighbors without transportation or communication technology connecting them to community. Screening tools developed for urban populations miss this fundamental rural social need.

Healthcare access itself constitutes a social determinant in rural areas. When the nearest primary care provider is 30 miles away, the nearest specialist 100 miles, and the nearest hospital an hour by ambulance, every other social intervention operates against this backdrop. Addressing food insecurity matters less if diabetic patients cannot access providers to manage their condition. The social care integration model assumes adequate clinical care exists; rural healthcare access gaps undermine this assumption.

Workforce desert limits who can perform social care functions. Effective SDOH screening and navigation require trained staff: social workers, community health workers, care coordinators, navigators. Rural facilities operating with skeleton staffing have no capacity for these roles. Adding social care requirements to clinical staff already working beyond capacity produces screening without intervention.

Evidence Review

#

Evidence Rating Table

#

The Evidence Problem

#

Social determinants research confronts fundamental methodological challenges that limit what we can confidently conclude about intervention effectiveness.

Most SDOH intervention studies measure process outcomes rather than health outcomes. A 2024 scoping review of 41 hospital-based SDOH initiatives found that the majority tracked screening completion rates, referral generation, and platform utilization. Far fewer measured whether referrals resulted in services received, and fewer still measured health outcomes like hospitalizations, emergency department visits, or clinical indicators. This evidence base tells us that healthcare organizations can screen patients and generate referrals. It does not tell us whether these activities improve health.

Randomized controlled trials remain rare in community-level social interventions. Randomizing neighborhoods to receive or not receive social investments raises ethical concerns. Pre-post study designs without control groups cannot distinguish intervention effects from regression to the mean. Observational studies comparing participants to non-participants face selection bias: patients who complete social care navigation may differ systematically from those who do not in ways that affect outcomes independently.

Follow-up periods in most studies are too short to capture long-term health effects. Addressing food insecurity this month may not measurably affect diabetes control for months or years. Housing stability requires sustained assessment over years to determine durability. Most SDOH intervention studies follow patients for weeks to months, missing the timeline over which social improvements would translate to health improvements.

Rural evidence is particularly sparse. The largest SDOH intervention studies operate in urban safety-net health systems: academic medical centers, federally qualified health centers in metropolitan areas, large integrated delivery systems. Extrapolating findings to frontier counties with fundamentally different social service landscapes requires assumptions the evidence does not support.

Food and Nutrition Interventions

#

Produce prescription programs represent the most rigorously studied food intervention category. A 2023 multisite evaluation of nine produce prescription programs across 22 sites in 12 states enrolled 3,881 individuals and found program participation associated with increased fruit and vegetable intake, reduced food insecurity, and improved self-reported health among adults and children. Clinical improvements included HbA1c reduction of 0.63%, BMI reduction of 0.36 kg/m2, and blood pressure improvements among adults with diabetes.

A 2024 microsimulation study modeling produce prescription implementation for diabetes patients nationally projected substantial health benefits and cost savings. The meta-analysis underlying this modeling found produce prescriptions increased fruit and vegetable consumption by 0.80 servings per day with statistically significant improvements in glycemic control.

Evidence strength varies by outcome. Food security improvements show consistent moderate effects across studies. Clinical outcome improvements appear primarily in patients with diet-sensitive conditions like diabetes. Broad population health improvements remain undemonstrated. Implementation matters: programs providing higher dollar values of produce subsidies for longer durations show stronger effects.

Rural implementation challenges limit produce prescription transferability. Programs require participating grocery stores or farmers markets that accept prescription vouchers. Rural food deserts lack these retailers. A 2022 pilot study of produce prescriptions in rural Michigan found participant enthusiasm but redemption challenges when the nearest participating vendor was 25 miles away. Online produce delivery addresses access in some areas but requires reliable internet and delivery infrastructure many rural areas lack.

Housing Interventions

#

Permanent supportive housing has the strongest evidence base among social interventions, though evidence comes almost exclusively from urban settings serving homeless populations with serious mental illness or substance use disorders.

A 2020 Lancet systematic review of 72 studies found permanent supportive housing significantly increased housing stability compared to usual care, with relative risk of being housed ranging from 1.23 to 1.42 depending on population. Housing First programs consistently outperform Treatment First approaches that require sobriety or treatment compliance before housing access.

Health outcome evidence is more limited and mixed. The review found no measurable effect on psychiatric symptom severity (10 studies), substance use (nine studies), or employment outcomes. Physical health outcomes showed inconsistent results across studies. Quality of life improved in Housing First participants compared to treatment as usual.

A 2024 Health Affairs study of Denver’s Supportive Housing Social Impact Bond found that two years after randomization, participants had six fewer emergency department visits than controls, eight more office-based psychiatric visits, and three more prescription medications. The program reduced emergency service utilization while increasing engagement with community-based care.

Cost-effectiveness evidence favors housing interventions through avoided emergency department, hospitalization, and criminal justice costs. Charlotte, North Carolina, documented $2.4 million in savings over one year from a Housing First program through reduced jail days and hospitalizations. However, societal cost savings do not consistently exceed intervention costs across all studies.

Rural applicability remains essentially untested. No permanent supportive housing studies identified in systematic reviews were conducted in rural settings. Rural homelessness differs from urban homelessness: more hidden, more episodic, more doubled-up housing rather than street homelessness. Whether housing interventions designed for urban chronically homeless populations transfer to rural contexts is unknown.

Closed-Loop Referral Systems

#

Community information exchange platforms like Unite Us, Findhelp, and state-specific systems like NCCARE360 enable healthcare organizations to make electronic referrals to community-based organizations and track whether referrals result in service provision. The “closed loop” describes knowing whether the referral was completed rather than sending patients into a void.

NCCARE360 in North Carolina provides the most extensive evidence base for statewide CIE implementation. A 2024 study comparing referral outcomes during periods with and without funding found dramatic differences: 88% referral resolution rate with COVID Support Services Program funding versus 30% resolution rate without dedicated resources. The technology platform alone does not produce resolution; funding for community-based organizations to actually provide services determines whether referrals translate to assistance.

A 2024 JAMA Network Open study introduced a distinction between closed-loop rate and successful connection rate. Closed-loop rate measures whether referrals were marked as closed in the system, a process metric. Successful connection rate measures whether closure resulted from actual receipt of services. In their analysis of Duke Health referrals through NCCARE360, closed-loop rates exceeded 90% in both study periods, but successful connection rates varied dramatically based on available funding and services.

Process versus outcome distinction matters critically for evaluating CIE effectiveness. Platforms demonstrably improve coordination, reduce duplicative screening, and enable tracking across organizations. Whether this coordination improves health outcomes rather than just improving referral processes remains undemonstrated.

North Carolina Healthy Opportunities Pilots

#

North Carolina’s Healthy Opportunities Pilots represent the most ambitious effort to date to test whether Medicaid-reimbursed social interventions improve health outcomes and reduce costs. The program reimburses 29 evidence-based, non-medical interventions addressing housing, food, transportation, interpersonal violence, and toxic stress for qualifying Medicaid members.

As of August 2025, the program has delivered over 265,000 services to over 20,000 enrollees in predominantly rural areas. Food services represent more than 85% of all services delivered. The program operates efficiently, with 95% of service authorizations approved and 93% of invoices processed.

The 2024 interim evaluation found encouraging results. A 2025 JAMA study using comparative interrupted time series analysis found that while Medicaid spending initially increased at enrollment, it subsequently decreased to $85 less per beneficiary per month compared to similar Medicaid beneficiaries ineligible due to county of residence. Emergency department visits and inpatient admissions showed declining trends for HOP participants.

Sustainability concerns cloud the program’s future. The 2025-2027 North Carolina state budget under negotiation as of January 2026 did not include funding for Healthy Opportunities Pilots ongoing operations or statewide scaling. New service delivery faced potential pause as of July 2025 pending legislative appropriations. The program demonstrates that social interventions can generate healthcare savings, but those savings accrue to Medicaid while intervention costs require state appropriation, creating a funding mismatch that threatens program continuity.

Technology Platform Landscape

#

Platform Comparison

#

Technology Limitations

#

CIE platforms require resources to refer toward. The most sophisticated technology cannot generate services that do not exist. Rural areas implementing CIE platforms discover the referral network contains few organizations because few organizations exist. A 2024 study of CIE implementation in rural Appalachia found that 42% of counties had fewer than five organizations participating in the regional network.

Resource directory maintenance presents ongoing challenges. Community-based organizations change services, eligibility requirements, hours, and contact information frequently. Maintaining accurate directories requires dedicated staff and ongoing verification. Stale directories generate failed referrals, erode user trust, and waste clinical time.

Platform interoperability remains limited. Most healthcare systems choose a single CIE vendor, but community-based organizations may participate in multiple networks. Patients referred through one platform may already be known to organizations through another. The fragmentation that CIE platforms were designed to address persists at the platform level.

RHTP Application Assessment

#

What States Proposed

#

Every state included SDOH/HRSN components in their RHTP applications, though specificity and investment levels vary dramatically.

Distinct social care initiatives with dedicated funding appear in roughly one-third of applications. Texas proposed specific CHW deployment for social needs navigation. Tennessee committed to expanding its Community Compass platform statewide. North Carolina planned HOP expansion to additional regions. These states treat social care as a funded priority.

Embedded social care appears more commonly. States mention SDOH screening and referral as components of broader care coordination or population health initiatives without distinct funding streams. California’s CalAIM Community Supports receives attention but rural RHTP integration remains unspecified. This approach risks social care becoming the first cut when implementation budgets tighten.

CIE platform investments vary from statewide procurement to regional hub-based approaches to no discernible platform strategy. Oregon and North Carolina operate statewide platforms. Texas and Kentucky show fragmented regional platforms. Mississippi and Georgia have major CIE gaps with limited RHTP investment to address them.

What Survived Partial Funding

#

Social care integration ranks among the most vulnerable RHTP elements. When states received less than requested funding, social care components faced disproportionate cuts.

A 2025 survey of rural hospital CEOs found social determinants work ranked around 15th on priority lists, behind workforce, technology, facilities, and operational concerns. This prioritization flows through to state implementation decisions. Georgia’s RHTP plan shows social care effectively absent from funded priorities. Ohio protected workforce and telehealth while leaving social care at risk.

The pattern makes operational sense but undermines RHTP’s comprehensive transformation intent. States that deprioritize social care investment perpetuate the screening-without-intervention problem: identifying needs without resources to address them.

Sustainability Planning

#

Few applications address what happens to social care infrastructure after RHTP ends. States proposing CIE platform procurement do not specify ongoing maintenance funding. CHW positions for navigation depend on grant funding without Medicaid reimbursement pathways established.

North Carolina’s Healthy Opportunities Pilots provide a cautionary example. Even with demonstrated cost savings and outcome improvements, the program faces potential service interruption due to state budget constraints. Social care generates Medicaid savings that accrue to federal and state Medicaid programs, but intervention costs require separate appropriation. This mismatch threatens every RHTP social care investment.

States with Medicaid expansion and mature managed care markets have more plausible sustainability pathways. MCOs can embed social care requirements in care management expectations. Non-expansion states lack this leverage and face greater sustainability challenges.

Implementation Reality

#

Screening Without Intervention

#

CMS mandated SDOH screening for hospitals beginning January 2024, requiring assessment of five domains: food, transportation, housing, violence, and utilities. The mandate does not require intervention. Healthcare organizations must identify needs; they need not address them.

The mandate creates risk of screening burden without patient benefit. Patients repeatedly asked about food insecurity, housing instability, and transportation barriers by multiple providers across multiple visits without receiving assistance may lose trust in healthcare. Screening without subsequent intervention documented in the medical record could create legal liability. The ethical implications of identifying needs an organization cannot address remain unresolved.

Rural facilities face this dilemma acutely. The mandate applies regardless of community resource availability. A critical access hospital in a county without food assistance programs must still screen for food insecurity, documenting unmet needs it cannot address.

Resource Capacity Gaps

#

Even where community resources exist, referral volume often exceeds service capacity. A food bank serving a county may have capacity for 200 families monthly. Hospital screening identifies 400 food-insecure patients monthly. The referral system works perfectly, yet half of identified patients receive no services.

Community-based organizations participating in CIE platforms report being overwhelmed by referral volume without corresponding increases in operating support. Healthcare systems refer; CBOs receive referrals; but CBO capacity constraints determine whether referrals translate to services. Platform technology cannot expand food bank warehouses or hire additional case managers.

Who pays for the social service remains the fundamental challenge. Medicaid can reimburse screening, navigation, and care coordination. Medicaid generally cannot reimburse the food itself, the rent payment, the utility bill. RHTP can fund platform infrastructure and navigation staff. Whether communities have resources to address identified needs depends on funding streams entirely outside healthcare.

The 2030 Cliff

#

RHTP social care investments face the same sustainability cliff as all RHTP components. Five years of federal investment cannot permanently transform social care infrastructure if states do not establish ongoing funding mechanisms.

CIE platforms require continuous maintenance. Without ongoing investment, resource directories become stale, organizations drop participation, referral networks atrophy. The infrastructure built by 2030 degrades rapidly without sustaining investment.

Navigation workforce depends on continued employment. CHWs trained for social needs navigation face the same post-grant employment uncertainty as clinical CHWs. States that build navigation capacity through RHTP without establishing Medicaid reimbursement pathways will see that capacity disappear when grants end.

The states best positioned for social care sustainability already have infrastructure: Medicaid expansion providing a reimbursement vehicle, mature managed care markets enabling MCO requirements, existing CIE platforms requiring only expansion rather than creation. States building from scratch face the steepest sustainability challenges.

The 2030 Question

#

Five years from now, what will RHTP social care investment have accomplished?

Optimistic scenario: States with funded social care initiatives demonstrate health outcome improvements sufficient to justify ongoing Medicaid investment. CIE platforms mature and expand. Community-based organizations develop sustainable funding models incorporating healthcare partnerships. Screening identifies needs that intervention addresses.

Realistic scenario: Process metrics improve dramatically. Screening rates increase. Referral volumes grow. Platform utilization expands. Health outcome evidence remains mixed, with some populations showing benefit and others showing no measurable impact. Sustainability depends on state-by-state policy choices with substantial variation.

Concerning scenario: States under budget pressure cut social care first. Screening mandates continue while resources to address identified needs stagnate or decline. Patients experience repeated social needs assessment without assistance. Trust in healthcare erodes. The gap between social care rhetoric and social care reality widens.

The evidence supports cautious optimism for targeted populations and specific interventions. Produce prescriptions for diabetic patients show clinical benefit. Permanent supportive housing stabilizes homeless populations. Comprehensive Medicaid programs like North Carolina’s generate measurable savings.

The evidence does not support universal enthusiasm for SDOH integration as healthcare transformation. Most interventions show process improvements without demonstrated health outcomes. Rural evidence remains sparse. Sustainability mechanisms are underdeveloped. The fundamental challenge, that social needs require social resources healthcare cannot provide, persists regardless of how efficiently healthcare identifies those needs.

States implementing RHTP social care components should:

Prioritize evidence-based interventions with demonstrated health outcomes rather than assuming all social care produces equivalent benefit. Produce prescriptions for diet-sensitive chronic conditions have stronger evidence than general food insecurity referral. Housing interventions for high-utilizer homeless populations show clearer benefits than housing screening for general populations.

Ensure resources exist before screening identifies needs. Platform technology and screening protocols are easier to implement than community resources to address identified needs. Expanding screening without corresponding resource expansion creates screening burden without patient benefit.

Build sustainability mechanisms concurrently with implementation. Medicaid reimbursement pathways, MCO contract requirements, state appropriation mechanisms must develop alongside RHTP-funded infrastructure. Waiting until 2029 to address sustainability ensures sustainability will not be achieved.

Acknowledge uncertainty honestly. The evidence base for many SDOH interventions remains preliminary, particularly in rural settings. States should evaluate their programs rigorously, contribute to the evidence base, and adjust approaches based on observed outcomes rather than assumed effectiveness.

Rural health transformation requires addressing social determinants. Whether the current approach of healthcare-based screening, referral, and navigation accomplishes this goal remains an open question the next five years of RHTP implementation may help answer.