{"subscriber":false,"subscribedOffers":{}} Buying Health For North Carolinians: Addressing Nonmedical Drivers Of Health At Scale | Health Affairs

Cookies Notification

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Find out more.


Integrating Social Services & Health

Buying Health For North Carolinians: Addressing Nonmedical Drivers Of Health At Scale

  1. Zack Wortman ([email protected]) is the chief of staff and deputy chief data officer for the North Carolina Department of Health and Human Services (NCDHHS), in Raleigh.
  2. Elizabeth Cuervo Tilson is the state health director and chief medical officer for the NCDHHS.
  3. Mandy Krauthamer Cohen is the secretary of the NCDHHS.
PUBLISHED:Free Accesshttps://doi.org/10.1377/hlthaff.2019.01583


Since 2017 the North Carolina Department of Health and Human Services has asked how its resources could be optimized to buy health, not only health care. This has led the department to incorporate whole-person care into all of its priorities, including building a statewide infrastructure and implementing incentives to address nonmedical drivers of health—focusing on food, housing, transportation, employment, and interpersonal safety/toxic stress. This article describes four interconnected initiatives that the department has implemented or is implementing to begin integrating medical and nonmedical drivers of health. This multifaceted effort involves many partners and includes financial incentives for commercial payers, Medicare, and Medicaid that are aligned with whole-person care; a standardized screening process to identify people with unmet social resource needs across all populations; NCCARE360, the first statewide network linking health care and human services providers to one another with a shared technology platform; and a large-scale Medicaid pilot to evaluate the impact of nonmedical health interventions on health outcomes and health care costs. North Carolina’s interconnected initiatives can help inform efforts around the US and generate needed evidence on how to implement systems through public-private partnerships to address nonmedical drivers of health at scale.


As health care payers and providers look for opportunities to better align health care costs and outcomes, a growing body of research is prompting many people to look outside the four walls of a hospital or practice and beyond the scope of what has traditionally been considered health care. There is clear evidence that having an unmet resource need—such as food insecurity, housing instability, unmet transportation needs, or interpersonal violence—can negatively affect health while also increasing health care use and costs.15

However, there is a lack of evidence about and experience in how best to address unmet social needs in the health care context at scale, which is inhibiting further policy development and investment. We believe that North Carolina’s work can meaningfully contribute to the implementation experience and research base needed to delineate a blueprint for implementing the integration of medical and nonmedical drivers of health at scale; better understand which interventions affect health, for which patients, and in what contexts; and define which payment models and incentives can succeed in promoting the adoption and sustainability of the changes needed to address nonmedical drivers of health as a routine part of health care delivery.6

Since 2017 the North Carolina Department of Health and Human Services (NCDHHS)—which oversees a Medicaid program with 2.2 million beneficiaries as well as a broad array of public health, behavioral health, and social services programs—has asked how its resources, and those of its partners, could be optimized to buy not only health care but also health. Using “buying health” as the primary lens for all policy and operational decisions has led the department to prioritize building a more coordinated and whole person–centered system that addressed both medical and nonmedical drivers of health. The department has incorporated whole-person health into every priority initiative—particularly its Early Childhood Action Plan, Opioid Action Plan, and the implementation of Medicaid managed care. This focus has also led the department to launch an interconnected set of initiatives to build the incentives and infrastructure needed to begin integrating medical and nonmedical drivers of health at scale.

North Carolina had several helpful preconditions in place or emerging when this effort began in early 2017. The North Carolina Hospital Association and North Carolina Medical Society had identified addressing nonmedical drivers of health as a priority and focused on the drivers at their annual meetings. North Carolina’s transition toward value-based payment in health care also began to accelerate in 2017, with rapid growth of upside and downside risk contracting in the Medicare and commercial health insurance markets. Additionally, the upcoming transition to Medicaid managed care provided a rare opportunity to redesign the state’s Medicaid program and test new interventions under a Section 1115 waiver, which was approved by the Centers for Medicare and Medicaid Services (CMS) in fall 2018.

Building The Infrastructure And Incentives

This article describes four interconnected initiatives that the NCDHHS has implemented or is implementing to begin integrating medical and nonmedical drivers of health. The initiatives are to better align financial incentives for health care providers and Medicaid health plans to address both medical and nonmedical drivers of health; introduce a standard screening process to identify people with unmet resource needs; establish NCCARE360, an innovative technology platform that allows health care providers and human services organizations to connect people with social resources and track the outcomes; and launch large-scale Healthy Opportunities Pilots to evaluate the impact of nonmedical health interventions on the health outcomes and health care costs of high-need Medicaid beneficiaries. The financial incentives and pilots are initiatives within North Carolina’s Medicaid program, while the screening process and NCCARE360 are intended for all populations. Together, these initiatives form the basis of the strategies that the NCDHHS is using to turn the concept of “buying health” into a reality.

As noted above, full implementation is still in process. The standardized screening tool was developed in 2018 and implemented in 2019. NCCARE360 implementation is under way. The platform is now fully operational in fifty of North Carolina’s hundred counties and will be operational statewide in 2020. The Healthy Opportunities Pilots are early in the implementation process and are expected to be fully operational in 2021.

The NCDHHS has focused on five domains of nonmedical drivers of health: food, housing, transportation, employment, and interpersonal safety/toxic stress. These domains were selected for three reasons. First, there was strong evidence that linked unmet needs in these domains to poor health outcomes, increased health care costs, or both.15 Second, both data and extensive stakeholder input identified significant unmet need in North Carolina in these domains.7 Third, these domains are similar to those of the Accountable Health Communities Model of the Centers for Medicare and Medicaid Services (CMS).8

Aligning Financial Incentives

Integrating medical and nonmedical drivers of health at scale is ultimately dependent on the degree to which entities have aligned incentives to buy health, not just health care. For example, the siloed and fractured environment of the fee-for-service payment system often disincentivized providers from delivering whole-person care. However, in a value-based payment environment, where the focus is on the total cost of care and health outcomes, financial incentives can drive insurers, health systems, and others to embed the nonmedical drivers of health in overall care delivery. Aligned financial incentives allow the state to direct dollars from other federal and state economic programs in a strategic and effective manner.

The adoption of value-based payment models has accelerated in North Carolina in the past three years. The state has experienced a proliferation of value-based payment arrangements in Medicare and now has thirty Medicare accountable care organizations.9 Building on that momentum, there has been a rapid growth of upside and downside risk contracting in North Carolina’s commercial market by Blue Cross and Blue Shield of North Carolina—particularly with its launch in 2019 of the Blue Premier program, a value-based contracting model in which seven major North Carolina health systems now participate. The move to Medicaid managed care in North Carolina creates an opportunity to use Medicaid, the state’s second-largest payer, to generate further movement to value-based payment.

Under the NCDHHS’s value-based payment strategy, Medicaid managed care organizations must make at least 50 percent of their total medical expenditures under value-based payment arrangements by the end of the second year of the contract period. Within four years nearly all Medicaid managed care contracts between health plans and providers must include a value-based component, with at least 45 percent of health plans’ medical expenditures involving upside risk and at least 15 percent involving downside risk.10

The NCDHHS also embedded financial requirements within Medicaid managed care to further align Medicaid health plans with whole-person care. Medicaid plans can count investments in nonmedical interventions (such as providing a healthy food box) and community resources (such as donating to a food pantry to increase its capacity) as part of their “medical loss ratio,” the percentage of their revenue that they must use for direct health care expenses. Plans that invest at least one-tenth of a percent of their total revenue in community resources additionally receive a modest preference in the auto-assignment algorithm for beneficiaries who do not proactively choose a health plan.11

These policies are designed not only to align financial incentives but also to create financial flexibility for providers and plans to invest in buying health for their populations.

Standardized Screening

In North Carolina, providers in health care and human services settings have not consistently asked patients about their nonmedical health needs. To address this, in 2018 the NCDHHS worked with a technical advisory committee of medical and social services experts to create a screening tool to identify patients with nonmedical health needs. (Online appendix exhibit 1 shows North Carolina’s screening tool.)12 The tool consists of nine simple yes-or-no questions. The questions map to four of North Carolina’s five targeted nonmedical health domains (food, housing, transportation, and interpersonal safety/toxic stress) and align with existing questions from validated national screening instruments that are used to identify unmet social needs. These instruments include the Hunger Vital Sign; the Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences; and the Accountable Health Communities screening.13

In a field test conducted at eighteen clinical sites in 2019, most patients (95 percent) and clinic staff members (97 percent) felt that the length of the screening was appropriate, that they understood the questions (93 percent of patients and 95 percent of staff members), and that they felt comfortable with the questions (92 percent of patients and 89 percent of staff members). Eighty-one percent of the staff members said that the questions were easily integrated into their work flows. Unexpectedly, most people who were screened for a need indicated that, for a variety of reasons, they did not want to be connected to a resource. The NCDHHS will release a best practice guide on screening in 2020 based on state and national experience, and it plans to conduct further evaluation.14

Standardization in screening allows for consistent data collection to inform policy decisions. North Carolina Medicaid will require its new managed care organizations to use this tool within ninety days of enrolling a new member and periodically thereafter, share information from the screening with the person’s primary care provider, incorporate the identified needs into care management risk stratification, and include social resource needs in written care plans.

The NCCARE360 Referral Platform

Standard screening processes would have limited impact in isolation, in part because many health care and human services providers may be reluctant to ask about needs they do not have the ability to address. When health care or human services providers identify a person with an unmet social need, they need a consistent way to refer that person to an organization that can intervene, and they need to know whether and when that intervention is actually provided. To meet that need, the NCDHHS and several nonprofit, private-sector, and philanthropic partners have implemented NCCARE360, the first shared statewide technology platform to link a network of health care and human services organizations.

NCCARE360 has three primary functions. First, it is a statewide community resource directory that includes a call center with navigators and a data team that verifies the accuracy of community resource information (for example, a resource’s address, hours, and types of services provided). Second, it is a community resource repository that integrates existing resource directories across the state. And third, it offers secure communications and data analytics functionality powered by Unite Us. The shared digital platform enables health care and human services providers to securely send and receive electronic referrals and learn whether a person had their unmet resource need met. This closed-loop referral functionality is embedded in electronic health records and other case management platforms and is designed for use across all populations—although its use is required for Medicaid health plans.

To illustrate how NCCARE360 works, consider a single pregnant woman who recently presented in crisis at the local office of the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC)—an assistance program that provides food and other supports. The woman’s hours had been reduced at work, which destabilized her economic situation. Though already enrolled in WIC, she did not have enough food or stable housing. The WIC office used NCCARE360 to quickly connect her to organizations that provided emergency food and assistance finding affordable housing, to link her to a program to support her long-term self-sufficiency, and to connect her to care management services during her pregnancy. Typically, it might have taken weeks to coordinate such services, with the onus of navigating a fragmented system on a woman who was already in crisis. The stress and lack of food could have contributed to a poor health outcome for her and a poor birth outcome for her infant.

NCCARE360 employs an extensive local implementation process that provides training and technical assistance for the organizations that join the network, while also fostering relationships among these organizations and driving the alignment of their efforts around shared goals and responsibilities.

The network’s twelve-week, multistep implementation process is led by regional community engagement managers, who hold a series of introductory, leadership, and strategy sessions in the community; train organizations; and help integrate NCCARE360 into organizations’ existing work flows. A manager stays in the community, helping additional organizations join the network and providing ongoing support and technical assistance.

NCCARE360 was launched in many North Carolina counties in 2019. As noted above, it is operational in fifty of the state’s hundred counties and will be operational statewide by the end of the year. A total of 741 organizations have joined the network, which has made 2,031 referrals for 1,214 unique clients, collectively. The plurality of referrals have been for housing (30 percent) and food (14 percent). The median time for a receiving organization to accept a referral is about twenty-four hours, and the median time to close a referral is nine days.15

A key design decision was building a single standardized statewide platform, given the problems that a fragmented referral platform landscape could pose. If different payers and health systems adopted different platforms, smaller organizations would likely need to use—and pay for—multiple platforms to accept the full array of referrals in their community. The NCDHHS feared that multiple platforms would mean multiple siloed networks, which would undermine the interoperability and communication that referral platforms exist to facilitate. The use of a single statewide platform also made it possible to analyze statewide data on both the supply of and demand for different community resources, which can help to match investment with the most acute needs.

Launching a single statewide platform such as NCCARE360 steers resources toward a shared effort and gives a large and varied group of organizations a stake in the platform’s success. The platform is directly administered by the nonprofit Foundation for Health Leadership and Innovation. A diverse array of philanthropic organizations, commercial payers, and health systems have also committed long-term financial resources to sustain the platform.

One early barrier has been slow uptake of NCCARE360 by some health systems. Conversations with health systems revealed concerns about the originally planned pricing model for licenses to use the platform, according to which a health system would receive a limited number of licenses and would pay for any additional licenses. To incorporate health system feedback, instead the platform will charge large organizations such as health systems and payers a flat fee for unlimited licenses. This will allow health systems to scale use up across their inpatient and outpatient systems and clinically integrated networks. Community organizations and social services agencies will not pay to use NCCARE360.

Another limitation has been the relatively low initial rate of referral acceptance, with only 56 percent of referrals (not including pending referrals) being accepted by the receiving organization. This appears to be driven partly by referring organizations’ lack of understanding of eligibility criteria. For example, an organization might refer someone with an unstable housing situation (such as sleeping on a family member’s couch) to an organization that provides services only to people who are literally homeless. This problem likely predates the implementation of NCCARE360, was occurring in informal and ad hoc referrals, and is now surfacing because of the ability for referrers to see in NCCARE360 that their referral was not accepted. The NCCARE360 team is exploring strategies to address this issue.

Though NCCARE360 is still being implemented and important issues need to be addressed for it to increase its reach and effectiveness, it is facilitating new communication between health care and human services providers and revealing solvable problems with existing informal referrals. It also provides a foundational piece of infrastructure upon which other initiatives, such as the Healthy Opportunities Pilots (described below), can be built.

Healthy Opportunities Pilots

The final major initiative in North Carolina’s effort is the Healthy Opportunities Pilot program, which—as noted above—is still being implemented but is expected to be fully operational in 2021. Authorized under North Carolina’s Medicaid Section 1115 waiver, the pilots will use Medicaid funds to pay for CMS-approved interventions in four domains: food, housing, transportation, and interpersonal violence/toxic stress. The waiver authorizes North Carolina’s Medicaid program to spend up to $650 million over five years to fund pilots operating in two to four geographic regions16 and requires rigorous evaluation of the impact of the interventions on health outcomes and health care costs.17 The pilots provide an unprecedented opportunity to test the efficacy of using health care dollars to pay for targeted, health-related social interventions. In turn, this will position North Carolina for statewide scaling of effective interventions in managed care and value-based payment arrangements.

To illustrate how the pilots will work, consider a woman enrolled in Medicaid who has chronic diabetes and experiences repeated hospitalizations due to uncontrolled high blood sugar. A standardized comprehensive screening conducted by her care manager reveals that she lives in a rural food desert and lacks access to affordable transportation, which makes it difficult for her to obtain fresh, healthy food. Through NCCARE360, the care manager can make a referral to a local food pantry. However, the food pantry is oversubscribed, and the woman lacks transportation to get there. Under the pilots, Medicaid can pay for a healthy food box to be delivered to the woman’s home for a set period of time, with ongoing reassessment. The funding helps remove the transportation barrier to obtaining healthy food, increases the likelihood that the pantry will have the resources to help her, and facilitates comparative evaluations of the degree to which the healthy food box improves health or provides a positive return on investment.

The pilots will operate regionally, with Medicaid managed care health plans authorizing and paying for pilot services from separately designated funds. Human services organizations in regional networks will enter into agreements with the state and health plans to provide approved services. To be eligible for the pilots, Medicaid enrollees must meet at least one state-defined health criterion (for example, being an adult with two or more chronic health conditions or a woman with a high-risk pregnancy) and at least one state-defined social risk factor. Appendix exhibit 2 outlines the eligibility criteria for the Healthy Opportunities Pilots.12 Reimbursement for twenty-nine approved services is set in a fee schedule published by the NCDHHS. Appendix exhibit 3 shows one sample entry—the home-delivered healthy food box—from the fee schedule.12

The ability to reimburse human services organizations for these nonmedical health services is critical. Payment helps ensure that often underfunded community organizations have the resources to consistently provide support when a referral is made. In turn, this creates an environment conducive to conducting rigorous evaluations of which nonmedical health initiatives improve health or lower health care costs in which contexts—evaluations that are currently lacking. An evaluation of the pilots will be conducted by the University of North Carolina’s Sheps Center for Health Services Research.

While it is too early to report results, designing the pilots has required the NCDHHS to develop an unprecedented degree of operational detail about how to deliver and pay for nonmedical health interventions using health care dollars. The pilots have required the state to create a delivery, payment, and evaluation infrastructure for health-related social interventions and to define roles and responsibilities for health plans, providers, and human services networks and organizations.

In addition, the fee schedule, for the first time, prices twenty-nine discrete health-related social interventions, drawing on an independent actuarial analysis, experience with existing Medicaid waiver services, and substantial stakeholder and expert input. The fee schedule describes the unit of service, payment amount, anticipated frequency and duration of the service, its setting, and the eligibility criteria for a person to receive it. These operational details can provide a starting point for others looking to launch similar efforts.


Interest in addressing nonmedical drivers of health is unlikely to result in long-term change without a clearer understanding of how to implement interventions at scale. Payers and providers will need more information to guide evidence-based investments not only in effective interventions but also in the infrastructure, incentives, and payment models needed to promote and sustain adoption. While it will still be a few years until North Carolina has outcomes data from these interconnected initiatives, the work to design and implement them offers important policy and operational details for others looking to explore implementation at scale. As this work moves forward, the NCDHHS will continue to evaluate and refine its approach and share its findings publicly.


The authors of this article are employees of the North Carolina Department of Health and Human Services and have been leading the design and implementation of the initiatives described in the article. The authors thank Erika Ferguson, Amanda Van Vleet, Emma Sandoe, Sejal Hathi, Tracy Zimmerman, Jessie Tenenbaum, Keval Desai, Rebecca Planchard, Elyse Powell, Melinda Dutton, and the Manatt Health team for the feedback and expertise they contributed to this article.


Loading Comments...