Implementation of a Hub-and-Spoke Partnership for Opioid Use Disorder Treatment in a Medicaid Nonexpansion State

Expanding access to evidence-based interventions to address the opioid crisis, especially medications for patients with opioid use disorder, requires significant innovation at the health care system level. Challenges include high rates of uninsured patients, unintegrated substance use disorder treatment facilities, geographic disparity in medication access, systemic stigma, and insufficient educational opportunities for providers (1–3). Medicaid expansion has funded evidence-based opioid use disorder treatment and community-based psychosocial interventions for individuals who are either uninsured or enrolled in Medicaid, but not all states have elected to apply for expansion (4).

In Medicaid expansion states, hub-and-spoke (H&S) partnerships have successfully delivered opioid use disorder treatment by leveraging newly available funding to incentivize innovation and collaboration among agencies to support care for previously uninsured patients (1–3, 5). In the H&S model, an anchor site (hub) provides comprehensive and specialized care, complemented by satellite partner sites (spokes) offering more limited services (5). H&S partnerships have successfully expanded the pool of providers and treatment centers employing evidence-based treatments for opioid use disorder through hub-delivered training (1, 2) and set specific patient enrollment targets for H&S networks (3).

We describe implementation of an H&S network in Tennessee, a Medicaid nonexpansion state, and how we have met challenges not seen in Medicaid expansion states. To our knowledge, this is the first report of the application of the H&S partnership in a Medicaid nonexpansion state.

Landscape of Opioid Use Disorder Treatment

Tennessee and other southeastern states have had higher rates of opioid use per capita than other U.S. regions (6). With the 2012 Prescription Safety Act, Tennessee expanded use of the Controlled Substance Monitoring Database, published prescribing guidelines, and increased regulation of pain management clinics (6). Despite these efforts, between 2012 and 2016, Tennessee’s opioid overdose rate increased dramatically, from an age-adjusted rate of 11.0 to 18.1 incidents per 100,000 residents (6). Concurrently, access to evidence-based opioid use disorder pharmacotherapy was low: Tennessee legislators overruled a 2016 federal law allowing nurse practitioners and physician assistants to prescribe buprenorphine, and Tennessee Medicaid did not reimburse all forms of medication for opioid use disorder (7). Tennessee lawmakers declined Medicaid expansion in 2015, limiting resources for one of the largest payers for substance use disorder treatment (7). In 2016, 61% of Tennessee patients who needed psychiatric services could not access them (7).

State Opioid Response–Funded H&S Partnership in Tennessee

Funded by State Opioid Response (SOR) grants in 2018 (8), the state mental health department invested in an H&S model with four regional hubs, each partnering with between four and seven spoke agencies (see table in the online supplement to this column). Parallel to the treatment service H&S model, an education-focused extension for community health care outcomes (ECHO) tele-education H&S model was established to provide educational support for community partners.

Middle-Tennessee Treatment Hub

As a regional hub for addiction services, Vanderbilt University Medical Center (VUMC) houses most requisite levels of opioid use disorder care. Specific services include a general hospital with an interdisciplinary addiction consultation service; an inpatient, co-occurring disorder–focused psychiatric unit; an intensive outpatient program for individuals with co-occurring disorders; a low-barrier bridge clinic for individuals with substance use disorders who are discharged from acute care settings; and a longitudinal outpatient clinic for patients with co-occurring disorders. Overall, this emphasis on comprehensive, “med-psych integrated” care is designed to support treatment of patients with opioid use disorder and the most complex conditions. To date, the hub has enrolled 110 uninsured patients. Patients had a mean age of 43 years, and 74% (N=81) were male and 70% (N=77) were White. On average, patients traveled 27 miles (range 2.3–200) each way to VUMC. Nine patients (8%) patients obtained insurance after grant enrollment.

VUMC currently lacks certain levels of addiction care, including longer-term residential and outpatient methadone dispensation (our hub is licensed to offer only buprenorphine and intramuscular naltrexone in the outpatient setting). As we will discuss, this limitation required creation of partnerships with community providers—beyond the usual referral relationships for insured patients—to ensure effective transitions for the uninsured population.

The middle-Tennessee hub leverages SOR funding to support a team of clinicians whose salaries are supported by a blend of federal-to-state grant funds, third-party reimbursement, and hospital gap funding. Our primary clinical team—comprising addiction psychiatrists, other physician specialists in our bridge clinic (infectious diseases, pain medicine, and internal medicine), psychiatric nurse practitioners, social workers, recovery coaches, and nurse case managers—draws on the Massachusetts nurse case manager model and collaborative care model (9).

VUMC also serves as the statewide project ECHO hub, funded with SOR funds. Project ECHO is a model for technology-enabled education and mentoring meant to address the national shortage of a trained health care workforce, particularly in underserved areas (10). Spoke-based providers attend twice-monthly tele-education sessions and can access hub-curated materials via an online learning platform. In the 2020–2021 academic year, VUMC held 18 education sessions with a mean enrollment of 22 trainees per session.

Middle-Tennessee Spoke Partners

Spokes (see the online supplement) offer a variety of services; four are nonprofit agencies offering residential and intensive outpatient psychosocial services (intensive outpatient program or partial hospital program), one is a federally qualified health clinic (FQHC), and one is an opioid treatment program dispensing methadone and buprenorphine. Spoke partners were selected by our state mental health agency because they had a history of serving uninsured populations (all providers); working with underrepresented minority groups, especially historically Black communities (Matthew Walker Community Health Center [MW] and Meharry Medical College [MMC]); working with peripartum women (Mending Hearts); and offering medication for opioid use disorder and other addiction services (Buffalo Valley, MMC, Samaritan Recovery Community, and Behavioral Health Group). The partnerships with MW and MMC (see online supplement) have launched an important trend toward addressing health disparities by sharing resources between a larger academic medical center (VUMC) and a smaller, historically Black medical school (i.e., MMC) and one of its historically affiliated FQHCs (MW).

In our H&S system, spoke partners contract directly with the state, whereas the hub administers the social services budget for the network and employs all grant-funded staff. The hub was also consulted on the establishment of spoke sites, made site visits to several spokes during contracting, and strongly supported FQHC inclusion in the partnership. The operational aspects that cement these partnerships primarily involve real-time troubleshooting of referral challenges, communication about grant-related administrative matters, and the ECHO H&S network itself. However, state-sponsored incentives to promote communication and collaboration are lacking, and, as a result, many spoke agencies continue to operate in isolation. This has provided a valuable lesson for subsequent iterations of SOR funding. In the next round of funding, spokes will be incentivized through microgrants that will be contingent on meeting enrollment targets and on greater ECHO participation. These microgrants will fund case manager effort at the spokes. In addition, hub recovery coaches will also spend more time at the spoke sites, which should support relationships among our partners and facilitate patient care across our network.

Despite the aforementioned challenges to collaboration, an unanticipated benefit of the ECHO hub was a mutual understanding of the treatment philosophy and implementation processes between the hub and spokes through regular educational forums. Clinical leaders at one spoke partner improved their ability to identify quality gaps, in one instance revising inflexible policies requiring buprenorphine discontinuation as an automatic response to substance misuse. For some agencies, the partnership also supported a transition from abstinence-based models of treatment to lower-barrier, medication-integrated harm-reduction approaches more closely aligned with national guidelines.

Care Delivery: Medicaid Expansion States Versus Nonexpansion States

H&S in a Nonexpansion State

Using federal grant funding for opioid use disorder to pay for treatment required establishing independent documentation and billing systems and locating funding for care of patients with both opioid disorder and co-occurring disorders. In contrast, Medicaid expansion does not typically require significant changes to documentation and billing systems, nor does it exclusively cover a single disease category.

Parallel Documentation and Billing

For documentation and reimbursement, our grant requires use of an additional, government-contracted, Web-based system. Initial enrollment in this system requires a 60- to 90-minute patient encounter to capture multiple data points, including an Addiction Severity Index assessment. This enrollment must be repeated during any transition between partner sites regardless of how recently it was completed. Documentation in this database is also required for all subsequent clinical encounters. This duplicate layer of documentation required the development of new processes, particularly the creation of additional billing codes within our electronic medical record system to track grant-funded patients for reconciliation.

Of note, time-consuming enrollment into this system created a barrier to enrolling acute patients leaving our hospital or emergency department, which represent critical care transition points for individuals with opioid use disorder. To address this challenge, our middle-Tennessee hub site provides uncompensated care for several weeks during acute stabilization and engagement but before official grant enrollment. Although costly, we have found that this model is necessary to retain patients in treatment during the transition to outpatient settings.

Payment for Co-Occurring Disorders

The SOR grant pays only for opioid use disorder care, posing a clinical challenge because a significant proportion of patients have co-occurring disorders. These patients often cannot afford private insurance and do not meet the strict Medicaid inclusion criteria in Tennessee (e.g., having dependent children or a permanent disability). In response, we created processes to provide unreimbursed, medical and psychiatric care for conditions other than opioid use disorder. Fortunately, our medical center had long-established processes for charity care, including a medication assistance program and an Office of Outpatient Referral Assistance (OORA) to either locate compatible community providers or approve unreimbursed care. We estimate that >90% of our SOR-funded outpatients receive medications for chronic conditions other than opioid use disorder and therefore enroll all patients in the medication assistance program. Approximately 50% require referrals to community FQHCs, and <5% require access to internal OORA referrals, in part because our transitional bridge clinic is itself a multispecialty facility. With the exception of one spoke partner, MW, an FQHC, most spokes must outsource general medical and psychiatric care to regional FQHCs and community mental health centers.

Access to Full Continuum of Care

Although our middle-Tennessee H&S network offers multiple levels of care on the addiction continuum, grant funding for care of inpatients with co-occurring disorders is not included in SOR, and availability of residential and sober living facilities that accept SOR funding is limited, especially facilities that support medication for opioid use disorder. This shortfall puts greater pressure on grant-funded outpatient clinics to treat patients with needs requiring higher levels of care. In our clinical experience, many patients are unable to stabilize in the outpatient setting because of a lack of recovery support or suitable living conditions, a situation that can have negative implications for the program environment and staff morale.

Discussion and Conclusions

H&S partnerships for the management of opioid use disorder leverage public funding to address a stigmatized and largely underfunded public health problem (5). Although previous H&S models in a Medicaid expansion context have yielded valuable proof-of-concept results for facilitating improved patient access, partnerships among agencies, and educational cooperation, H&S implementation in a Medicaid nonexpansion context has not been previously described. Our experience in Tennessee underscores some common themes from other H&S models, such as the value of funding administrative leadership, seed funding of novel clinical approaches, and developing project ECHO hubs (1–3). We have also observed unique obstacles, including duplicate billing and documentation costs, lack of funding for co-occurring disorders, and challenges in accessing high-quality inpatient and residential care.

These findings have implications for policy makers and other medical centers. Present funding levels are inadequate to meet the needs of the uninsured population with opioid use disorder. Under Medicaid expansion, the Centers for Medicare and Medicaid Services has supported expansion both by extending coverage to young adults without children with incomes ≤138% of the federal poverty level and by approving Section 1115 and Section 1915 waivers of the Social Security Act that allow states additional flexibility to cover addiction services and recovery supports. Even for covered individuals, TN Medicaid pays for the continuum of addiction services but has not yet applied for any waivers that would allow for expansion of home- and community-based supports such as sober living or other long-term recovery supports for patients with substance use disorders. These resource limitations have minimized the utility of systematic enrollment for TN Medicaid–eligible patients within our programs.

We recommend future value-based cost analyses to quantitatively compare Medicaid expansion and targeted-grant approaches in order to address the opioid crisis. Health care systems will be most motivated to pursue grant funding for infrastructure innovation if public insurers like Medicaid are empowered to offer new payment models. Conversely, obstacles to integrating grant-funded opioid use disorder care into large health systems may discourage such innovation and contribute indirectly to systemic stigma toward this vulnerable patient population.