Abstract

Context: In 2015, the Centers for Medicare and Medicaid Services (CMS) urged state Medicaid programs to use 1115 waiver demonstrations to expand substance use treatment benefits. We analyzed four critical points in states' decision-making processes before expanding benefits.

Methods: We conducted qualitative cross-case comparison of three states that were early adopters of the 1115 waiver request. We conducted 44 interviews with key informants from CMS, Medicaid, and other state agencies, providers, and managed care organizations.

Findings: Policy makers expanded substance use treatment in response to “fragmented” care systems and unsustainable funding streams. Medicaid staff had mixed preferences for implementing new benefits via 1115 waivers or state plan amendments. The 1115 waiver process enabled states to provide coverage for residential benefits, but state plan amendments made other services permanent parts of the benefit. Medicaid agencies relied on interorganizational networks to identify evidence-based practices. Medicaid staff secured legislative support for reform by focusing on program integrity concerns and downstream effects of substance use rather than Medicaid beneficiaries' needs.

Conclusions: Decision-making processes were influenced by Medicaid agency characteristics and interorganizational partnerships, not federal executive branch influence. Lessons from early-adopter states provide a road map for other state Medicaid agencies considering similar reform.

In 2015, the Centers for Medicare and Medicaid Services (CMS) invited state Medicaid agencies to test their ability to implement new service delivery models that promote treating substance use disorders (SUDs) as chronic diseases (Wachino 2015). In a letter to state Medicaid directors, CMS highlighted the urgency of improving substance use treatment to combat the national opioid epidemic, which has disproportionately impacted Medicaid beneficiaries. More than 12% of Medicaid beneficiaries have a diagnosed SUD (Donohue, Raslevich, and Cole 2020), and Medicaid beneficiaries account for nearly 50% of all opioid-related deaths in states at the epicenter of the opioid epidemic (McMullen 2016; WVVIPC 2017). In response to the growing need for substance use treatment, CMS called on states to design full evidence-based SUD care continuums including early intervention, outpatient treatment, residential treatment, recovery support services, and medications to support individuals throughout their recovery process (Wachino 2015). States willing to test new service delivery models could do so by applying for a waiver under Section 1115 of the Social Security Act.

Section 1115 waivers allow states to conduct five-year demonstration projects to experiment with alternative ways to use Medicaid dollars to deliver services (CMS 1994). Although waivers enable Medicaid agencies to tailor services to state needs, CMS reserves authority over which waiver initiatives it promotes and approves. Section 1115 waivers allow executives from the federal government (e.g., presidential administration, CMS leaders) and state government (e.g., governor, Medicaid commissioners/directors) to sidestep Congress and state legislators to implement Medicaid program changes that are more responsive to evolving conditions than is often possible to achieve through the legislative process. Political executives at the federal and state levels can use 1115 waiver authority as a tool to administer Medicaid programs in ways that accommodate state preferences and political agendas without requiring changes in law (Thompson 2013; Thompson and Burke 2007). Presidential administrations may exert executive federalism by inviting states to apply for waivers, such as CMS's 2015 request for states to implement evidence-based substance use treatment service delivery models. However, state executives must accept the federal government's invitation (Thompson 2013). Despite the proliferation of state waiver demonstrations, little is known about states' internal decision-making processes for pursuing waiver-based systems transformation (Wright et al. 2018). Examining Medicaid agencies' pursuit of the 2015 1115 “SUD waiver” reveals critical insights into how policy makers within state Medicaid agencies respond to federal invitations to revise state benefits, initiate and formulate policy, and engage with state legislatures.

Medicaid agencies that participated in the 2015 SUD waiver opportunity were required to transform their benefits in significant ways. Medicaid agency staff had to discard existing SUD treatment benefits that in many states had not changed in decades, and to select evidence-based practices (EBPs) aligned with national standards for clinical care (Wachino 2015). The SUD waiver allowed states to make more comprehensive changes than would be possible through the state plan amendment process. However, waiver authority also invited greater CMS oversight and scrutiny. Medicaid agencies' ability to achieve comprehensive benefit transformations required reaching organizational and political consensus on the need to improve SUD services, selection of EBPs, and the policy vehicle for reform.

We aimed to better understand the internal decision-making processes and experiences that state Medicaid staff and their colleagues engaged in when pursuing reform. Why did state Medicaid agency leaders and other relevant stakeholders decide to enhance SUD benefits? Why did Medicaid agencies opt to use the SUD waiver, rather than state plan amendments, as a vehicle for implementing new services? How did policy makers select new EBPs for planned SUD care continuums? Once plans were in place, how did Medicaid agencies secure state matching funds to pay for new services? We address these questions by comparing three states' SUD waiver experiences as natural experiments in enhancing benefits for Medicaid beneficiaries.

Overview of Prewaiver Medicaid SUD Service Variation and Political Environment

There are few federally imposed requirements on what state Medicaid agencies must include in their plan benefits (Boozang et al. 2014). Consequently, there is significant variation in Medicaid-reimbursable SUD services across states. A national survey found that 10 states did not reimburse for intensive outpatient services, and other states limited access to SUD services by requiring prior authorization, annual maximum limits, or copayments and deductibles (Grogan et al. 2016). Access is further limited by a dearth of providers willing to deliver care to Medicaid beneficiaries because of low reimbursement rates (Gordon 2018; Rosenbaum 2014) and stigma against this population (Allen et al. 2014).

The 2015 SUD waiver opportunity coincided with two major health care challenges: Medicaid expansion and the opioid epidemic. In 2015 many states were debating whether to expand Medicaid eligibility under the Affordable Care Act. At the time only 26 states and Washington, DC, had adopted Medicaid expansion (KFF 2020). Controlling Medicaid costs was a major point of debate across states (Grogan, Singer, and Jones 2017). Nationally, Medicaid was already the largest payer for behavioral health services, with annual mental health and SUD treatment costs estimated at $67 billion (SAMHSA 2014). Research suggested that expanding Medicaid eligibility criteria would almost double the number of beneficiaries receiving SUD services (Busch et al. 2013).

Simultaneously, the urgency of the national opioid epidemic presented some states with a compelling reason to invest in improving substance use treatment benefits within their Medicaid programs. Medicaid beneficiaries are disproportionately represented in SUD prevalence and opioid-related overdose rates. Additionally, substance use-related diagnoses are among the top 10 most common reasons for Medicaid hospital readmissions within 30 days (Hines et al. 2014; Weiss and Jiang 2021). The SUD waiver provided states with an opportunity to improve the quality of SUD care and, potentially, health outcomes.

Selection of States as Case Studies

To date, 28 states have approved SUD waivers, and six states are awaiting waiver approval (table 1). Fifteen states are implementing comprehensive reform to create full SUD care continuums; 19 states are using waiver authority for limited benefit changes. Examples of limited reform include reimbursing for residential treatment without seeking to develop a complete care continuum or adding a few new services to a nearly complete existing care continuum (e.g., Massachusetts).

We identified California, Virginia, and West Virginia as ideal case study states. These states were early adopters of the SUD waiver, planned comprehensive SUD reforms, and were engaged in SUD service benefit implementation at the time of data collection. Examining how decisions were made in states at a similar implementation stage allows for a better understanding of key factors within the inner and outer contexts that prompted the SUD waiver process.

California was the first state to pursue and implement the SUD waiver in 2015 and thus served as a test case for CMS and a model for other states (CHCF 2018). Virginia and West Virginia began implementing SUD waivers in 2017 and 2018, respectively (CMS 2016, 2017). Before the SUD waiver, all three states had limited SUD benefits and planned to use the SUD waiver to implement comprehensive SUD care continuums (table 2), which would require significant changes to their benefits and possibly their budgets (VDMAS 2016; Douglas 2014). Additionally, all three state Medicaid agencies contracted managed care organizations to administer benefits, and therefore they had similarly complex contract and stakeholder relationships to consider when reforming covered services or reimbursement rates.

States were also selected for their contrasts in organizational characteristics and size. California's 58 county welfare departments are responsible for locally administering Medicaid benefits. Conversely, Virginia and West Virginia Medicaid programs are administered at the state level. California and West Virginia are Medicaid expansion states that cover nearly a third of their respective state populations with Medicaid benefits (CMS n.d.[a]). Before adopting the SUD waiver, Virginia had not expanded Medicaid, and only 12% of Virginians were Medicaid beneficiaries (CMS n.d.[a]). At the time states applied for their SUD waivers, California's governor was a Democrat, and Democrats controlled the state legislature. Virginia state leadership was politically divided, with a Democratic governor and a majority Republican legislature. West Virginia had recently elected a Democratic governor, who switched to the Republican party six months into his term, right as the state was seeking SUD waiver approval. Additionally, Republicans controlled the West Virginia state legislature. West Virginia's SUD waiver application preparation mostly occurred under the Obama administration while the state's approval-seeking process lingered into the Trump administration. Comparing SUD waiver experiences in California, Virginia, and West Virginia enabled us to explore the potential influence of partisan politics on SUD states' decision-making processes.

Data Collection and Analysis

We conducted 44 semistructured interviews with key informants from each state Medicaid agency, other state agencies (e.g., behavioral health agencies), Medicaid managed care organizations (MCOs), Medicaid providers, consulting partners working with Medicaid agencies, and CMS staff involved in states' waiver application processes. We reviewed agency and policy documents to generate a list of individuals with SUD waiver or benefit responsibilities whom we initially contacted for interviews. Snowball sampling techniques were used to identify additional stakeholders with key insights on state decision-making processes (Green and Thorogood 2014). Interviews inquired about each states' internal decision-making processes, experiences interacting with state leadership and CMS, as well as other stakeholders involved in pursuing SUD reform via the 1115 waiver. Data collection continued until thematic saturation was achieved for each state, and additional data collection would not yield new insights (Guest, Bunce, and Johnson 1995; Saunders et al. 2018). All interviews were recorded and professionally transcribed. Two qualitative researchers independently double-coded 25% of the transcripts to develop and refine a qualitative code book. Passages were coded to constructs related to states' broad policy context (e.g., state leaders, opioid epidemic), Medicaid agencies' inner context (e.g., organizational priorities), and bridging factors (e.g., intermediaries) linking both contexts (Aarons, Hurlburt, and Horwitz 2011). Once coding consensus was achieved, one researcher coded the remaining transcripts. Qualitative themes were derived from thematic analysis and consensus discussions with the larger research team. This study was approved as exempt by the Boston University Medical Campus Institutional Review Board (H-38990).

Why Did Policy Makers Decide to Expand SUD Services?

Policy makers were motivated to expand SUD services for four reasons: (1) existing Medicaid SUD benefits were limited, (2) benefits that did exist were difficult to access because of inadequate provider networks, (3) relying on block grant funds to pay for SUD services was unsustainable, and (4) the opioid epidemic disproportionately impacted Medicaid beneficiaries in Virginia and West Virginia.

Stakeholders referred to Medicaid SUD treatment service environments as “addiction treatment deserts” because benefits were sparse and Medicaid service providers were difficult to access. Across states, key informants said accessing medications for opioid use disorder (MOUD) was challenging. For example, in California, methadone providers did not operate in every county, meaning beneficiaries had to travel significant distances to access MOUD.

Low Medicaid reimbursement rates resulted in inadequate provider networks, further hindering beneficiaries' access to services. In Virginia and West Virginia, low reimbursement rates disincentivized prescribers to contract with Medicaid, and beneficiaries frequently had to pay out of pocket for MOUD. One Virginia Medicaid MCO representative said, “The payment levels for Medicaid substance use were so low that nobody did it. . . . Although methadone was covered, 98% of people basically paid out of pocket.” One Medicaid agency representative described reimbursement rates as so low that “you could make more waiting tables than you could providing addiction treatment to Medicaid members.” Another state agency representative said low rates discouraged providers because “you can go to Wal-Mart or anywhere else . . . and make more money. And, you know, and the work would be easier.”

Key informants described states as having “patchwork” and “cobbled together” treatment systems that blended Medicaid and state safety net funds to ensure access to care. Before the SUD waiver, none of the Medicaid programs reimbursed for residential treatment for nonperipartum adults. West Virginia created access to a residential-like treatment by having Medicaid reimburse facilities for intensive outpatient treatment, while the state behavioral health agency used block grant dollars to pay for room and board. Medicaid staff described this approach as financially unsustainable and poor-quality care. Without Medicaid reimbursement for residential treatment, providers used subjective logic to ration access to treatment:

“Counties used their substance abuse block grant funding to buy as much residential treatment as they could. What that generally meant was, you sort of parsed it out carefully over the course of a year and you bought really small, like, short lengths of stay. You typically wouldn't fund multiple trips to residential for the same person, because like, you know, counter to all best practices [they] had, you know, ‘proven’ themselves not able to make the change. . . . It just was a terrible system where everybody felt terrible about it. You're constantly sort of deciding who gets the golden ticket, and how much could they get.”—California Medicaid Staff

Before waiver approval, the statewide prevalence of diagnosed SUD was 8.1% in California, 7.2% in Virginia, and 5.8% in West Virginia, compared to a national average of 7.5% (SAMHSA 2019). However, these figures do not account for undiagnosed individuals living with SUD or substance misuse. In California, stakeholders said the opioid epidemic was not a prominent factor influencing SUD. As one provider said, “The opiate crisis didn't hit the public airways, if you would, in a large way until after California already made the decision” to pursue the SUD waiver. However, the opioid epidemic was a significant catalyst for Virginia and West Virginia Medicaid agency staff exploring the need to improve SUD services. Key informants were skeptical that any changes to their benefits would have occurred without their states experiencing major increases in opioid-related overdoses among Medicaid beneficiaries. One Virginia Medicaid MCO representative said, “I don't think anything would have happened . . . if it wasn't for the opioid crisis. . . . Nothing speaks louder than people dying.”

Why Did Medicaid Agencies Use 1115 Waivers Instead of State Plan Amendments to Enhance Their SUD Benefit?

State Medicaid agencies can define reimbursable services within a state plan amendment (SPA) or within a waiver authority. All states have an SPA that describes how the Medicaid agency defines eligible populations and administers services in ways that are compliant with federal rules (CMS n.d.[b]). SPAs must be reviewed and approved by CMS whenever the state makes changes to program policies (CMS n.d.[b]). Waivers allow states to relax some federal statutory Medicaid program requirements with the goal of testing out innovative approaches for administering Medicaid benefits in ways that are tailored to state needs. Waivers grant states flexibility but require lengthier application and review processes than SPAs, and they are viewed as time-limited service delivery experiments (MACPAC 2016). State Medicaid agencies primarily used 1115 waivers to enhance SUD services because the waiver allowed them to reimburse for residential treatment, which was not otherwise permitted. Despite the flexibility granted by 1115 waiver authority, none of the states embedded all of their SUD services within the waiver.

Across states, Medicaid agency leadership differed in their preferences for using waivers and SPAs to transform benefits. As one key informant said, “Please don't do an 1115 if you don't have to,” citing the administrative burden of developing, managing, and negotiating waivers with CMS, and the temporary nature of waiver services. Virginia Medicaid leaders had a strong preference for including benefits in their SPA to make them permanent. Ultimately, they only included residential services in their SUD waiver. A Virginia behavioral health agency stakeholder said, “We were fortunate enough that we did not need to use an 1115 for the entire benefit. We embedded it permanently in our program.” Conversely, stakeholders at the West Virginia Medicaid agency favored waiver authority. Medicaid agency staff believed that revising the SPA exposed the state to federal scrutiny: “A lot of times we have found if you open up a SPA that's been in place for a while, you know, you can have detrimental effects from that as well. You could possibly lose things that you've had in place.” One West Virginia Medicaid representative described having previously edited the SPA only to have CMS investigate all aspects of their services. As a result, Medicaid leadership said “you take a risk” when you edit the SPA. West Virginia ultimately split their new services between the SPA and the 1115 waiver.

Before the SUD waiver, state Medicaid agencies were unable to reimburse most residential treatment. The Medicaid Institutions for Mental Disease (IMD) exclusion prohibits states from using federal Medicaid dollars to pay for services in facilities with more than 16 beds. The SUD waiver allowed states the flexibility to reimburse for residential treatment for the first time. As previously described, Medicaid agencies relied on creative workarounds to deliver residential-treatment-like reimbursing for outpatient services and using block grant dollars to cover room and board. A CMS official described the SUD waiver as “an attractive option for both CMS and states to say, ‘we don't know the extent of Medicaid federal funds going out on the IMD exclusion, but here is an opportunity to come out into the light and do it the right way.’ And here you have no risks of compliance issues; we can do it safely.”

The SUD waiver became an attractive option for states because it allowed them to provide a full, high-quality SUD care continuum. Although pursuing the SUD waiver invited federal oversight of their benefits, stakeholders believed strongly that beneficiaries needed a high-quality SUD care continuum inclusive of residential treatment.

How Were Evidence-Based Practice Medicaid Benefits Selected?

The 2015 SUD waiver guidance from CMS stated that participating Medicaid agencies must use “industry standard models” to structure SUD care continuums (Wachino 2015). CMS gave states some flexibility to define most evidence-based practice (EBP) SUD benefits, but they required states to use the American Society of Addiction Medicine (ASAM) Criteria to define residential services and standardize patient assessments (Wachino 2015).

The ASAM Criteria are proprietary guidelines that define clinical service requirements for SUD care, patient assessments, care transitions, and discharge (Mee-Lee 2013). A CMS official said states needed to use ASAM Criteria to structure residential services and the patient assessment process to help ensure “that the quality of care that those beneficiaries are receiving is current, or at least consistent, with the industry standard.” According to a CMS official, CMS leadership also “wanted to make sure that the beneficiaries are able to basically enter into levels of care that are indicated by a comprehensive and evidence-based assessment and they're not necessarily just getting residential treatment because, you know, for other non-evidence-based reasons.” Leaders at CMS worried that clinicians might provide residential treatment to patients to receive a higher reimbursement rate than outpatient services, or offer residential treatment simply because it was available, regardless of whether that level of care was medically necessary (Wachino 2015).

Early-adopter states ultimately used the ASAM Criteria to structure their entire SUD service model. States perceived CMS's requirement to use ASAM Criteria for residential services and patient assessment criteria as an instruction that applied to the entire care continuum. One Virginia Medicaid representative said, “The way it was written, it was very clear that CMS wanted us to use ASAM.” Medicaid staff tasked with designing the new benefit said they were unable to find any alternative EBP model. One West Virginia Medicaid representative noted, “You have to have a national, standardized criteria for this. It has to be evidence-based. And ASAM is the only one.” When asked whether Medicaid leadership considered an alternative model, a key informant from California replied, “Well, what else have you got?”

Using the ASAM Criteria also helped generate early stakeholder buy-in for benefit reform. ASAM Criteria represented a standardized model or “the shell that, essentially, we needed to create the benefit.” Medicaid agency staff were excited to use ASAM Criteria as a way to raise the overall quality of care in the SUD benefit. A California Medicaid consultant highlighted the comprehensiveness of ASAM Criteria: “I think if you're looking to design a system today with a clean slate, you know, ASAM is probably a good way to do it.” Interviewees did not specifically explain why a structure, such as the ASAM Criteria, was needed for SUD services when one is not used for other chronic diseases.

The ASAM Criteria do not speak in detail to the role for MOUD or recovery support services (i.e., nonclinical services that promote recovery, such as peer coaching). To define these services, Medicaid agency benefit developers relied on interorganizational networks of SUD experts, colleagues from state behavioral health agencies, and other state Medicaid programs. Medicaid staff established small teams that were responsible for leading waiver development. Small teams worked “behind the scenes” to initially select EBPs before further vetting the potential benefits to other stakeholders.

California and Virginia Medicaid agencies formed stakeholder workgroups to assess early buy-in for identified EBPs. Workgroup members included Medicaid MCO representatives, providers, behavioral health experts, and advocates. Workgroup members said they made recommendations to clarify EBPs described in the state terms and conditions. “[We were] recommending edits to them: ‘You need to expand this language. You need to also make sure that we're highlighting certain services . . . so that they're on the radar.’” In these states, workgroups did not have authority over final benefit decisions, which were made by Medicaid staff. One Virginia Medicaid MCO representative described the process this way: “We certainly didn't have veto authority, but we certainly could say . . . ‘There's no way we can do this.’ . . . Because you know, our systems are not perfect systems. At the end of the day, you have to be able to authorize and pay for services. And if it gets too complicated, and we can't pay for services, or pay it correctly, it gets ugly pretty quickly.”

Medicaid staff also described having an informal network of state Medicaid agency partners that they regularly consulted. A Virginia Medicaid representative described the process as looking at “what services other states were using, anything that we could find that would be, like, an evidence-based spectrum of services.” Virginia Medicaid staff described using a copy of California's SUD waiver application to help identify potential EBPs. Medicaid staff also looked to other states without SUD waivers. “Just because they didn't have a waiver didn't mean they didn't have a really good spectrum of services,” one West Virginia Medicaid representative said. For example, West Virginia Medicaid staff reached out to their Georgia Medicaid peers to learn how they had defined and set reimbursement rates for peer support services. A West Virginia Medicaid representative said informal information exchanges between Medicaid agencies were common because “we're all in the trenches together. . . . We'll just pick up the phone and call each other.” These peer-to-peer calls among Medicaid agencies were used to vet EBPs and gather intelligence about how CMS might react to their proposed benefits.

How Did Medicaid Agencies Secure State Funds to Pay for New Services?

Medicaid agencies needed to secure state funding to pay for new services included in their revised benefit. Representatives from CMS noted that successful efforts to secure additional funds “depends on obviously the support that the state Medicaid agency would enjoy from their leadership,” noting that leaders in nonexpansion states had “less of an urgency” to enhance SUD benefits. This study did not investigate the magnitude of funds required to adopt new SUD services in each state, although interviews revealed that Medicaid staff invested time and effort to secure legislative support and enhance their Medicaid program budgets. States were only able to secure political and financial support for their planned SUD waivers after producing palatable evidence of the local need for SUD reform.

In California, the need for reform was intrinsically tied to the state legislature's desire to improve program integrity and oversight within Medicaid. In July 2013, CNN aired an investigative series revealing that the California Medicaid program had reimbursed $94 million in fraudulent billing to 56 SUD treatment facilities (Zamost and Griffin 2015). The media attention generated by the investigative report prompted the state legislature and the Medicaid agency to begin searching for ways to improve program oversight.

At the time of the investigation, providers contracted directly with the state Medicaid agency to deliver services. The SUD waiver provided the Medicaid agency with an opportunity to establish new services and require providers to contract directly with their county-run Medicaid programs. This contract change was intended to improve local oversight of provider credentialing, service provision and billing. California key informants said that without the fraud scandal, it was unlikely that the SUD benefit would have evolved. As one California provider noted, “We were able to finally get traction with the legislature, the governor and the Department of Healthcare Services to design a much more rich benefit and to put in place yet more controls, more regulations, to ensure some of the past abuse couldn't happen again. And so both had to go hand in hand or we never would have got anywhere with that effort.”

Both Virginia and West Virginia generated statewide legislative support for a waiver by focusing on the downstream effects of substance use rather than the immediate needs of adult Medicaid beneficiaries. One Virginia Medicaid staff member recalled that the general assembly “were like, ‘why should we care about the guys in the alley shooting up heroin?’ . . . You could still feel and see the stigma attached with all of that.” Medicaid agencies saw their request to enhance adult Medicaid benefits as potentially politically infeasible. In response, Medicaid staff in Virginia and West Virginia reframed Medicaid SUD reform as a way to reduce both the prevalence of neonatal opioid withdrawal syndrome and the number of children entering foster care as a result of parental substance use. “We decided to take a different approach . . . and said, ‘Well, what about the children?’ And they said, ‘What do you mean the kids?’ And so we said, ‘This is what's happening. Kids are being placed into foster care. And also, lots of babies are being born with neonatal abstinence syndrome.’ So we made it very personal.” Virginia Medicaid staff even took state legislators to a neonatal intensive care unit to further demonstrate how many babies were being treated for neonatal opioid withdrawal syndrome. Reframing SUD reform as a way to stop the negative downstream health outcomes of SUDs helped Virginia and West Virginia Medicaid staff convince legislature that substance use was a “systemic problem,” not just a Medicaid problem.

The aforementioned strategies helped early adopters secure the matching funds needed to implement SUD waivers. However, Medicaid staff remain concerned about their ability to sustain new services beyond the demonstration period. One West Virginia Medicaid agency representative said, “There's that, fear, right? So, like, yes, you have this for five years, but now what? Like, how are you going to sustain it?” Stakeholders said they will need to continue to present persuasive evidence to their legislature to sustain services over the long term.

Study results should be interpreted in light of our methodological limitations. The multiple case study design enabled us to examine three states' SUD waiver experiences in depth. Identifying common themes across three states highlights the transferability of our findings, but important differences likely exist between early-adopter and late-adopter states. This study focused on identifying lessons from states that were early adopters of the 1115 SUD waiver opportunity and pursued comprehensive reform of their SUD benefit arrays. States who decided to pursue reform later under CMS leadership appointed by the Trump administration may have experienced different waiver review processes. Contextual factors such as states' financial health (e.g., state budgets) and the nature of their existing substance use treatment system may differ from the states included in this study, thereby limiting the transferability of our findings. States that were undertaking a smaller-scale reform (e.g., only adding residential treatment services to their benefit array) may have experienced differential internal decision-making processes when pursuing reform. Snowball sampling techniques resulted in us conducting interviews with CMS and Medicaid agency staff, employees from Medicaid MCOs and Medicaid contracted providers. Including other stakeholder perspectives (e.g., patient advocates via interviews or additional document review of legislative testimonials) could have produced insights about additional influences on states' 1115 SUD waiver-planning and decision-making processes.

Conclusion

Little is known about state Medicaid agencies' internal decision-making processes in pursuing large-scale health systems reform. Comparing three states' experiences with exploring the need for enhanced benefits and preparing SUD waiver applications revealed four key decision points that other states could face when pursuing similarly complex and controversial reform. Early adopters remain engaged in their five-year demonstrations, and midpoint evaluations indicate that they have increased beneficiaries' access to SUD services (CMS 2019; Urada et al. 2019). Our study contextualizes the internal decisions that led to improvements in service access.

Previous research highlights the important influence of federal executive leadership on state Medicaid agency actions, such as Medicaid expansion (Thompson, Gusmano, and Shinohara 2018). However, we found that states' decisions were largely influenced by state stakeholder needs, not the federal executive branch. Early-adopting states did not immediately pursue the waiver once CMS issued its guidance document. Instead, states spent time assessing their individual SUD treatment needs. While states did eventually heed CMS's guidance to structure care continuums by using the ASAM Criteria, they also invited state SUD experts to help select SUD EBPs rather than exclusively relying on the menu of SUD services suggested in CMS's invitation for reform. Additionally, key informants in California, Virginia, and West Virginia reported that their respective state legislatures needed to approve budgetary changes related to enhancing Medicaid SUD benefits.

These results indicate that there are limits on when and how much executive leadership exerts influence over states' internal decision-making processes. The initial CMS guidance to reform SUD benefits was issued by the Obama administration in 2015 (Wachino 2015), then reissued by the Trump administration in 2017 with relaxed administrative requirements (Neale 2017); for example, states did not need to submit an evaluation plan with their initial waiver application. Presidential administration-driven alterations to waiver requirements also promote variation in states' desire to pursue reform and in their actual implementation of services, highlighting what some have called “variable speed federalism” (Conlan and Posner 2016). Future research should investigate how relaxed waiver requirements (if maintained by the Biden administration) or other waiver directives influence states' decisions to pursue innovation.

Additionally, Medicaid agencies' past experiences with CMS could impact their pursuit of future innovations. States had different preferences regarding incorporating new SUD benefits under waiver authority or an SPA. States concerned about opening SPAs to federal oversight could opt to submit waiver renewals rather than making new services a permanent part of their benefit. However, this approach would be onerous for state Medicaid staff, since waivers require lengthy justifications and public comment periods. Future research is needed to better understand states' decision-making processes for adding benefits through waiver demonstration and SPAs. Previous research has focused on waivers as tools to further federal executive branch policy goals (Gusmano and Thompson 2020), but it is less clear how SPAs could be vulnerable to executive influence.

Medicaid agencies do not envision benefit reform as a closed process within the state agency. Rather, Medicaid agencies work closely with interorganizational partners from other state agencies, Medicaid MCOs, providers, and other community organizations to define beneficiaries' unmet care needs and identify EBPs. Interorganizational networks act as bridging factors, or relational ties (Lengnick-Hall 2021), to support Medicaid innovation. Research on behavioral health integration has similarly demonstrated that Medicaid and behavioral health agencies are breaking down administrative siloes and collaborating (MACPAC 2016). Interagency collaborations were critical to states' abilities to prepare SUD waivers that were supported by state policy makers and CMS. Collaborations promote sharing resources, expertise, and the burden of tackling large-scale reform efforts. Collaborations can also help state agency staff navigate common waiver implementation challenges, including promoting provider awareness and fidelity to EBPs (Crable 2022).

Finally, early adopters of SUD waivers highlighted the challenge of receiving legislative approval to pay for Medicaid innovations. Medicaid agencies reframed SUD reform as an opportunity to reduce negative downstream effects of substance use on children and decrease the potential for fraud. This reframing is not surprising given the controversial political rhetoric surrounding Medicaid and widespread stigma toward individuals with SUDs (Olsen and Sharfstein 2014). Some conservative-dominated states framed their motivation for Medicaid expansion as a way to improve quality and control costs (Grogan, Singer, and Jones 2017). A recent survey revealed that 43% of Medicaid directors believe their agencies are under increasing scrutiny from the state legislature, and 51% reported a general increase in political pressure (NAMD 2019). Legislative scrutiny and political pressure highlight the need for Medicaid agencies to frame program changes in ways that are palatable to state leaders. This reframing will likely be unique to local state needs. For example, the opioid epidemic motivated benefit enhancement in Virginia and West Virginia, but California had not been substantially impacted by overdose deaths, and their legislature was instead focused on seeing reform as a way to reduce Medicaid fraud and waste. Not all states require Medicaid agencies to seek legislative approval for benefit changes or spending. In states without legislative approval requirements, reform efforts could rely more heavily on convincing Medicaid agency leadership and governors of the need for change, emphasizing the important influence of state-driven executive federalism (Thompson 2013).

Examining three states experiences with pursuing SUD waiver reform revealed critical insights about the types of decisions state Medicaid agencies must make and the diverse influences that guide their choices. Lessons from these early-adopter states can serve as a road map for other state Medicaid agencies and policy makers considering similar reform as well as advocates and researchers interested in better understanding policy makers' internal decision-making processes.

Acknowledgments

This research was funded through a grant from the Lifespan/Brown Criminal Justice Research Training Program on Substance Use, HIV, and Comorbidities (NIDA R25DA037190). The authors would like to thank the individuals who took time from their schedules to participate in study interviews. This qualitative research would not be possible without their generous contributions of both time and insights. The authors are also appreciative to Sara Bachman, Jonathan Purtle, Frank Thompson, and Heather Howard, who provided external feedback on different versions of this manuscript.

The authors would like to express their sincere gratitude to our late coauthor, Dr. David K. Jones, whose mentorship and support were instrumental to completing this study and manuscript.

References

Aarons, Gregory A., Hurlburt, Michael, and Horwitz, Sarah McCue.
2011
. “
Advancing a Conceptual Model of Evidence-Based Practice Implementation in Public Service Sectors
.”
Administration and Policy in Mental Health
38
, no.
1
:
4
23
.
Allen, Heidi, Wright, Bill J., Harding, Kristin, and Broffman, Lauren.
2014
. “
The Role of Stigma in Access to Health Care for the Poor
.”
Milbank Quarterly
92
, no.
2
:
289
318
.
Boozang, Patricia, Bachrach, Deborah, Detty, Andrew, and Solutions, Manatt Health.
2014
. “
Coverage and Delivery of Adult Substance Abuse Services in Medicaid Managed Care
.” Technical Assistance Brief, May. https://www.manatt.com/getattachment/abd7994e-7f5c-4aff-99db-6484139d110b/attachment.aspx.
Busch, Susan H., Meara, Ellen, Huskamp, Haiden A., and Barry, Colleen L.
2013
. “
Characteristics of Adults with Substance Use Disorders Expected to Be Eligible for Medicaid under the ACA
.”
Pscyhiatric Services
64
, no.
6
:
520
26
.
CHCF (California Health Care Foundation)
.
2018
. “
Medi-Cal Moves Addiction Treatment into the Mainstream: Early Lessons from the Drug Medi-Cal Organized Delivery System Pilots
.” Issue Brief, August. https://www.chcf.org/wp-content/uploads/2018/08/MediCalMovesAddictionTreatmentToMainstream.pdf.
CMS (Centers for Medicare and Medicaid Services). n.d.(a)
. “
Medicaid Enrollment Data Collected through MBES
.” www.medicaid.gov/medicaid/national-medicaid-chip-program-information/medicaid-chip-enrollment-data/medicaid-enrollment-data-collected-through-mbes/index.html (accessed
February
21
,
2022
).
CMS (Centers for Medicare and Medicaid Services). n.d.(b)
. “
Medicaid State Plan Amendments
.” www.medicaid.gov/medicaid/medicaid-state-plan-amendments/index.html (accessed
February
21
,
2022
).
CMS (Centers for Medicare and Medicaid Services)
.
1994
. “
Content Details: 59FR—Medicaid Program; Demonstration Proposals Pursuant to Section 1115(a) of the Social Security Act; Policy and Procedures
.”
Federal Register
59
, no.
186
. FR document number 94-23960, September. https://www.govinfo.gov/app/details/FR-1994-09-27/94-23960.
CMS (Centers for Medicare and Medicaid Services)
.
2016
. Letter to Cynthia B. Jones re the “
Virginia Governor's Action Plan (GAP) and Addiction Recovery Treatment Services (ARTS) Delivery System
.”
December
15
. https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/va/Governors-Access-Plan-GAP/va-gov-access-plan-gap-appvl-amdmnt-12152016.pdf.
CMS (Centers for Medicare and Medicaid Services)
.
2017
. Letter to Cynthia Beane re “
Creating a Continuum of Care for Medicaid Enrollees and Substance Use Disorders
.”
October
6
. https://dhhr.wv.gov/bms/CMS/Documents/WV%20Creating%20a%20Continuum%20of%20Care%20for%20Medicaid%20Enrollees%20with%20Substance%20Use%20Disorders%201115%20Demonstration%20%2810.06.2017%29.pdf.
CMS (Centers for Medicare and Medicaid Services)
.
2019
. “
Medicaid Section 1115 SUD Demonstration Monitoring Report—Part B: West Virginia, Creating a Continuum of Care for Medicaid Enrollees with Substance Use Disorders
.”
May
30
. https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/wv/continuum-care/wv-creating-continuum-care-medicaid-enrollees-qtrly-monitoring-rpt-dy2q1-2019.pdf.
Conlan, Timothy J., and Posner, Paul L.
2016
. “
American Federalism in an Era of Partisan Polarization: The Intergovernmental Paradox of Obama's ‘New Nationalism.’
Publius
46
, no.
3
:
281
307
.
Crable, Erika L., Benintendi, Allyn, Jones, David K., Walley, Alexander Y., Hicks, Jacqueline Milton, and Drainoni, Mari-Lynn.
2022
. “
Translating Medicaid Policy into Practice: Policy Implementation Strategies from Three US States' Experiences Enhancing Substance Use Disorder Treatment
.” Implementation Science,
January
6
. https://doi.org/10.1186/s13012-021-01182-4.
Donohue, Julie, Raslevich, Amy C., and Cole, Evan.
2020
. “
Medicaid's Role in Improving Substance Use Disorder Treatment
.”
September
25
. https://www.milbank.org/publications/medicaids-role-in-improving-substance-use-disorder-treatment-in-the-us/.
Douglas, Toby.
2014
. “
RE: California Bridge to Reform Demonstration (No. 11-W-00193/9) Amendment for Drug Medi-Cal Organized Delivery System (DMC-ODS) Waiver
.”
November
6
. https://cchealth.org/aod/pdf/RFQ-DMC-ODS-2015-05-Amendment-Request.pdf.
Gordon, Sarah H., Gadbois, Emily A., Shield, Renee R., Vivier, Patrick M., Ndumele, Chima D., and Trivedi, Amal N.
2018
. “
Qualitative Perspectives of Primary Care Providers Who Treat Medicaid Managed Care Patients
.”
BMC Health Services Research
18
, no.
1
: article no. 728. https://doi.org/10.1186/s12913-018-3516-9.
Green, Judith, and Thorogood, Nicki.
2014
.
Qualitative Methods for Health Research
. 3rd ed.
London
:
Sage
.
Grogan, Colleen M., Andrews, Christina, Abraham, Amanda, Humphreys, Keith, Pollack, Harold A., Smith, Bikki Tran, and Friedmann, Peter D.
2016
. “
Survey Highlights Differences in Medicaid Coverage for Substance Use Treatment and Opioid Use Disorder Medications
.”
Health Affairs
35
, no.
12
:
2289
96
.
Grogan, Colleen M., Singer, Phillip M., and Jones, David K.
2017
. “
Rhetoric and Reform in Waiver States
.”
Journal of Health Politics, Policy and Law
42
, no.
2
:
247
84
.
Guest, Greg, Bunce, Arwen, and Johnson, Laura.
1995. “How Many Interviews Are Enough? An Experiment with Data Saturation and Variability and a Good Number of Journals.”
Field Methods
18
, no.
1
:
59
82
. https://doi.org/10.1177/1525822X05279903.
Gusmano, Michael K., and J, Frank.
Thompson. 2020. “The Administrative Presidency, Waivers, and the Affordable Care Act.”
Journal of Health Politics, Policy and Law
45
, no.
4
:
633
46
. https://doi.org/10.1215/03616878-8255553.
Hines, Anika L., Barrett, Marguerite L., Jiang, Joanna, and Steiner, Claudia A.
2014
. “
Conditions with the Largest Number of Adult Hospital Readmissions by Payer, 2011
.” Agency for Healthcare Research and Quality, Statistical Brief #172, April. https://www.hcup-us.ahrq.gov/reports/statbriefs/sb172-Conditions-Readmissions-Payer.jsp.
KFF (Kaiser Family Foundation)
.
2020
. “
Status of State Medicaid Expansion Decisions: Interactive Map
.” www.kff.org/medicaid/issue-brief/status-of-state-medicaid-expansion-decisions-interactive-map/ (accessed
February
21
,
2022
).
Lengnick-Hall, Rebecca, Stadnick, Nicole A., Dickson, Kelsey S., Moullin, Joanna C., and Aarons, Gregory A.
2021
. “
Forms and Functions of Bridging Factors: Specifying the Dynamic Links Bbetween Outer and Inner Contexts during Implementation and Sustainment
.” Implementation Science,
April
1
. https://doi.org/10.1186/s13012-021-01099-y.
MACPAC (Medicaid and CHIP Payment and Access Commission)
.
2016
. “
Integration of Behavioral and Physical Health Services in Medicaid
.” March. https://www.macpac.gov/wp-content/uploads/2016/03/Integration-of-Behavioral-and-Physical-Health-Services-in-Medicaid.pdf.
McMullen, Erin.
2016
. “
Maryland Medicaid and Opioid Epidemic
.” Maryland Department of Health and Mental Hygiene,
June
23
. https://mmcp.health.maryland.gov/Documents/MMAC%20presentation%20DUR%20and%20Overdose%20Jun%2016.pdf.
Mee-Lee, David, ed.
2013
.
The ASAM Criteria: Treatment Criteria for Addictive, Substance-Related, and Co-Occurring Conditions
.
Chevy Chase, MD
:
American Society of Addiction Medicine
.
NAMD (National Association of Medicaid Directors)
.
2019
. “
Seventh Annual State Medicaid Operations Survey
.” November. https://medicaiddirectors.org/publications/seventh-annual-medicaid-operations-survey/.
Neale, Brian.
2017
. “
SMD #17–003 RE: Strategies to Address the Opioid Epidemic
.” Centers for Medicare and Medicaid Services,
November
1
. https://www.medicaid.gov/federal-policy-guidance/downloads/smd17003.pdf.
Olsen, Yngvild, and Sharfstein, Joshua M..
2014
. “
Confronting the Stigma of Opioid Use Disorder—And Its Treatment
.”
JAMA
311
, no.
14
:
1393
94
.
Rosenbaum, Sara.
2014
. “
Medicaid Payments and Access to Care
.”
New England Journal of Medicine
371
, no.
25
:
2345
47
.
SAMHSA (Substance Abuse and Mental Health Services Administration)
.
2014
. “
Projections of National Expenditures for Treatment of Mental Health and Substance Use Disorders, 2010–2020
.” https://store.samhsa.gov/sites/default/files/d7/priv/sma14-4883.pdf (accessed
February
21
,
2022
).
SAMHSA (Substance Abuse and Mental Health Services Administration)
.
2019
. “
Behavioral Health Barometer, Volume 5
.” https://store.samhsa.gov/sites/default/files/d7/priv/sma19-baro-17-us.pdf (accessed
February
21
,
2022
).
Saunders, Benjamin, Sim, Julius, Kingstone, Tom, Baker, Shula, Waterfield, Jackie, Bartlam, Bernadette, Burroughs, Heather, and Jinks, Clare.
2018. “Saturation in Qualitative Research: Exploring Its Conceptualization and Operationalization.”
Quality and Quantity
52
, no.
4
:
1893
907
. https://doi.org/10.1007/s11135-017-0574-8.
Thompson, Frank J.
2013
. “
The Rise of Executive Federalism
.”
American Review of Public Administration
43
, no.
1
:
3
25
.
Thompson, Frank J., and Burke, Courtney.
2007
. “
Executive Federalism and Medicaid Demonstration Waivers: Implications for Policy and Democratic Process
.”
Journal of Health Politics, Policy and Law
32
, no.
6
:
971
1004
.
Thompson, Frank J., Gusmano, Michael K., and Shinohara, Shugo.
2018
. “
Trump and the Affordable Care Act: Congressional Repeal Efforts, Executive Federalism, and Program Durability
.”
Publius
48
, no.
3
:
396
424
.
Urada, Darren, Antonini, Valerie P., Teruya, Cheryl, Joshi, Vandana, Lee, Anne B., Iturrios-Fourzan, Isabel, Tran, Elise, Huang, David, and Vazquez, M. Eva.
2019
. “
Drug Medi-Cal Organized Delivery System: 2019 Evaluation Report
.” UCLA Integrated Substance Abuse Programs,
September
16
. http://uclaisap.org/dmc-ods-eval/assets/documents/DMC-ODS-Year-4-Evaluation-Report-FY-2018-19.pdf.
VDMAS (Virginia Department of Medical Assistance Services)
.
2016
. “
Virginia GAP Program for Seriously Mentally Ill 1115 Demonstration Application for Amendment
.” Virginia's Addiction and Recovery Treatment Services Delivery System Transformation, August. https://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Waivers/1115/downloads/va/Governors-Access-Plan-GAP/va-gov-access-plan-gap-amend-req-sud-demo-08162016.pdf.
Wachino, Vikki.
2015
. “
SMD #15–003 Re: New Service Delivery Opportunities for Individuals with a Substance Use Disorder
.” Centers for Medicare and Medicaid Services,
July
27
. https://www.medicaid.gov/federal-policy-guidance/downloads/smd15003.pdf.
Weiss, Audrey J., and Jiang, H. Joanna.
2021
. “
Overview of Clinical Conditions with Frequent and Costly Hospital Readmissions by Payer, 2018
.” Agency for Healthcare Research and Quality, Statistical Brief #278, July. https://www.hcup-us.ahrq.gov/reports/statbriefs/sb278-Conditions-Frequent-Readmissions-By-Payer-2018.jsp.
Wright, Brad, Porter, Anna, Singer, Phillip M., and Jones, David K.
2018
. “
The Devolution of Health Reform? A Comparative Analysis of State Innovation Waiver Activity
.”
Journal of Health Politics, Policy and Law
44
, no.
2
:
315
31
.
WVVIPC (West Virginia Violence and Injury Prevention Center)
.
2017
. “
2016 West Virginia Overdose Fatality Analysis: Healthcare Systems Utilization, Risk Factors, and Opportunities for Intervention
.”
December
20
. https://dhhr.wv.gov/bph/Documents/ODCP%20Reports%202017/2016%20West%20Virginia%20Overdose%20Fatality%20Analysis_004302018.pdf.
Zamost, Scott, and Griffin, Drew.
2015
. “
CNN Investigation into Massive Fraud Leads to Charges, Convictions
.”
November
12
. https://www.cnn.com/2015/11/12/us/rehab-racket-charges/index.html.