Ensuring access to high-quality substance use disorder treatment for Medicaid enrollees: A qualitative study of diverse stakeholders’ perspectives

Abstract

Background:

Medicaid programs are vital to ensure low-income individuals have access to substance use disorder (SUD) treatment. However, shifts in Medicaid policies may alter coverage and SUD care for this population, who already face difficulties receiving high-quality SUD treatment. Using a policy implementation research approach, we sought to identify barriers and facilitators when transitioning from Medicaid fee-for-service to managed care plan structures and opportunities for improving SUD care in New York State (NYS).

Method:

Study staff conducted semistructured, in-depth qualitative interviews (N=40 total) with diverse stakeholders involved with different aspects of SUD treatment in NYS, including policy leaders (n = 13), clinicians (n = 12), Medicaid managed care plan administrators (n = 5), and patients (n = 10).

Results:

Findings from thematic analysis centered on three themes: 1) while transitions to managed care have benefited clinicians, certain policies affect patients’ Medicaid enrollment and quality of care; 2) stakeholders perceived individuals with dual diagnoses, older adults, and linguistic minorities to be at higher risk for inadequate care; and 3) current quality metrics may not adequately capture treatment quality.

Conclusion:

Policy changes should focus on promoting increased collaboration among stakeholders, expanding Medicaid coverage, and reducing stigma. Resources should be diverted to facilitate psychiatric care for patients with dual diagnoses and to build workforce capacity to adequately meet the needs of older adults and linguistic minorities. Opportunities for NYS Medicaid include adapting performance metrics to capture meaningful patient outcomes and link reimbursements to improvements in patients’ quality of life.

1. Introduction

In 2017, 19.7 million individuals in the United States battled a substance use disorder (SUD), amounting to 7.2% of the entire country’s population aged 12 and older (Substance Abuse and Mental Health Services Administration, 2018). Low-income communities evidence a disproportionately high prevalence of SUDs (Andrabi et al., 2017; Collins, 2016; Karriker-Jaffe, 2013). Medicaid and Medicaid expansions sponsored by the Affordable Care Act are critical in ensuring access to SUD treatment for low-income individuals (Humphreys & Frank, 2014). However, the shifting policies, landscape, and structure of Medicaid may affect care for low-income populations, who already face greater barriers to quality and adequate SUD treatment (Olfson et al., 2018). High quality SUD care includes timely assessment of disease severity, wraparound services for substance use, physical health, and provision of long-term supports (Fields & Roman, 2010). Research can point to how best to improve access and deliver high quality SUD treatment.

New York State (NYS) has the second highest prevalence of opioid use or dependency in the United States (National Survey on Drug Use and Health, 2017). The number of overdose deaths among NYS residents tripled between 2010 and 2017, increasing from 5.4 to 16.1 deaths per 100,000 people (New York State Department of Health, 2019). Beginning in 1997, NYS began to transition most of its Medicaid beneficiaries from a traditional fee-for-service system to Medicaid managed care (Centers for Medicare & Medicaid Services, 2019a), to centralize services and lower overall health care expenditures (Berwick & Hackbarth, 2012; Centers for Medicare & Medicaid Services, 2019b). Despite this change, coverage for behavioral health services remained limited for more than a decade (Centers for Medicare & Medicaid Services, 2019a). In 2011, NYS launched a Medicaid redesign project that carved out behavioral health services previously excluded from managed care plans (New York State Department of Health, 2011). While this change allowed for more insurance coverage among low-income individuals, the rates of SUD treatment remained the same (Olfson et al., 2018). Implementation of parity laws that allow equal coverage of behavioral and physical health have been posited as primarily Medicaid’s responsibility (Geissler & Evans, 2020). However, Medicaid managed care organizations (MCOs) continue to face challenges in the provision of behavioral health coverage and care integration (Tuck & Smith, 2019). Most notably, MCOs report CFR 42 regulations as a primary barrier, since limitations on information sharing hinder interinstitutional care coordination for SUD treatment (Tuck & Smith, 2019). Other documented barriers to accessing quality SUD care for Medicaid enrollees include treatment center norms and providers’ biased attitudes (Foney & Shannon Mace, 2019). Thus, managed care plans, policy-makers, and clinicians hold critical roles for increasing access to quality SUD care for Medicaid enrollees.

Existing administrative data are useful in assessing the performance of Medicaid MCOs, but they tend to be incomplete (Pawlson et al., 2007). While administrative data may identify problems, they cannot elaborate on policies and practices that cause these problems or their potential solutions (McGinnis et al., 2014). The goal of policy implementation research is to understand factors that influence the translation of policies (e.g., rules, laws, and regulations) into practice, as many of the greatest health advances have been attributed to successful policy change and implementation (Nilsen et al., 2013). Using a policy implementation research approach (Nilsen et al., 2013), we sought to engage different stakeholders to identify the effects on SUD care when transitioning from Medicaid fee-for-service to managed care, and the outstanding barriers and facilitators to receiving or providing SUD quality care through managed plan structures. We used a qualitative approach to gather the perspectives of different stakeholders to understand barriers and facilitators that influence policy implementation (Head & Alford, 2015).

2. Method

2.1. Participants and recruitment

The Disparities Research Unit at Massachusetts General Hospital in Boston coordinated this research. Purposeful sampling was used to recruit diverse types of “information rich” (Patton, 2002) stakeholders: policy leaders, including state-level employees, leaders within advocacy organizations, or academics focused on SUD research and policy; clinicians or clinic administrators who provided SUD treatment services (herein referred as clinicians); plan administrators working within SUD care in Medicaid MCOs; and patients receiving SUD treatment through Medicaid. Recruitment started with seven policy leaders who had agreed to participate in connection with this research. The study recruited additional participants through snowball sampling techniques, where policy leaders referred us to plan administrators and/or clinicians. Concurrently, we used NYS listings of Medicaid substance use plans and providers to contact additional participants. For patient recruitment, the research team contacted state-listed SUD treatment and harm-reduction centers in New York City, asking them for assistance in recruiting patients. Participating clinics then recruited and specified dates and times for in-person interviews. Before recruitment, we estimated needing an overall sample of 40–50 total interviews to reach valid conclusions across stakeholder groups (Guest et al., 2006).

2.2. Procedure

We conducted semistructured, in-depth interviews, a qualitative data collection technique that uses an interview guide while allowing flexibility for interviewers to ask additional questions and elicit new topics (DeJonckheere & Vaughn, 2019). The research team developed interview guides using an iterative process and tailored them to ask each diverse stakeholder group about the same domains. Final interview guides focused on (1) general barriers and facilitators to receiving or providing SUD quality care, (2) differences in SUD services received or offered from Medicaid fee-for-service in contrast to managed care, (3) and evaluation of SUD quality of care. PhD researchers trained in conducting qualitative interviews conducted all interviews. Patient participants were interviewed in person, while the other stakeholders were interviewed over the phone. Interviews lasted 30 to 60 minutes, averaging around 45 minutes. Study staff obtained verbal consent, or written consent for patients, to participate in the study and be audio-recorded. Study staff conducted all patient interviews in private rooms at the clinics. All participants received a $100 gift card for their participation, although most policy leaders declined compensation because of payment restrictions associated with their agencies. The Partners Healthcare Institutional Review Board approved the study.

2.3. Analysis

Research staff transcribed the interviews verbatim, and de-identified and imported them into Dedoose, an online application for qualitative data management and analysis (Dedoose, 2020). This study used a thematic analysis framework, consisting of the following stages: familiarization with the data, creating codes, coding transcripts, generating and reviewing themes, and describing and finalizing themes (Braun & Clarke, 2006). As we collected data , the research team met regularly to review and discuss ongoing interviews and become familiar with the data. Next, the team developed an intial code book, and two trained coders (PK & SM), coded transcripts independently, identified coding differences, and met with the larger team to discuss discrepancies and suggest codebook modifications. The team iteratively refined the codebook through this process until the team developed a final codebook. The two coders coded all interviews independently, reviewed discrepancies between themselves until they reached consensus (Corbin & Strauss, 2014), and the first author (JZD) confirmed their codes. We grouped patterns of responses generated from codes into potential themes across stakeholder groups (Braun & Clarke, 2006). Refinement and clarification of themes occurred collaboratively among research team members until we solidified these themes and excerpts.

3. Results

3.1. Sample and study overview

Of the 88 stakeholders invited, 40 agreed to participate, including policy leaders (n=13), clinicians (n=12), plan administrators (n=5), and patients (n=10; see Table 1 for demographic information). Refusals came mostly from plan administrators and clinicians who did not have interest or clearance to participate.

Table 1.

Participant information for individual interviews (N=40).

	#of Participants	Male	Female	Mean Age (SD)	Race	Ethnicity
Patients	10	7	3	40.8 (range 27–65)	3 White; 1 Black; 2 Asian	4 Latino
Plan administrators	5	5	0	62.0 (range 41–74)	5 White
Clinicians*	12	4	6	54.7 (range 46–58)	6 White; 4 Black; 1 Asian	1 Latino
Policy leaders**	13	4	6	52.7 (range 34–61)	7 White; 1 Asian	2 Latino

Note:

^*Two clinicians missing gender;

^**Three policy leaders missing demographic data

Policy leaders, plan administrators, and clinicians varied in locality (rural, urban). Most clinicians worked in outpatient settings, and 10 reported that their clinics offered medication for addiction treatment. Of clinicians, six worked in community outpatient/inpatient treatment centers, four participants were linked with university hospitals/hospital clinics, one worked in private practice, and one worked within a forensic setting. Four of these clinicians worked for clinics providing methadone treatments, three mentioned provision of naltrexone, Vivitrol, and buprenorphine within their care settings, and one described care based on an abstinence model. About half of patients were mandated to receive treatment.

Based on the qualitative interviews, the research team identified three major themes: 1) while transitions to managed care have benefitted clinicians, certain policies affect patients’ Medicaid enrollment and quality of care; 2) individuals with dual diagnoses, older adults, and linguistic minorities were perceived to be at higher risk for poor quality care; and 3) current Medicaid quality metrics may not adequately capture treatment quality. Themes presented here represent common experiences across multiple stakeholder interviews and are accompanied by illustrative quotations. We present additional supporting quotes in Table 2.

Table 2.

Additional illustrative quotes.

Theme 1: While transitions to managed care have benefitted clinicians, certain policies affect patients’ Medicaid enrollment and quality of care
Sub-theme	Representative quote(s)
Managed care provides accountability and support to providers	“I think there’s better care coordination, at least potentially, with the managed care patients versus fee for service, which I think is going to be key if you’re looking at what needs to be done regarding substance use disorder. I think a lot is haphazard under fee for service and I think there’s a lot that can be gained with better care coordination.”
	“So [with managed care], every client has an individualized treatment plan and we try very hard to make sure that the needs of the individual is met so that we can provide good quality care.”
	“…with managed care, some of the restrictions about seeing more than one provider in a day has been lifted. So folks can more conveniently access things at times, or have a psychiatry appointment and another provider appointment, and I think that has, where that can happen, that is helpful, and in the traditional Medicaid system that was always challenging.”
Certain Medicaid policies exacerbate problems with healthcare navigation and enrollment	“I was already in treatment. So, they [Medicaid] were asking me for proof of address, if I was taking money and proof of the program. So, I came and told [name of clinician] ‘Hey, they are asking me for proof of the program,’ so she made me the letter. Then I brought the letter to them [Medicaid] and finally I have it [insurance]. [I lost insurance for] about two weeks”
There are unintended consequences of some policies crafted to improve SUD care	“one of the issues that I have with how the state has set that up is in order for practitioners to bill for SBIRT (Screening, Brief Intervention and Referral to Treatment), they have to take a 4-hour continuing education course about that, which, it’s 4 hours to do that when in order to prescribe opioids, we have to have a 3-hour course. There’s something wrong with this”
	“I don’t care if someone stays 10 days, 20 days, 30 days, or 40 days, if they need to be there. And if you’re working a transition that makes sense. If you are focusing, which is what the state is doing with the regulation, on time, you’re not focusing on quality of care. You can’t do both.”
	“[When selecting SUD care options for enrollees] You’re reduced to these programs that are insurance-driven models that are going to be like ‘We’ll get you in for two and half minutes and then we’ll throw you out and we’ll put you into an IOP [intensive outpatient program] and we’ll see you for six times, and then we’re gonna move you down a level and move you down a level, and then we’re going to declare you well in 6 to 12 months, then you have no contact and that’s it. Goodbye! If you’re left!”
Theme 2: Individuals with dual diagnoses, older adults, and linguistic minorities were perceived to be at higher risk for poor quality care
Sub-theme	Representative quote(s)
Dual diagnoses	“for instance, somebody getting buprenorphine maintenance in primary care who has a primary psychiatric disorder- we still struggle to tie them into our behavioral health services in addition to the primary care-… they will often be evaluated by general psychiatry and told they need specialty addiction treatment, which they’re already kind of getting, but not in the sense of a traditional outpatient program, and so they may essentially be refused entry into a general mental health clinic for say, management of their bipolar disorder, because the mental health clinic has the kind of triage algorithm that puts people with active substance use disorders over in intensive outpatient treatment where they can get psychiatric services and addiction treatment, including buprenorphine, but [where] that person’s already on buprenorphine in their primary care, they’re not gonna stop doing that, they’re not looking for an intensive outpatient program, and the result is, 3 to 6 months later, the person still doesn’t have a psychiatrist.”
Older adults	“say for instance, a[n older] patient who comes in a day early, or two days earlier, they don’t have any medication, and then it’s in the computer that they were issued medication for those two days, what happened is one of the policy in the clinic is that if they don’t return that bottle then they’re supposed to not get any more take home until their case has been discussed in a disciplinary meeting. So then, we, the medical directors, sometimes they have to decide that they have to give that, depending on the medical situation with that person, that they will have to document that this policy cannot be applied to this person. You know, which can be a little tricky, because it’s there as a guideline, so these are things that I think OASAS needs to look at, revise in their policy, to ensure that we are not being too hard on the population that’s aging and having these difficulties.”
	“.…that aging population is absolutely being affected. We don’t discharge them. I take a financial loss, I’ll tell you that right now (laughs) and we just do a sliding fee scale for those who can afford it, for those who can’t, we’re just carrying them. It’s unfair for someone who’s had treatment for 30 years to all of a sudden tell them you can’t receive services here [when aged out of Medicaid]. So those are the main barriers, is the patient needing education and then our aging population not getting the services they need as a result of not having Medicaid.”
Linguistic minorities	“…it’s not always easy to find the facility you need. I’ve got a fairly large population of Orthodox Jews. They only speak Yiddish. I have Yiddish [speaking] counselors but I don’t always have enough and I can’t always get them in.”
Theme 3: Current Medicaid quality metrics may not adequately capture treatment quality
Sub-theme	Representative quote(s)
Utility of quality metrics and assessments	“I think that Medicaid data itself, the records that you have now, I think would not allow any kind of analysis about patient experience. There would just be a lot of assumptions of – it looked like this patient or this person was continuing with care, and then all of a sudden they stopped. We don’t know what that means, right? Does it mean they were disenrolled, or does it mean they moved, you know, it’s really hard to know what that means..”
Quality measures lack patients’ voices	“DOH [Department of Health] knows what they do, and OMH [Office of Mental Health] knows what they do, but no-one is looking at the patient, they’re looking at the system. And we don’t know how a patient who has a disease that is affected by all these elements can navigate all these systems.”
Quality of SUD care should include measures for individuals’ daily and social functioning	“as a social worker, that’s just something that is innately within me as far as understanding how culture, how a socioeconomic background, how education, how environment, how all these things affect an individual and how we can’t put, it’s not a one, as they say, cookie-cut treatment. You know, like one way does not work for everyone.”
	“I think if we could measure and capture the length of a person’s recovery further out, you know, how they are doing a year from now, and if we could even measure their quality of recovery, and I think that gets back to some of their other health needs, you know, work, the family connections, their physical health, you know, they made their initial primary care appointment very shortly after we discharged them and a health issue was identified and then they maintained their connection with that primary care physician to address that health issue, which then improves their quality of life, their quality of recovery. Those would be very very helpful measurements, that I think that would then, and family connections-did we do a good job with the family treatment program? Do we need to improve on that based upon the connection that people make with their family, work, do we need to do more with vocational training?”

3.2. Theme 1: While transitions to managed care have benefitted clinicians, certain policies affect patients’ Medicaid enrollment and quality of care

Policy changes in the past decades have allowed for easier access to SUD treatment for Medicaid enrollees, but certain policies continue to hinder quality of SUD care. Three subthemes emerged: (1) Managed care provides accountability and support to providers, (2) certain Medicaid policies exarcerbate problems with health care navigation and enrollment, and (3) unintended consequences occur with some policies crafted to improve SUD care.

3.2.1. Managed care provides accountability and support to providers

Participants believed certain aspects of managed care were critical to improving SUD access for Medicaid beneficiaries. One participant stated “…with managed care, there’s more oversight, accountability, and ability from the managed care plans to contact a provider or nudge the provider on things.”(P14, Plan administrator). Stakeholders described Medicaid as an exceptional resource for people seeking SUD treatment and that managed cared structures have allowed for more opportunities to provide accountability and support, and thus improved quality of care provided: “Medicaid seems to be the gold card here.” (P29, Clinician). This experience was particularly salient among patients, who valued access to Medicaid:

Interviewer: “What do you think that we as providers, as society, should do to make sure that people get the services they need?”

P34, Patient: “Give them Medicaid.” (Laughs)

Clinicians and plan administrators largely favored the transition from fee-for-service to managed care, particularly because MCOs provided oversight and support to providers, and included MCOs’ clinical audits: “Managed care audits are different from [state] audits because they focus more on clinic-issues and it’s created a whole other level of accountability and work to be done” (P21, Clinician).

One clinician described the utility and breadth of clinical audits, explaining that MCOs: call our main headquarters and say ‘Hey, you billed us for [name] on [date]’ and that note has to be produced, what they look for is the quality care within the note, what specific therapy was used. Did you use MI [motivational interviewing], did you use CBT [cognitive behavioral therapy], reality therapy, what was used, is it documented appropriately, is it a billable note as per the APG [Ambulatory Patient Groups] guidelines. So, it’s very, very specific.”

(P30, Clinician).

Clinicians and plan administrators believed that these audits were important to establish accountability and ensure that clinicians provided quality care. Increased accountability also meant that clinicians were responsible for providing quality of care and producing proper documentation to be reimbursed for services:

If you’re not meeting the quality of care that’s supposed to be provided, or your documentation is not correlating with the services that were billed, you lose money on that. They’re [MCOs] very strict on that.”

(P30, Clinician)

3.2.2. Certain Medicaid policies exacerbate problems with healthcare navigation and enrollment

All patients in this study acknowledged that they would not have been able to receive care without Medicaid due to costs, particularly in inpatient settings. However, some patients stated they liked the “old Medicaid” better because they could go to any clinic that accepted Medicaid without restrictions. One of the biggest barriers for patients following the transition to MCOs was not understanding the constraints of managed care:

Patients don’t really understand managed care. They don’t understand why they can’t go to Dr. V anymore and they have to go to Dr. S, because Dr. V is not in that HMO [health maintenance organization] … The patient then changes to [MCO] but doesn’t understand that there’s a waiting period to go to [MCO], so they have to wait to see the doctor. So, in the meantime they’ll go to the emergency room.

(P23, Clinician)

All stakeholder groups acknowledged the negative impact of MCOs on patients’ preferences for care, including SUD care.

Stakeholders described that certain related policies can hinder enrollment and therefore affect SUD care. For example, patients often relied on public assistance, which can be revoked if patients do not follow through on mandated treatment. One patient described some of the practical barriers outside of Medicaid that can affect SUD care for Medicaid beneficiaries:

If you don’t go to the appointments… the person that’s giving you the Medicaid will… shut you[r public assistance] down real quick. You don’t know why they didn’t go to the appointment. You don’t [know if they] have received it in the mail. Not only that, but they tell you that [you have to go] face-to-face with the HRA [Human Resources Administration]. How you gonna go to HRA if you don’t have Medicaid or Metrocard?

(P34, Patient)

Outpatient clinicians helped uninsured individuals to enroll in Medicaid, yet one policy leader who also performs clinical duties explained that “Medicaid has a 45-day waiting period” before benefits are in place so providers may offer treatment but are not guaranteed reimbursement:

While they may be in our treatment program, we’re not getting reimbursed, and if they leave treatment early, we don’t get reimbursed at all, and they’re not connected. Because they left treatment so they lose their benefits. So that’s a bureaucratic function—that’s really what it is, it’s a processing and a bureaucratic function.”

(P25, Policy leader)

Participants also reported that Medicaid enrollees experienced lapses in coverage due to income fluctuations that impacted their eligibility and SUD care:

People fall in and out of enrollment… I think [it] speaks to a very vulnerable population whose care is going to be compromised because, if their income changes modestly, then they might not be eligible for several months. Lapses in being actually enrolled in Medicaid might in and of itself be a marker of poor quality of care.”

(P6, Policy leader)

Stakeholders across groups described how policies can leave people without insurance coverage or treatment, hindering quality of SUD care. Describing patients who lose SUD coverage in the transition from Medicaid to Medicare, one clinician states: “they’re afraid to go to emergency rooms because they know they’re going to get a bill. So what are they doing? They’re avoiding medical care at all costs” (P23, Clinician). However, most clinicians described working around coverage regulations by providing services on a sliding scale, reserving funding for the uninsured (“a little pot of gold,” P30, Clinician), or providing free services to ensure SUD treatment was continuously offered. While policy leaders explained that NYS reimburses treatment for uninsured individuals, front-line providers in this study did not seem aware of this coverage or could not access it.

3.2.3. Some policies crafted to improve SUD care have unintended consequences

One NYS regulation that was often mentioned among plan administrators and clinicians was utilization management, a process whereby plan administrators assess the appropriateness of the care provided (e.g., prior authorizations). NYS prohibits plans from performing utilization management or denying coverage for the first 14 days of inpatient treatment; this policy elicited mixed reactions among stakeholders. For example, one plan administrator believed that prohibiting plans from utilization management was detrimental as plan administrators could not help members to choose the best course of treatment:

Because we see people who get admitted to facilities over and over. We reach out to the facility and say ‘We want to help you manage the transition because this guy’s flunking transition, time after time after time.’ And the facilities say to you ‘Hey, the law says I don’t have to talk to you for 14 days.’ So what you get is a managed care expert who has a network and has the ability to really help with transitions of care, engaging with a provider who says ‘I don’t need to talk to you, I’ll do it myself’ but they don’t do it, and we’re not allowed to touch them… If you create wedges between managed care and the provider community, the person who gets ripped apart and falls in the hole that’s created is the member.

(P20, Plan administrator)

However, clinicians and policy leaders viewed the prohibition of utilization management as necessary to protect patients’ rights to acute treatment. Some clinicians perceived difficulties getting MCOs to reimburse inpatient stays, which created serious disruptions to clinical care:

Utilization management says “this guy keeps going to rehab or to inpatient, something’s wrong,” so we already determined that he needs a higher level of care, and [plans say] “he can only stay for 10 days,” and we’re like, “Noooooo! He needs to be at least through the 28th to get him out of the environment, away from the places, people, and things that are causing these problems and then he can start to heal.” So that’s a big thing because of not enough time being okayed for these managed cares.

(P29, Clinician)

The study did not specifically ask participants about utilization management and patients did not speak to the implementation and effectiveness of this policy. However, many talked about the impact that “treatment leverage” policies had on their lives. Treatment leverages included mandated treatment to avoid incarceration or regain custody of their children, the latter being particularly salient among female patients. Some patients, clinicians, and plan administrators perceived these leverages as punitive rather than focused on their healing process. Here some patients describe their experiences with mandated treatment and ways that it influenced their perception of quality of SUD care:

[Treatment centers would say] “You’re doing this, you’re doing that, you’re not complying” and then they’d go back to the court and say this and this and that, and then the ACS worker will say “Oh yes, she’s not complying and she doesn’t want her kids.” No, that’s not it. You’re putting me in a place that I’m not comfortable with, that I don’t feel like I’m getting what I need. But they don’t care about that, they just want you to go to the program and complete what you’ve got to do.

(P31, Patient)

I came on my own this time around, and it works. When you’re mandated, it seems like you’re doing it for something, you know, you want to stay clean because you’re going to get in trouble or go back to jail, lose your kids, lose your apartment, lose whatever the mandation was about. I’m doing this on my own now, and I feel good about it.

(P34, Patient)

Thus, while some patients found mandated treatment helpful, most believed that treatment leverages did not help them to stay engaged in treatment. Many programs have strict requirements for daily check-ins as a condition of receiving medication, with consequences for missing visits including inability to see their families. Patients described that these leverages excacerbated the stigma, shame, and punitive nature of SUD treatment, hindered their ability to stay engaged, and therefore reduced the perceived quality of SUD care.

3.3. Theme 2: Stakeholder perceived individuals with dual diagnoses, older adults, and linguistic minorities to be at higher risk for inadequate care

Participants described that the transition to managed care improved access to SUD treatment in general and that most patients could readily access most SUD treatments if they wanted. However, policies transitioning to managed care practices failed to address disparities in care for some groups, which included individuals with (1) dual diagnoses, (2) older adults, and (3) linguistic minorities.

3.3.1. Dual diagnoses

Clinicians and policy leaders often described that individuals with dual diagnoses of SUD and severe psychiatric illness were likely to receive inadequate care despite policy transitions aimed at improving access and quality of care. Some stakeholders stated that the disparity resulted from the lack of centers equipped to adequately address dual diagnoses, which had not changed much despite multiple policies enacted to improve SUD care:

We need [to do] more to address the patients with severe mental health … they come here [to the SUD outpatient clinic, and] we end up sending them back to the shelter that they came from … but after we’ve tried all the places that we know [that treat dual diagnoses] and people saying ‘no, no no no,’ we end up sending [them] back.

(P16, Clinician)

Policy leaders and clinicians explained that patients with dual diagnoses, in particular, have poorer treatment quality because “only 10%” of psychiatrists accept Medicaid in NYS” and because of the burden placed on patients to attain psychiatric care:

One thing that I think has been a long-standing difficulty for all types of patients, is accessing psychiatry in an efficient, low barrier way that doesn’t involve huge commitments.

(P22, Clinician)

Lack of psychiatrists and shortages of prescribers to provide medication for opioid use disorder treatment was also problematic for the larger SUD community in general:

While it is very easy to get into a traditional medication-assisted treatment program where you can be prescribed methadone, it is not that simple to find a doctor that can induce you on buprenorphine or Suboxone. [Places are] booked 6, 8, 10 weeks out in the future, and if someone wants to come off heroin and get on Suboxone they can’t wait 6, 8, or 10 weeks.

(P21, Clinician)

3.3.2. Older adults

Certain stakeholders also believed that older adults were more likely to receive inadequate SUD care, unlike those in other age groups. For example, some patients shared observations that older adults were missing “methadone appointments” by going at the wrong times due to cognitive decline and that some had chronic conditions (e.g., chronic pain) that were compounded by withdrawal symptoms. Patients explained that SUD treatment was rarely tailored to older adults and noted their older peers’ difficulties to remain in care. One policy leader explained: “I think older adults get short shrift in the treatment community, and we don’t have tailored programs that meet their needs and don’t necessarily look into or address the issue of physical co-morbidities in that population” (P11, Policy leader).

3.3.3. Linguistic minorities

Participants also described care challenges for linguistic minorities as a result of language barriers. For example, one clinician talked about patients they see more intensively in their outpatient clinic because of language barriers:

[these patients] need a high level of service, which means they are not really meeting the criteria of outpatient, but because of their language barriers there is no residential or inpatient provider available for people who do not speak English.

(P24, Clinician)

Despite making accommodations, resulting care could still be insufficient to meet patients’ needs. Evidence-based treatment, such as peer advocacy trainings, was sometimes unavailable to linguistic minorities, despite stakeholders’ acknowledgement of the importance of peer advocacy in recovery. Per participants, despite the linguistic diversity in NYS, language barriers were particularly salient for those whose primary language was less common (e.g., Arabic, Chinese, French).

3.4. Theme 3: Current Medicaid quality metrics may not adequately capture treatment quality

Participants discussed ways that quality care was determined and assessed in SUD treatment following the transition to MCOs. Three subthemes emerged: (1) the utility of quality metrics and assessments; (2) quality measures lack patients’ voices; and (3) quality of SUD care should include measures for individuals’ daily and social functioning.

3.4.1. Utility of quality metrics and assessments

Participants (policy leaders, administrators, and clinicians) reported the multiple quality metrics and assessments that they utilized. Participants explained that Medicaid in NYS evaluates SUD quality by tracking “service utilization, follow-up after hospitalization, detoxification services” and other claims data. The state tracked claims data but also required licensed providers to report patient outcomes data to track clinicians’ performance, whereas clinics often had their own internal quality assessments measures (e.g., patient satisfaction surveys, clinician evaluations). Stakeholders viewed the utility quality metrics differently by the role they had. For instance, policy leaders were interested in initiatives and practices to improve statewide care and explained that indicators were intended to assess quality and to “pinpoint how well the services are utilized, and to some extent, how well are people” (P6, Policy leader). Clinicians had to comply with state-mandated requirements by reporting indicators intended to demonstrate they were providing good care. However, clinicians were skeptical that the data reported to the state related to patient progress or outcomes:

You know, outcome measures in substance abuse treatment are very crude, and are not really excellent measures of progress in treatment or the health of the patient even, and counselors know that intuitively and so sometimes they kind of resent it, you’re ordering me to do this or that with an eye towards value based payment for the state, and that doesn’t really help my patient at all, or me, or even really measure the progress that they made.

(P21, Clinician)

Plan administrators approached state quality metrics in ways that ensured that they were achieving high quality ratings and remained competitive as a managed care organization: “[Managed care organizations are] competing on quality ratings, they’re competing on network … but they are trying to basically maximize their overall profitability” (P4, Plan administrator). However, most stakeholders questioned the validity of claims data metrics because they failed to capture the full story: “are we seeing low rates of utilization because there are no providers or is it because there are providers but for some reason, people are not utilizing [them]?” (P6, Policy leader). Instead, most clinicians and some plan administrators advocated for metrics that were actionable and could be used to improve services.

3.4.2. Quality measures lack patients’ voices

Despite some stakeholders’ use of state quality metrics to evaluate themselves, one common concern expressed across stakeholder groups was that patients’ voices were not included in developing quality metrics. For instance, one policy leader said: “When you look at the data collection system, it’s about what we found about the patients, it’s never about what the patient wants” (P1, Policy leader). Stakeholders saw the need to incorporate patient voices and experiences as significant for crafting metric evaluations of quality SUD and decision making:

It was really interesting to me one day, sitting in a room with all of these leaders in the community, and the question was ‘How come people with Medicaid, high risk people on Medicaid, keep on going to the ER for services they don’t need to go to the ER for?’ And you have all these people who, none of us have Medicaid, right? None of us are fitting the demographic, and I’m like ‘why can’t we ask someone who’s doing that?

(P28, Policy leader)

To bridge the gap in Medicaid state metrics and provide patient-centered care, clinicians in this study attempted to make treatment more accessible for patients, often making internal changes to care based on patients’ feedback. Inclusion of patients’ experiences in their assessments were intended to help move them away from outcome measures, such as hospitalizations, and toward more process-oriented data to inform immediate changes that clinicians can make within their organizations, such as getting feedback on how to make visits more convenient.

3.4.3. Quality of SUD care should include measures for individuals’ daily and social functioning

Policy leaders and clinicians alike wanted to incorporate social needs information and functioning levels as foundational SUD treatment quality measures, in particular, with securing and tracking housing information. One clinician explained:

it’s not something that we’re accustomed to tracking because again our focus has been the substance use and getting that stabilized, but because of the homelessness issue and residents in shelters or transient living family member living with friends, we’ve started talking about how we can possibly measure that [housing] given the impact that it has on their substance use.

(P23, Clinician)

Clinicians and policy leaders expressed difficulties with knowing whether SUD care was being optimally evaluated when outcome measures failed to capture how patients were functioning: “[We need to track] how they return to the workforce, and can maintain a job, have a good relationship/reunion with a family member, [if] they have more social leisure activities compared with before they came for treatment” (P24, Clinician).

Patients echoed this sentiment. Besides feeling welcomed and safe, patients generally regarded high quality treatment according to how well they functioned in daily activities, such as maintaining a job, resuming activities they enjoyed, and reconnecting with family and friends.

4. Discussion

The current study examined diverse stakeholders’ perspectives on factors related to access to quality of SUD care for those insured through NYS Medicaid. The study team conducted semistructured interviews and identified three themes. First, while transitions to managed care have benefitted clinicians, certain policies affect patients’ Medicaid enrollment and quality of care. Second, three groups were believed to be at high risk for inadequate treatment, including those with dual diagnoses, older adults, and linguistic minorities. Third, current Medicaid quality metrics may not adequately capture SUD treatment quality.

Medicaid remains critical to ensure that low-income communities receive SUD services (Olfson et al., 2018), as participants described the relatively accessible and adequate services available through outpatient care. Medicaid expansions and transitions to managed care have allowed for more comprehensive and high-quality SUD services through benefits expansions (Grogan et al., 2016). MCOs are well-positioned to better allocate resources to avoid excessive health care expenses and oversee staffing and accountability to ensure high quality SUD care. However, enrollment gaps hinder access to care, interfere with medication adherence, and are more likely to lead to hospitalization (Mojtabai, 2019; Tarazi et al., 2017). Our findings are consistent with existing literature that identifies insurance lapses due to patient income fluctuations as a significant barrier to SUD care (Maclean et al., 2019). Research has proposed extending eligibility 12 months after initial enrollment as a way to avoid insurance lapses, balance health care expenditures, and improve health of beneficiaries (Swartz et al., 2015). However, even though NYS became the first to adopt Medicaid eligibility extensions (Brooks et al., 2015), more research is necessary to understand why Medicaid disenrollment remains a barrier and why front-line providers might not be receiving NYS funds intended to cover uninsured SUD care for those in coverage transition.

Stakeholders had mixed opinions about utility of some SUD treatment policies and regulations crafted to improve SUD care. A critical component of successful policy implementation is examining where diverse stakeholders’ needs overlap and diverge to address gaps (Damschroder & Hagedorn, 2011; Greenhalgh et al., 2004). Whereas clinicians and policy-makers believed that prohibiting utilization management ensured emergency treatment access, plan administrators described it hindering successful transitions to outpatient care and patient-centered treatment. Lack of communication and coordination between stakeholders appears evident throughout the care continuum, which remains a barrier to improving SUD treatment and merits prioritization (Kehn et al., 2015).

Treatment that is leveraged or mandated has been posited as an efficient component of treatment for SUD recovery (Wild et al., 2012). However, these leverages are not associated with greater adherence to SUD care and patients can perceive them as coercive and punitive (Redlich et al., 2006). Mandated treatment can be stigmatizing if punishment is embedded in the broader way SUD care is structured, as patients in this study described. Policies and clinical guidelines that encourage the shifting of the philosophy of SUD care toward chronic disease management models (McLellan et al., 2014) and move away from acute episodic treatment should be implemented across managed care plans serving Medicaid enrollees. Changing overarching treatment models should include embedding addiction education and treatment in general medical training (Haack & Adger Jr, 2002), and prioritizing training to reduce provider stigma of patients with SUDs (Livingston et al., 2012). Only by shifting SUD care policies can we address provider shortages, lessen bias, and implement a whole-person approach to achieve high-quality care (Skidmore & Budd, 2017).

Participants in our study perceived SUD care to be inadequate for patients with dual diagnoses, older adults, and linguistic minorities, echoing previous findings (Antai-Otong et al., 2016; Guerrero et al., 2013; Wu & Blazer, 2011). The dearth of treatment centers to address comorbid SUD and severe psychiatric illneses was presented as the main barrier for those dually diagnosed. Our findings are consistent with existing literature that identifies the lack of treatment centers for dual diagnoses as a persistent problem in SUD care (Antai-Otong et al., 2016). Therefore, expanding centers and building on personnel capacity to work with dual diagnosis remains crucial, particularly as limited treatment centers in the United States are equipped to adequately treat them (McGovern et al., 2014). Research has shown certified community behavioral health centers to be a promising solution to increase access to SUD care for people with comorbid mental health and substance use issues (Kikkert et al., 2018). Treatment expansions into different settings (e.g., integrated in primary care, community behavioral health centers) and through varied modes of service delivery (e.g., telehealth) may also allow for low-barrier psychiatric services (Urada et al., 2014) to benefit patients with dual diagnosis and ease medication access barriers outlined by participants. Recent SAMHSA amendmends to the CFR 42 regulations allowing facilities to engage in more information sharing may improve quality of SUD care (U.S. Department of Health & Human Services, 2020) by facilitating better care coordination by MCOs (Tuck & Smith, 2019). However, insufficient availability of personnel to provide pharmacotherapy, regarded as critical for guideline concordant care for people with SUDs, remains a challenge (Kermack et al., 2017).

Stakeholders in this study perceived that SUD treatment was inadequate for older adults and linguistic minorities given the scarcity of tailored treatments and resources for both populations. Older adults with opioid use disorders are more commonly treated with methadone than their younger counterparts, increasing the need for monitoring and, subsequently, the risk for inadequate adherence (Harris et al., 2005; Neighbors et al., 2019). Researchers and practitioners should consider efforts to increase case management programs tailored to older adults and use of buprenorphine among this subgroup. Guidelines for treating patients of culturally diverse and linguistic-minority backgrounds have been outlined in the past (Center for Substance Abuse Treatment, 2006). However, these guidelines are loosely enforced nationally and rarely put into practice (Guerrero et al., 2014; Torrey et al., 2011). For instance, despite the federal implementation of the National Culturally and Linguistically Appropriate Services (CLAS) Standards in 2000 to increase cultural and linguistic competence (Office of Minority Health, 2001), most states still do not meet all standards almost two decades later (Aggarwal et al., 2017). While comprehensive adoption and implementation of CLAS standards may prove challenging (Aggarwal et al., 2017; Barksdale et al., 2014), the lack of payments to managed care plans to attend to specific populations may influence the lack of SUD care for these subgroups. SUD care providers should implement novel strategies to improve SUD care for linguistic minorities, older adults, and those with dual diagnosis.

Stakeholders in our study perceived current quality metrics as not representing SUD treatment quality and also failing to capture the experiences of individuals not interacting with the health care system. Our findings echo previous studies (Dunigan et al., 2014; Laudet, 2011; Morgenstern et al., 2008), suggesting that secure housing and employment are social needs that should be incorporated into these measures. Research should explore functional domains salient in promoting long-term recovery (Laudet, 2011). Currently, no consensus exists on which social outcomes should be assessed in clinical settings (Cantor & Thorpe, 2018) due to a lack of policy standards around utilization of social outcomes data (Freij et al., 2019). In response, NYS has an opportunity to create ways for SUD treatment centers to systematically track and improve patients’ social needs and quality of life. For example, based on the participants’ perspectives in this study, NYS could incorporate more patient-centered metrics of recovery (Neale et al., 2016; Neale et al., 2015) and require improved quality of life for value-based care (i.e., integration of primary care and SUD care history, including patient’s satisfaction scores while optimizing treatment) (Tseng & Hicks, 2016) and reimbursements (New York State Department of Health, 2015).

This study has several limitations. We interviewed stakeholders to understand their experiences with SUD care broadly, rather than focusing on specific substances or treatments, which would have allowed for more targeted responses than we could obtain. Our purposeful sampling recruitment approach resulted in patient participants only from New York City clinics, thus our patients’ experiences may not be reflective of patients’ experiences in NYS broadly. Stratified sampling techniques could help to more explicitly compare the experiences of stakeholders in diverse settings and evaluate the breadth of these findings. Finally, we had difficulty recruiting plan administrators for this study. Additional research should help us to further understand the differences among stakeholder groups. Despite these limitations, our study provides vital data for improvements on access and quality SUD care and opportunities for meaningful practice and policy changes.

5. Conclusion

Policies implemented in the past decade have facilitated comprehensive SUD services for low-income and vulnerable patients on Medicaid. Policy changes that allow for more collaborations among stakeholders, extend Medicaid coverage, and reduce stigmatizing care are still needed. Resources should be diverted toward facilitating psychiatric care for patients with dual diagnoses, and building capacity to work with older adults and linguistic minorities. Opportunities for NYS include tracking meaningful outcomes for patients (e.g., improvement in quality of life and social needs), and making insurance reimbursement contingent upon improvement in patients’ quality of life. Future research could examine stakeholders’ perspectives on these proposed solutions, including how to balance the need for quality SUD care with the extensive related costs.

Highlights:

• Medicaid remains critical for accessing SUD care for low-income communities.

• Policies around treatment leverages were perceived to add to the stigma, shame, and punitive nature of SUD care.

• Policies transitioning to managed care practices failed to address disparities in care for some individuals with dual diagnoses, older adults, and linguistic minorities.

• Medicaid quality metrics were not perceived to be adequate indicators of high-quality SUD treatment, particularly because they fail to capture improvements in patients’ quality of life and outcomes.