Variations in Acceptance of American Society of Addiction Medicine (ASAM) Continuum Levels of Care for Substance Use Disorder Treatment in an Urban Safety Net Primary Care Health Setting: A Qualitative and Quantitative Analysis and Implications for Health Equity

Abstract

Objectives:

To assess patients’ demographic factors associated with declination of recommended American Society of Addiction Medicine (ASAM) Continuum levels of care and analyze patients’ self-reported reasons for declination.

Methods:

This cross-sectional observational study examined data collected during clinical care delivery in a safety net health system’s primary care setting for patients receiving medication treatment and recovery support services for opioid and/or alcohol use disorder (OUD, AUD). We evaluated intake assessments to identify recommended levels of care, acceptance or declination of the recommendation, and self-reported reasons for declination. We stratified the acceptance data by demographic factors and analyzed reasons for declination to identify themes.

Results:

Of 1399 completed intakes, 42% declined all ASAM levels of care as a complement to medication and within-clinic recovery support services. Patients who identified as non-White, male, and/or 60 years or older were significantly more likely to decline all levels of care. Among the 125 patients who provided a documented reason for declining, 26% endorsed work, financial, or family obligations; 26% indicated a preference for alternative forms of psychosocial support; and 14% indicated possible interest in the future.

Conclusions:

It is important for substance use disorder treatment providers and policymakers to understand the barriers patients face regarding engagement in treatment and reasons for the declination of services. Recognizing variations in care acceptance by demographic factors can help us understand common barriers and guide future directions to address health inequities.

The United States remains in a drug overdose public health crisis, largely driven by illicit opioids. A key component to addressing this crisis includes improving access to evidence-based substance use disorder (SUD) treatment. The Substance Abuse and Mental Health Services Administration (SAMHSA) estimates that less than a quarter of individuals who need SUD treatment receive it.¹ Among adults with an SUD who did not receive treatment, 94.7% either did not seek treatment or did not believe they should seek treatment.¹ Patients’ reasons for not seeking treatment are underexplored and may inform efforts to lower barriers.

Current guidelines call for the availability of formal psychosocial counseling and supportive environments of varying intensity along the American Society of Addiction Medicine (ASAM) levels of care (subsequently referred to as “levels of care” or “level of care”). These are meant to complement medications for opioid and alcohol use disorder (MOUD and MAUD) and recovery support services, but not to be considered a requirement for access to medication.² One approach to improving access to and acceptability of SUD care is offering multimodal treatment options integrated into or colocated with primary care clinics, including MOUD/MAUD, a variety of patient-centered psychosocial treatments, harm reduction services, and recovery supports.³ We performed a cross-sectional observational study to assess patients’ demographic factors associated with a declination of recommended ASAM levels of care and identify patients’ self-reported reasons for declining formal SUD psychosocial treatment.

METHODS

Study Setting and Participants

This pragmatic study utilized data collected during usual clinical care delivered at a safety net health system in Chicago, IL, and surrounding suburbs. The health system offers MOUD/MAUD and recovery support services integrated into its adult-serving primary care health centers. If a patient has an OUD or AUD and is interested in these integrated services, an in-person or virtual intake assessment is completed by a recovery coach, who is a Certified Alcohol and Drug Counselor (CADC). The intake assessment is a multidimensional assessment of physical, mental, and social health, based on the ASAM Criteria.⁴ The ASAM Criteria is a resource to guide patient assessment and recommendation of the most appropriate level of care, from inpatient setting to community-based care. In Illinois, the CADC certification includes training on comprehensive assessments, and the recovery coaches receive annual retraining on the ASAM criteria. Based on the intake assessment, the recovery coach usually recommends a level of care to the patient as a complement to medication treatment, recovery support, and wraparound services (ie, primary care, transportation support) provided within our system. The coaches use motivational enhancement techniques to address concerns, are trained to address structural/social determinants of health, and are trained to use basic terms and avoid acronyms and jargon such as “IOP.” The recovery coaches use their independent clinical judgment to develop a recommendation, with time for asking and answering questions. Recommendations are reassessed over time. The recovery coach documents patients’ acceptance or declination of the recommended level of care and provides referrals for any services accepted but not offered internally. If the patient declines the recommended level of care, the recovery coach may document the reason provided by the patient. Regardless of whether accepting or declining the recommended level of care, a patient is eligible to and encouraged to receive internal services, including MOUD/MAUD and recovery support. This study was approved by the Institutional Review Board of Cook County Health. Confidentiality of participant information was maintained throughout the study, and data were anonymized for analysis.

Data Collection

We evaluated intake assessment data from three calendar years, 2019–2021. Intake assessments were completed in person or by telephone. Recovery coaches recorded patients’ verbal responses in standardized and structured electronic medical record (EMR) forms. The level of care recommendations and acceptance/declination were captured in the following structured fields: “outpatient” to represent ASAM level 1, “intensive outpatient” to represent ASAM level 2, and “residential” to represent ASAM level 3 or higher.⁴ Patient responses to intake assessments, including reasons for declination, were extracted from the EMR. Study participant age, race, ethnicity, and gender data were retrieved from the health system’s research data warehouse. Co-occurring mental health diagnosis and educational attainment were captured by self-report and extracted from the EMR.

Data Analyses

For our analyses, we grouped age into categories (20–39 y old; 40–59 y old; 60–79 y old), and we evaluated self-reported mental health diagnoses (yes/no). We grouped patients who declined an ASAM level of care with those for whom an ASAM level of care was not recommended (ie, solely recovery coach support) with the rationale that the recovery coaches sometimes solely recommended recovery coach support based on person-centered considerations. We grouped those already connected to an ASAM level of care with those who accepted the recommended level of care during the intake process with the rationale that our goal was to understand the acceptance of levels of care. Descriptive statistics were calculated for categorical variables and results of intake assessments—ASAM-level recommendation and acceptance or declination—were reported stratified by the aforementioned categories: outpatient, intensive outpatient, and residential. χ² statistics were calculated using Stata Statistical Software (Version 18.0; StataCorp) to analyze differences in treatment acceptance by demographic factors.

Free-text reasons for declining referral to ASAM levels of care were grouped into thematic categories using inductive coding methodology. Declination categories were created by 3 authors who independently identified themes and then agreed on the categories by consensus (J.N.K., E.G., and D.C.), with input from a fourth author (K.H.). Descriptive statistics of count and percentage were calculated for each declination category, allowing for an individual to cite >1 reason.

RESULTS

During the study period, 1448 intakes were conducted with individuals with an active OUD or AUD who expressed interest in commencing MOUD/MAUD, with 1399 (97%) confirmed to be complete (Table 1). Of these 1399, 3 were completed in the inpatient hospital, 8 in the emergency department, and 1388 (99.2%) in primary care. The vast majority of patients presented to initiate buprenorphine, with a handful of patients managed with naltrexone for OUD and/or AUD. Over half were 40–59 years old, and the majority identified as male, African American/Black, and of non-Hispanic origin. About half of individuals reported a co-occurring mental health diagnosis, and ~1 in 3 had not completed high school (Table 1).

TABLE 1.

Study Participant Demographics

N	1399
Age groups	n (%)
20–39	434 (31)
40–59	742 (53)
60–79	223 (16)
Gender
Females	515 (37)
Males	879 (63)
Transgender or not defined	5 (<1)
Race
African American/Black	754 (54)
White	540 (39)
All other races	105 (7)
Ethnicity
Hispanic/Latino/Spanish origin	118 (8)
Non-Hispanic/Latino/Spanish origin	1278 (91)
Not defined	3 (<1)
Mental health condition*
Co-occurring mental health diagnosis	711 (51)
No co-occurring mental health diagnosis	685 (49)
No answer provided	3 (<1)
Level of education attained*
Less than high school	61 (4)
Some high school	375 (27)
High school/G.E.D./vocational diploma	539 (39)
Some college	325 (23)
College graduate or higher	96 (7)
No answer provided	3 (<1)

^*By self-report.

Of the 1399 complete intake assessments (Fig. 1), nearly all resulted in a documented referral to a level of care, with approximately one third referred to each category: residential, intensive outpatient, and outpatient. Forty-four patients did not have a documented referral to a level of care and were referred for recovery coach services only. A little over half of patients (n=810, 58%) accepted a level of care referral or indicated already being engaged in a level of care. Thirteen percent (n=181) of patients accepted referral to residential treatment, 14% (n=202) to intensive outpatient, and 31% (n=427) to outpatient. Five hundred eighty-nine (42%) chose not to engage in any level of care. Of these 589 individuals, 125 (21%) provided one or more reasons for the decline.

FIGURE 1.

Flowchart displaying results from all completed intake assessments, 2019–2021. Recommendations to American Society of Addiction Medicine (ASAM) level of care and patient acceptance or declination of recommended levels of care.

This study found variation in the acceptability of levels of care by participant characteristics (Table 2). Compared with white patients, black patients and patients of other non-white races were significantly more likely to decline all levels of care. Male patients were significantly more likely to decline all levels of care compared with female patients, and patients 60 years and older were significantly more likely to decline all levels compared with patients 20–59 years old. Patients without a self-reported co-occurring mental health diagnosis were significantly more likely to decline levels of care compared with patients with a co-occurring diagnosis. The level of educational attainment was not associated with acceptance of a level of care.

TABLE 2.

Count and Percentage of Acceptance of ASAM Levels of Care, or Declination of All Levels of Care, Stratified by Patient Demographic Factors

	Total	Residential	Intensive outpatient	Outpatient	Declined all levels
Demographic variable	N	n (%)	n (%)	n (%)	n (%)
Race
Black/African American*	754	83 (11)	98 (13)	210 (28)	363 (48)
White	540	78 (14)	87 (16)	193 (36)	182 (34)
All other races*	105	20 (19)	17 (16)	24 (23)	44 (42)
Ethnicity
Hispanic/Latino/Spanish origin	118	13 (8)	19 (15)	28 (25)	58 (52)
Non-Hispanic/Latino/Spanish origin	1278	166 (12)	183 (14)	399 (31)	530 (43)
Gender
Female	515	105 (20)	71 (14)	145 (28)	194 (38)
Male**	879	76 (9)	129 (15)	281 (32)	393 (45)
Age
20–39	434	70 (16)	82 (19)	146 (34)	136 (31)
40–59	742	94 (13)	100 (13)	215 (29)	333 (45)
60–79*	223	17 (8)	20 (9)	66 (30)	120 (54)
Mental health diagnosis (Dx)^
Co-occurring Dx	711	116 (16)	123 (17)	216 (30)	256 (37)
No co-occurring Dx*	685	65 (9)	78 (11)	209 (31)	333 (49)
Education level†
Less than 12th grade	438	63 (14)	64 (14)	134 (30)	177 (42)
High school graduate/G.E.D./vocational diploma	540	61 (11)	83 (15)	172 (31)	224 (43)
Some college	325	45 (14)	43 (13)	93 (29)	144 (44)
Some graduate or higher	96	12 (13)	12 (13)	28 (29)	44 (46)

Data field with <11 observations are suppressed.

G.E.D. indicates General Educational Development test.

^* P<0.01.

^** P<0.05.

^† By self-report.

Of the 589 patients who did not accept a referral to any ASAM level of care, 125 (21%) provided at least one free-text reason documented by the recovery coach for a total of 148 reasons (Table 3). More than 1 in 5 (n=28, 22%) individuals identified work commitments or financial obligations as their reason for declining. More than 1 in 5 (n=27, 22%) indicated not being interested in more intense care, not feeling it was necessary, or wanting to try medication and recovery support in their health center first (combined under “not interested” in Table 3). Eighteen (14%) individuals declined ASAM levels at the time of intake but indicated they may be interested in the future. Thirty-two individuals (26% of respondents) indicated a preference for peer support groups, utilizing within-clinic recovery coach support, and/or currently receiving psychosocial support at a recovery home environment, with 6 individuals reporting 2 of these 3 reasons for declination. Eight individuals (6%) reported being on electronic monitoring with the criminal legal system and perceiving that this would interfere with engagement. Five individuals (n=4%) reported family obligations such as “caring for 7-year-old granddaughter” and “caring for medically frail mother.” Additional reasons are provided in Table 3.

TABLE 3.

Reasons for Declining ASAM Levels of Care

N=125 respondents providing N=148 response	N=148 responses (% of respondents)
Social/structural constraints
Prioritizing work commitments	27 (22)
Perceived challenges related to electronic monitoring	8 (6)
Prioritizing family obligations	5 (4)
Structural concerns: not want to be in a facility, rules, feels like a jail, concern for COVID	4 (3)
Prioritizing financial obligations	1 (1)
Geographic barrier: moving soon	1 (1)
Treatment preference-associated
Prefers peer support groups	16 (13)
Prefers recovery coach support	12 (10)
In a recovery home or sober living environment	10 (8)
Prefers outpatient therapist	6 (5)
Recently completed treatment	2 (2)
Prefers faith-based support	1 (1)
Current general lack of interest
Not interested	27 (22)
May be interested in the future	18 (14)
Health concerns
Health-related reasons	10 (8)

DISCUSSION

This pragmatic, cross-sectional study analyzed intake assessments completed in an urban underserved primary care setting among patients seeking MAUD or MOUD. The participants were connected to a recovery coach for integrated recovery support services as well as further assessment including ASAM level of care treatment recommendations.

Although more than half of the study participants accepted a referral to an ASAM level of care through our primary care clinical site, a substantial minority (42%) declined participation in any ASAM level of care at the time of assessment. A level of care different from the one recommended was accepted by 129 patients (9%). This pragmatic study reveals the sometimes unexpected outcomes of clinical conversations between patients and recovery coaches. We do not have access to data that could inform the clinical significance of discordance between recommendations and placements.

These findings suggest that acceptance of ASAM levels of care for many individuals is based on considerations and factors that are independent of the decision to take medications for addiction treatment. Patients in our study had the opportunity to decide what types and settings they prefer for the provision of SUD treatment and recovery support services, and >40% chose medication therapy in the primary care setting with integrated recovery support services and without more intensive engagement. Patients offered a variety of reasons for declining referral to external clinical sites for intensive psychosocial support, which should inform efforts to provide patient-centered options. More than 1 in 4 (26%) described prioritization of employment, family, and financial commitments as their reason for declination. One patient indicated concern for losing their employment if they engaged in intensive psychosocial support. For individuals with employment that is insecure, temporary, or does not offer paid sick leave, taking off time on a regular basis for outpatient or intensive outpatient, or missing work for several weeks for residential treatment, is unlikely to be a realistic option, especially for those who are the primary financial provider for themselves and/or family. Similarly, the patient describing serving as her granddaughter’s caregiver may be interested in intensive psychosocial care but may not have a realistic pathway for that to be an option at that moment.

Our findings are consistent with those reported in other studies. In focus groups with women, Elms et al⁵ found that key concerns and hesitations toward accessing intensive psychosocial treatment were related to the welfare of their children. In a systematic review of barriers to accessing medications for addiction treatment, Hall et al⁶ identified the lack of flexibility or impact on daily life (maintaining job, childcare, and daily commitment interference) as a key theme. SAMHSA describes health, home, purpose, and community as the pillars of sustained recovery.⁷ Acknowledging and respecting the factors that feed these pillars—such as caring for a child or maintaining employment to afford a home—when supporting a patient to develop their treatment plan may be crucial for improving health outcomes and sustaining these improvements.

Individuals who identified as Black/African American were significantly more likely to decline all ASAM levels of care compared with patients who identified as white. Past research has explored whether differences in perceptions of treatment needs by race/ethnicity are associated with disparities in treatment utilization.⁸ Our study adds to the literature in 2 important ways. First, all individuals in this study were seeking MOUD/MAUD and engaging with recovery support services. In other words, all were indicating a perceived need and seeking treatment. Second, this study identifies a racial disparity in acceptance, specifically of the ASAM levels of care, and suggests more attention may be needed on alternative pathways to sustained recovery.

The National Survey on Drug Use and Health (NSDUH) indicates that almost 95% of individuals with a substance use disorder who did not receive SUD treatment did not seek treatment or did not feel they needed treatment.¹ One in 4 individuals in our study identified their recovery home’s support, within-clinic recovery coaches, and peer-led support meetings as their preferred psychosocial support. Yet those are not included in NSDUH’s definition of treatment. Perhaps it is time to formally expand what is considered “treatment.”

Twenty-two percent endorsed feeling they did not need ASAM level of care treatment, they were not interested, they had similar prior treatment episodes and wanted something different, and/or they wanted to try medication alone. Despite ASAM recommending offering medication as a key treatment component for opioid use disorder regardless of psychosocial engagement in their National Practice Guideline for the Treatment of Opioid Use Disorder,⁹ there is a perspective among some in the health care community that for medication treatment to be effective, it must be paired with intensive psychosocial support.¹⁰ How to most effectively make voluntary psychosocial support available within or external to primary care is not known,³ but only offering limited options in a siloed fashion will leave some individuals out. Had our patients been evaluated in a setting where adjunctive counseling was required by policy or an unspoken expectation, they may have been denied access to life-saving medication treatment.

Many individuals indicated interest in psychosocial supports, but not enrollment in ASAM levels of care. Thirty-two individuals (26% of respondents) indicated a preference for peer support groups, utilizing within-clinic recovery coach support, and/or currently receiving psychosocial support in a recovery home environment. These community and clinic-based supports might enhance individual recovery capital, especially one’s sense of community, in a lasting and sustainable way. Eighteen (14%) individuals declined ASAM levels at the time of intake but indicated they may be interested in the future. Decreased cravings and withdrawal once a medication for addiction treatment is initiated and working may foster greater receptivity to ASAM levels of care, and recommendations were revisited regularly throughout care.

Lastly, a handful (n=8, 6%) of patients reported being on electronic monitoring with the criminal legal system as their reason for declination. Whether a real or perceived barrier, criminal legal system engagement should not interfere with access to any evidence-based SUD treatment modality.

It is noteworthy that the demographic of patients more likely to decline external intensive services (black, male, and older) reflects populations that have the highest age-adjusted rates of drug overdose death.¹¹ Failure to address barriers and develop policies and practices that enable patient-driven preferences for SUD care may contribute to disparities in mortality rates. It is essential for SUD treatment and care providers to understand that a patient choosing to decline intensive psychosocial treatment does not mean they do not want to reduce harm, improve health, and engage in positive change. Meeting patients where they are and hearing their reasons for their treatment engagement choices is critical, and regulation and payment mechanisms that support more flexible pathways to recovery ought to be explored.

Fortunately, policy and practice changes are evolving to be more patient-centered and elevate the patient's voice. In its first revision in over 20 years of the federal regulations of Opioid Treatment Programs, or providers of community-based methadone, SAMHSA clarifies that methadone clinics must offer counseling but can no longer require participation in counseling for patients to access methadone.¹² The ASAM Criteria is a resource to guide patient assessment and recommendation of the most appropriate level of care, from inpatient setting to community-based ambulatory care.⁴ The newest version of the ASAM Criteria includes a new Person-Centered Considerations Dimension, emphasizing the need to consider social determinants of health and patient preference in the level of care recommendations. This new dimension can support treatment providers to help patients enhance their pillars of recovery—health, home, purpose, and community. Offering bus cards to intensive outpatient treatment, identifying job training as part of the treatment plan, and addressing housing insecurity all feed these pillars and address barriers to engagement.

Another budding practice change that may address barriers identified by patients is the use of telehealth to deliver psychosocial supports including ASAM levels of care.¹³ Telehealth may offer flexibility to a person serving as a primary caregiver with transportation challenges or employed full-time. This is an area ripe for further attention.

Lastly, more than half of the study participants accepted a referral to an ASAM level of SUD treatment. This suggests interest among many patients presenting to primary care settings for MOUD/MAUD and recovery support services to engage in the more intensive psychosocial supports provided through the ASAM levels of care. Providers in similar settings should strive to ensure seamless pathways for patients entering their care to engage in optional, complementary, more intensive services not offered internally.

Limitations

There are several important limitations to this study. First, the data were collected in real-world, clinical settings during routine patient care. Documentation errors can occur in a busy clinic, and while coaches were expected to document verbatim responses for declining, some may have paraphrased patients' actual responses, and there is the potential for misrepresentation. Interviewer bias may have played a role, particularly if patients felt that the recovery coaches wanted them to accept an ASAM level of care. Just over 20% of patients provided a reason for declination, and the patients who provided reasons were not a random sample but rather a convenience sample and therefore may over or under-represent overall reasons for declination. Patients also may not have felt comfortable providing all reasons for declination to the coach. The authors hope to replicate this study with additional years of data to have a larger sample and post-assessment outcomes.

Forty-four patients did not receive a recommended ASAM level of care and instead were identified for recovery coach-only support. There were inconsistencies between coaches regarding how they would document when a patient was only interested in MOUD and recovery support services, with some documenting the referral and the decline but others not documenting any referral and indicating a recommendation for recovery coach services only.

We did not examine whether declination of care was associated with the recommended level of care; this is an area of potential future exploration. Comorbid mental health diagnosis was self-reported and not cross-checked with the diagnosis list. Lastly, the study period included the peak of the COVID-19 pandemic, and the availability of certain levels of care, such as residential, may have affected patients’ acceptance of such levels.

Our study setting is a safety net health care setting that serves a diverse, urban population. Findings may be limited to settings serving similar populations. In addition, a key component of our care delivery is the ability to offer recovery support services embedded in primary care.

It is not a standard of care locally or nationally to have this team member available in primary care, and reimbursement models are not readily in place to support this integrated model. This could be an area of policy exploration, including the cost-effectiveness of such a care model.¹⁴

CONCLUSIONS

As substance use disorder (SUD) treatment providers work to facilitate care for all patients, it is important for them to understand the barriers patients face regarding engagement in psychosocial treatment and reasons for declination of services. A better understanding of what factors affect a patient’s interest in and ability to engage in recommended levels of psychosocial SUD care may support treatment providers in the implementation of evidence-based and patient-centered treatment options. Recognizing variations in care acceptance by demographic factors can help practitioners and policymakers understand common barriers and guide future directions to address health inequities. Supporting patients to choose from a menu of evidence-based medications and psychosocial supports, including but not limited to ASAM levels of care, may result in greater engagement in SUD treatment overall as compared with a setting with offerings limited to ASAM levels of care. In other words, full implementation of the Person-Centered Considerations Dimension of the ASAM Criteria.