Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Sep 1.
Published in final edited form as: J Subst Abuse Treat. 2019 Jun 15;104:135–143. doi: 10.1016/j.jsat.2019.06.010

A comparison of buprenorphine and psychosocial treatment outcomes in psychosocial and medical settings

Ned J Presnall a,b, DA Patterson Silver Wolf a, Derek S Brown a, Sara Beeler-Stinn a, Richard A Grucza b
PMCID: PMC7075557  NIHMSID: NIHMS1563407  PMID: 31370977

Abstract

Background

Facing an epidemic of opioid-related mortality, many government health departments, insurers, and treatment providers have attempted to expand patient access to buprenorphine in psychosocial substance use disorder (SUD) programs and medical settings.

Methods

With Missouri Medicaid data from 2008 to 2015, we used Cox proportional hazard models to estimate the relative hazards for treatment attrition and SUD-related emergency department (ED) visits or hospitalizations associated with buprenorphine in psychosocial SUD programs and medical settings. We also tested the association of buprenorphine with hours of psychosocial treatment during the first 30 days of psychosocial SUD treatment. The analytic sample included claims from 7,606 individuals with an OUD diagnosis.

Results

Compared to psychosocial treatment without buprenorphine (PSY), the addition of buprenorphine (PSY-B) was associated with a significantly reduced hazard for treatment attrition (adjusted hazard ratio: 0.67, 95% CI: 0.62–0.71). Among buprenorphine episodes, office-based (B-OBOT), outpatient hospital (B-OPH), and no documented setting (B-PHA) were associated with reduced hazards for treatment attrition when compared to the psychosocial SUD setting (B-PSY) (adjusted hazard ratios: 0.27, 95% CI: 0.24–0.31; 0.46, 95% CI:0.39–0.54; 0.70, 95% CI: 0.61–0.81). Compared to B-PSY, B-OBOT and B-PHA were associated with significantly reduced hazards for a SUD-related ED visits or hospitalization (adjusted hazard ratios: 0.59, 95% CI: 0.41–0.85; 0.53, 95% CI: 0.36–0.78). There was no significant difference between B-PSY and B-OPH or B-PSY and PSY in hazard for an SUD-related ED visit or hospitalization.

Conclusions

Our findings support the conclusion that adding buprenorphine to Medicaid-covered psychosocial SUD treatment reduces patient attrition and SUD-related ED visits or hospitalizations but that buprenorphine treatment in office-based medical settings is even more effective in reducing these negative outcomes. Policy-makers should consider ways to expand buprenorphine access in all settings, but particularly in office-based medical settings. Buprenorphine treatment in an unbilled setting was associated with an increased hazard for patient attrition when compared to treatment in billed medical settings, indicating the importance of Medicaid-covered provider visits for patient retention.

Keywords: buprenorphine, intensive outpatient, Medicaid, OBOT, office-based opioid therapy, opioid use disorder, OUD, pharmacotherapy, primary care, psychosocial

1.1. Introduction

Opioids were involved in 42,249 U.S. deaths in 2016, a 27.6% increase over the prior year (Seth, Scholl, Rudd, & Bacon, 2018). Provisional estimates from 2017 suggest that fatal drug overdoses have continued to rise in most states; from December 2016 to December 2017, the U.S. had an estimated 72,306 fatal drug poisonings, an increase of 9.5% over the prior year (Ahmad, Rossen, Spencer, Warner, & Sutton, 2018).

A strong scientific consensus supports maintenance buprenorphine or methadone pharmacotherapy as the first-line treatment for opioid use disorder (Connery, 2015; Volkow, Frieden, Hyde, & Cha, 2014). Compared to psychosocial treatment alone, methadone and buprenorphine increase retention and reduce illicit opioid use (Clark et al., 2015; Mattick, Breen, Kimber, & Davoli, 2014). Treatment with either medication is associated with a reduction in mortality risk compared to being out of treatment (Sordo et al., 2017) and methadone has been shown to be associated with decreased criminal activity and disease transmission (Marsch, 1998; Sordo et al., 2017). In the largest cohort study of drug related poisoning to date, patients’ risk for fatal drug-related poisoning when enrolled in a psychological intervention without medication was double that of patients in buprenorphine or methadone treatment (Pierce et al., 2016). Methadone and buprenorphine were FDA-approved for OUD in 1972 and 2002 respectively, but in 2016, only 21% of Medicaid-funded outpatient treatment programs in the U.S. (excluding DUI/DWI programs) provided maintenance buprenorphine or methadone treatment (SAMHSA, 2018a). In 2016, an estimated 58% and 74% of Medicaid-funded OUD outpatient treatment episodes in the U.S. and Missouri, respectively, comprised psychosocial treatment without medication (SAMHSA, 2018b).

1.2. Medical and psychosocial OUD treatment

Three types of long-term OUD treatment program developed in the United States during the 20th century: long-term residential, outpatient “drug free” (i.e. psychosocial), and outpatient methadone (Hubbard, Craddock, & Anderson, 2003). Historically, long-term residential and outpatient psychosocial programs provided psychosocial services of various intensity but no opioid agonist therapy. Many of these programs, based in the 12-step tradition, viewed methadone as a barrier to full recovery, a position formalized by Narcotics Anonymous in 1997 (World Services Board of Trustees, 1997). Outpatient methadone programs (OTPs) offered counseling, but patients in OTPs generally received fewer counseling services than patients in long-term residential or psychosocial outpatient programs (Etheridge, Craddock, Dunteman, & Hubbard, 1995).

The FDA-approval of buprenorphine in late 2002 offered psychosocial programs a safer, less regulated medication to treat OUD at a time when the number and proportion of OUD treatment admissions were rising (SAMHSA, 2010). Psychosocial programs faced several barriers to buprenorphine adoption, including poor access to prescribers, financial constraints, insufficient training, and oppositional views toward pharmacotherapy among staff (Aletraris, Edmond, Paino, Fields, & Roman, 2016; Friedmann, Jiang, & Alexander, 2010). Hazelden-Betty Ford, one of the most famous and influential psychosocial treatment programs, reluctantly integrated buprenorphine and contrasted its practice of “transitional” pharmacotherapy with the long-term medication maintenance practiced in other settings (Seppala & Larson, 2015). Overall, publicly funded psychosocial programs have been less likely than privately funded programs to adopt OUD pharmacotherapies, creating an economic disparity in patient access (Abraham, Knudsen, Rieckmann, & Roman, 2013).

Several states have made systematic attempts to increase the availability of buprenorphine in settings that already employ a medical model of disease management. Vermont’s “hub and spokes” system uses OTPs as specialty providers that can refer to office-based settings for maintenance treatment (AHRQ, 2016; Brooklyn & Sigmon, 2017). Rhode Island and Maryland have implemented Medicaid Health Homes in OTPs and psychiatric clinics (Korthuis et al., 2017). Massachusetts has developed a “nurse care manager” model in its federally qualified health centers, and New Mexico has used its Project Extension for Community Healthcare Outcomes (ECHO) to support the expansion of buprenorphine treatment in primary care, particularly in rural areas (Komaromy et al., 2016; LaBelle, Han, Bergeron, & Samet, 2016). In contrast to these initiatives, Missouri’s Department of Mental Health has focused on expanding buprenorphine treatment in its network of more than 140 psychosocial treatment programs; the department began this initiative more than a decade ago but formalized and incentivized buprenorphine treatment with funds from the State Targeted Response to the Opioid Crisis grant (Knopf, 2018).

The overall benefit and optimal dosing of psychosocial treatment as an adjunct to pharmacotherapy is unclear. The most recent Cochrane review found that adding any psychosocial treatment, such as contingency management or cognitive behavioral therapy, to a standard OTP program had no effect on retention or illicit opioid use (Amato, Minozzi, Davoli, & Vecchi, 2011). A comparison of intensive (9 hours per week) versus supportive (2–8 hours per week) outpatient treatment in patients receiving buprenorphine therapy showed no difference in clinical outcomes between the two treatment conditions (Mitchell et al., 2013). In separate studies of office-based opioid treatment with buprenorphine (B-OBOT), neither the addition of low-dose (tapered from weekly to monthly) cognitive behavioral therapy nor enhanced weekly counseling sessions with a nurse care manager (45 minutes versus 20 minutes) were associated with significant differences in patient opioid use or study completion (Fiellin et al., 2006, 2013). In a 12-week trial of buprenorphine maintenance in persons who had returned to opioid use during a buprenorphine taper, there was no significant difference in treatment success between patients who received standard medical management and those who received standard medical management plus opioid dependence counseling (Weiss et al., 2011).

We planned our analyses to examine the association of buprenorphine treatment with patient retention and clinical outcomes (SUD-related emergency department (ED) visits or hospitalizations) in psychosocial and medical settings. We sought also to address conflicting assumptions in the SUD treatment field. In OTPs and OBOT settings, retention in pharmacotherapy is a central clinical objective due to its association with decreased illicit opioid use, mortality, disease transmission, and crime (Marsch, 1998; Mattick et al., 2014; Sordo et al., 2017). In psychosocial treatment programs, methadone and buprenorphine are commonly viewed as a barrier to full recovery (Aletraris et al., 2016). When used in these settings, pharmacotherapy is often subject to rapid tapering requirements or contingent upon compliance with a psychosocial treatment plan (Bentzley, Barth, Back, & Book, 2015). Missouri’s efforts to integrate maintenance buprenorphine treatment into traditionally psychosocial treatment programs have highlighted these philosophical tensions (Knopf, 2018). We attempted to analyze each approach on its own terms. We examined whether buprenorphine therapy enhanced retention and engagement in psychosocial treatment programs. Similarly, we compared buprenorphine treatment retention in psychosocial, outpatient hospital, and office-based settings. Finally, we examined the association of psychosocial treatment, buprenorphine therapy, and primary treatment setting on the hazards for a SUD-related ED visit or hospitalization.

2. Materials and Methods

2.1. Overall objectives and planned analyses

We sought to compare the outcomes of psychosocial OUD treatment with and without buprenorphine and, separately, to compare the outcomes of buprenorphine treatment in psychosocial OUD programs and office-based settings, such as federally qualified health centers, rural health clinics, and private medical practices. Preliminary analyses revealed two other types of buprenorphine episode: outpatient hospital and pharmacy-only episodes. In the latter, buprenorphine pharmacy claims had no associated prescriber visits. In these episodes, patients may have paid out-of-pocket for their prescriber visit but used pharmacy benefits for their medication.

We performed separate analyses of two types of treatment episode. In the first analysis, psychosocial treatment episodes were defined by continuous treatment in a state-certified psychosocial treatment program, a type of program generally required to offer partial hospitalization, intensive outpatient, and supportive outpatient treatment (9 CSR 30–3.10); psychosocial episodes were classified by the presence or absence of buprenorphine therapy during the episode. In a second, separate analysis, buprenorphine episodes were defined by continuous treatment with buprenorphine and were classified by treatment setting, which included psychosocial programs, office-based settings, outpatient hospitals, and episodes with no billed setting. Finally, we compared the association of psychosocial treatment without buprenorphine and buprenorphine treatment by treatment setting on the hazards for a SUD-related ED visit or hospitalization.

2.2. Data

As part of a broader study, we obtained Medicaid claims data for 2008 to 2015 from the Missouri Department of Social Services for all health services received by eligible members with a SUD diagnosis (ICD-9 291.x, 304.0x, 304.x, 303.x, 305.x, 760.71, and 779.5) during this period. Claims included eligibility status, enrollment dates, and basic demographics (age, sex, race/ethnicity, and county). Before any additional selection criteria were applied, the data included 200,164 individuals and 286,484 eligibility episodes (continuous periods of Medicaid enrollment of 45 days or longer) (Figure 1). For the present study, we extracted all data associated with OUD treatment, including all psychosocial services with an OUD diagnosis and all prescriptions for buprenorphine and extended release (XR) naltrexone.

Figure 1.

Figure 1.

Open in a new tab

Treatment episode and group definitionsa

2.3. Variables

2.3.1. Psychosocial and buprenorphine treatment episodes

SUD treatment for Medicaid beneficiaries and the uninsured in Missouri takes place primarily in Missouri’s Comprehensive Substance Abuse Treatment and Rehabilitation Programs (CSTARs) which are funded by the Missouri Department of Mental Health (DMH) and are required to offer three levels of care: “Primary Treatment,” “Intensive Outpatient Treatment,” and “Supported Recovery.” These levels approximate American Society of Addiction Medicine (ASAM) levels 2.5 (partial hospitalization), 2 (intensive outpatient), and 1 (supportive outpatient) respectively (Mee-Lee, Shulman, Fishman, Gastfriend, & M., 2013; 9 CSR 30–3.130). In general, CSTARs are the only settings permitted to bill Medicaid using SUD-specific CPT codes such as H0004 for individual counseling. Missouri’s menu of Medicaid-funded psychosocial SUD procedures does not include program-level codes for intensive outpatient (H0015) or partial hospitalization (H0035), codes that generally indicate multiple hours per day of bundled psychosocial services. Instead, it includes more basic units of psychosocial treatment such as group therapy (H0005), group education (H0025), and individual therapy (H0004). CSTARs optionally provide room and board to patients, but this service is not reimbursed or included in Medicaid claims. Outside the DMH-contracted CSTAR system, Medicaid providers can offer buprenorphine treatment with or without adjunctive psychosocial services in medical settings, such as federally qualified health centers, rural health clinics, and outpatient hospitals. These settings use office-based codes for psychosocial services such as 90834 for individual counseling. Although CSTARs and hospitals can each offer Medicaid-funded inpatient detoxification, we excluded these services from our analyses as we were primarily interested in long-term outpatient treatment. In summary, our data set includes psychosocial and buprenorphine treatment services rendered in the DMH-contracted psychosocial CSTAR system which uses SUD-specific procedure codes and in medical settings, which use more general office-based procedure codes for psychosocial treatment. Both systems use the same procedure codes for prescriber visits; prescriber settings can be distinguished by place of service and revenue codes.

Our treatment episode definitions and groupings are schematically illustrated in Figure 1. We defined two types of treatment episodes: psychosocial and buprenorphine. Psychosocial episodes (n=8,516) were identified using a revenue code that identifies psychosocial treatment. The start of an episode was the first date of service in a psychosocial program with a primary OUD diagnosis. The end of an episode was the last date of service prior to a gap in service greater than 45 days. Psychosocial episodes were classified by the presence (PSY-B) or absence (PSY) of buprenorphine during the episode using National Drug Codes for OUD-specific buprenorphine products on Missouri’s preferred drug list. These same National Drug Codes were used to identify buprenorphine episodes.

Buprenorphine episodes (n=4,277) were characterized by the continuous provision of buprenorphine therapy. The first buprenorphine pharmacy claim comprised the start of an episode; a gap in buprenorphine supply of more than 45 days marked the end of a buprenorphine episode. Buprenorphine episodes were further classified by buprenorphine prescriber setting. To characterize prescriber setting, we linked each buprenorphine pharmacy claim to the most recent prescriber claim with an OUD diagnosis during the two weeks prior to and including the date of the pharmacy claim. Prescriber claims were identified by CPT codes beginning with ‘99’ or indicating medication management by a psychiatrist (90792, 90805, 90807, 90809, 90811, 90813, 90815, 90817, 90819, 90822, 90824, 90827, 90829, 90862). In cases where there was no such prescriber visit, we linked pharmacy claims to the most recent prescriber claim in the past two weeks irrespective of diagnosis. Provider claims included place of service and revenue variables which we used to characterize treatment setting. The four treatment settings examined were: psychosocial (B-PSY), outpatient hospital (B-OPH), office-based (B-OBOT), and no associated treatment setting (B-PHA). We excluded episodes if the primary treatment setting was an inpatient hospital, an outpatient psychiatric facility, or ambiguous. We also excluded B-PHA episodes for patients who never received an OUD diagnosis during the observation period as these patients may have been treated with buprenorphine off-label for non-OUD related conditions such as pain. We required individuals to have a minimum six months of Medicaid eligibility prior to an episode to establish baseline comorbidities for each episode.

Two groups (PSY-B and B-PSY) comprise the same combination of treatment services--buprenorphine in a psychosocial treatment program—but were constructed using distinct episode definitions. PSY-B was defined by continuous psychosocial treatment and B-PSY by continuous buprenorphine treatment. These two groups were never included in the same model and were used to examine different associations: PSY-B, the association of buprenorphine (vs. no buprenorphine) with psychosocial treatment retention and engagement, and B-PSY, the association of psychosocial (vs. medical) setting with buprenorphine treatment retention. We chose to conduct both types of analysis because each has distinct relevance for treatment providers and policy-makers seeking to understand the clinical ramifications of buprenorphine and psychosocial treatment integration.

Because patients who achieve successful induction onto XR naltrexone likely differ significantly in clinical characteristics from patients receiving maintenance buprenorphine treatment, we excluded episodes that involved XR naltrexone. We also excluded episodes that included services in an OTP since we were primarily interested in the characteristics of buprenorphine treatment in psychosocial and medical settings.

2.3.2. Outcomes/dependent variables

Treatment retention comprised the length of psychosocial or buprenorphine treatment prior to a 45-day gap in services or a 45-day gap in medication supply respectively. When patients lost eligibility prior to a 45-day gap in treatment, episodes were coded as “censored.” In the subset of psychosocial episodes lasting 30 days or more, we characterized treatment engagement as the number of hours of psychosocial treatment completed during the first 30 days. Current Procedural Terminology (CPT) codes for counseling (H0004), family counseling (T1006), group counseling (H0005), group education (H0025), behavioral health day treatment (H2012), and case management (T1016) were used to identify psychosocial services in delivered in psychosocial treatment programs. A service units variable was used to calculate the hours of treatment in each category. Office-based CPT codes were used to calculate the hours of individual (90804, 90806, 90808, 90832, 90834, and 90837), family (90847), and group (90853) therapy received outside psychosocial SUD programs. Psychosocial categories were summed to calculate total psychosocial treatment engagement. Our primary clinical outcomes were ED visits and hospitalizations. ED visits were identified by a dichotomous variable generated by Missouri’s Department of Social Services in the claims; hospitalizations were identified by a hospitalization revenue code.

2.3.3. Covariates

We included as covariates known correlates of SUD treatment retention available in the claims data. Race/ethnicity, sex, and age were derived from enrollment. Age was categorized as under 21, 21–24, 25–29, and 30+ years of age. Urbanicity was defined according to the National Center for Health Statistics Urban-Rural Classification Scheme for Counties; “small” and “medium” metropolitan levels were collapsed so that our five levels included: large central metropolitan areas, large fringe metropolitan areas, small/medium metropolitan areas, micropolitan areas, and non-core (rural) areas (Ingram & Franco, 2014). We also included diagnosed physical and psychiatric comorbidities. Physical comorbidities were defined by the Elixhauser index for the six-month period prior to the start of an episode with psychiatric and SUD diagnoses removed (Elixhauser, Steiner, Harris, & Coffey, 1998). We created separate indicators for the presence of any psychiatric disorder and any non-OUD SUD to identify a potentially independent association of mental health and SUD comorbidities with clinical outcomes.

2.4. Statistical procedures

All analyses were performed in SAS 9.4. We tested for demographic differences in our treatment groups using the Taylor series method to produce variance estimates adjusted for the clustering of episodes within individuals (Procs Surveylogistic and Surveyreg). Group comparisons were performed separately for psychosocial (PSY and PSY-B) and buprenorphine (B-PSY, B-OBOT, B-OPH, B-PHA) episodes. We ran a third set of group comparisons, including all treatment groups except PSY-B; we included these five non-overlapping groups in our final analyses. We also used the Taylor series method to calculate the differences in psychosocial treatment engagement by buprenorphine status in first 30 days of psychosocial treatment episodes.

Our two primary outcomes were negative clinical events: treatment attrition and an SUD-related ED visit or hospitalization. We performed survival analyses to calculate hazard ratios for each of these outcomes, controlling for the covariates and adjusting for within person correlation (Proc Phreg, covs option). We calculated the hazard ratio for psychosocial treatment attrition with versus without buprenorphine (PSY-B vs. PSY). For buprenorphine treatment attrition, we calculated the hazard ratios associated with each medical setting (B-OBOT, B-OPH, and B-PHA) versus the psychosocial setting (B-PSY). Using buprenorphine in psychosocial treatment (B-PSY) as the reference group, we calculated the hazards for an ED visit or hospitalization associated with psychosocial treatment without buprenorphine (PSY) and buprenorphine treatment in office-based (B-OBOT), outpatient hospital (B-OPH), and no associated (B-PHA) treatment settings (Table 3).

Table 3.

Hazard of discontinuing psychosocial treatment by buprenorphine status (n=8516a)

Hazard Ratio (95% CI)
Buprenorphineb (vs. PSY)
 B-PSY 0.67 (0.62–0.71)c
Sex (vs. male)
 female 0.91 (0.86–0.95)c
Race (vs. non-White)
 white 0.95 (0.89–1.01)
Age (vs. 30+)
 <21 0.96 (0.88–1.05)
 21–24 1.10 (1.02–1.19)c
 25–29 1.03 (0.97–1.09)
Urbanicity (vs. central metro)
 fringe metro 1.05 (0.98–1.13)
 small/med metro 0.91 (0.84–0.99)c
 micropolitan 1.00 (0.92–1.09)
 rural 0.92 (0.86–1.00)
Eligibility Type (vs. other)
 pregnancy 1.12 (1.02–1.22)c
Psych Comorbidity (vs. none)
 any 1.06 (1.01–1.12)c
SUD Comorbidity (vs. none)
 any 1.16 (1.10–1.21)c
Elixhauser 1.03 (1.01–1.04)c
Open in a new tab
a

17 observations excluded due to missing demographic characteristics

b

PSY, Psychosocial SUD treatment without buprenorphine; PSY-B, Psychosocial SUD treatment with buprenorphine

c

Statistically significant at conventional threshold (p<0.05).

3. Results

3.1. Characteristics of the Medicaid sample

Eligibility for Missouri Medicaid (MO HealthNet) is relatively restrictive for adults. Adults without dependent children are ineligible unless disabled. Adults with dependent children qualify if their household income does not exceed 18% of the federal poverty level, $20,780 for a family of three in 2018. Pregnant women and infants, and children ages one to 18, are eligible with household incomes up to 196% and 150% of the federal poverty level, respectively. Children are eligible for the Children’s Health Insurance Program if their household income does not exceed 300% of the federal poverty level. The demographics of our data reflect Missouri Medicaid’s eligibility criteria. Eligibility was determined by disability, parenting in extreme poverty, and pregnancy in 43%, 20%, and 8% of episodes, respectively. Although men are more likely than women to receive publicly funded SUD treatment, most Medicaid patients in our data set are women (SAMHSA, 2015). This dichotomy is due to women’s over-representation as caregivers of children in poverty and their eligibility to receive Medicaid benefits while pregnant; women made up 74% and 100% of those eligibility categories in our data respectively.

3.2. Characteristics of psychosocial and buprenorphine treatment groups (Tables 1 and 2)

Table 1.

Patient demographics, comorbidities, and service characteristics in psychosocial treatment episodes by buprenorphine statusa

PSY (n=7497) PSY-B (n=1019) TOTAL (n=8516)
By Sex % (SE) NS
 female 68.5 (0.7) 69.8 (1.7) 68.7 (0.7)
 male 31.5 (0.7) 30.2 (1.7) 31.3 (0.7)
By Race % (SE) <.01
 white 77.2 (0.7) 81.6 (1.5) 77.8 (0.7)
 non-white 22.8 (0.7) 18.4 (1.5) 22.2 (0.7)
By Age % (SE) <.0001
 <21 7.9 (0.4) 2.5 (0.5) 7.3 (0.3)
 21–24 11.5 (0.4) 8.9 (1.1) 11.2 (0.4)
 25–29 21.6 (0.6) 24.8 (1.6) 22.0 (0.6)
 30+ 59.0 (0.7) 63.8 (1.8) 59.6 (0.7)
By Urbanicity % (SE) <.0001, missing=17
 central metro 21.0 (0.6) 16.2 (1.3) 20.4 (0.6)
 fringe metro 27.0 (0.7) 21.7 (1.5) 26.4 (0.6)
 small/med metro 15.8 (0.5) 19.8 (1.5) 16.2 (0.5)
 micropolitan 14.2 (0.5) 17.4 (1.4) 14.6 (0.5)
 rural 22.0 (0.6) 24.9 (1.6) 22.4 (0.6)
By Eligibility Type % (SE) NS
 other 93.4 (0.3) 92.2 (0.9) 93.3 (0.3)
 pregnancy 6.6 (0.3) 7.8 (0.9) 6.7 (0.3)
By Psych Comorbidity % (SE) <.0001
 none 37.8 (0.7) 28.5 (1.5) 36.7 (0.6)
 any 62.2 (0.7) 71.5 (1.5) 63.3 (0.6)
By SUD Comorbidity % (SE) <.001
 none 68.3 (0.6) 62.5 (1.6) 67.6 (0.6)
 any 31.7 (0.6) 37.5 (1.6) 32.4 (0.6)
Elixhauser <.0001
 mean (SE) 1.01 (0.02) 1.09 (0.05) 1.02 (0.02)
Hours/week Psychosocial Treatment <.0001
 mean (SE) 21.2 (0.3) 17.3 (0.6) 20.7 (0.2)
Psychosocial Retention
 median days 23 51 26
Open in a new tab
a

PSY, psychosocial SUD treatment without buprenorphine; PSY-B, psychosocial SUD treatment with buprenorphine

Table 2.

Patient demographics, comorbidities, and service characteristics in buprenorphine treatment episodes by settinga

B-PSY (n=269) B-OBOT (n=1832) B-OPH (n=537) B-PHA (n=1542) TOTAL (n=4180) TOTALb (n=11,677)
By Sex % (SE) <.0001a,b
 female 57.1 (3.5) 69.8 (1.3) 75.7 (2.1) 64.7 (1.6) 67.9 (1.0) 68.3 (0.6)
 male 42.9 (3.5) 30.2 (1.3) 24.3 (2.1) 35.3 (1.6) 32.1 (1.0) 31.7 (0.6)
By Race % (SE) <.0001a,b
 white 53.5 (3.5) 92.8 (0.7) 92.3 (1.4) 88.9 (1.0) 88.7 (0.7) 81.3 (0.6)
 non-white 46.5 (3.5) 7.2 (0.7) 7.7 (1.4) 11.1 (1.0) 11.3 (0.7) 18.7 (0.6)
By Age % (SE) <.05a, <.0001b
 <21 1.5 (0.7) 2.0 (0.4) 2.9 (0.7) 2.3 (0.5) 2.2 (0.3) 5.9 (0.3)
 21–24 5.5 (1.5) 8.7 (0.7) 10.7 (1.6) 8.1 (0.8) 8.5 (0.5) 10.4 (0.4)
 25–29 16.8 (2.6) 23.9 (1.1) 23.0 (2.1) 21.1 (1.3) 22.3 (0.8) 21.9 (0.5)
 30+ 76.2 (2.9) 65.5 (1.3) 63.4 (2.4) 68.4 (1.5) 67.0 (1.0) 61.9 (0.6)
By Urbanicity % (SE) <.0001a,b
 central metro 47.3 (3.4) 7.7 (0.7) 8.6 (1.5) 13.8 (1.1) 12.6 (0.7) 18.0 (0.5)
 fringe metro 26.0 (3.0) 28.5 (1.2) 36.8 (2.4) 31.4 (1.5) 30.5 (1.0) 28.3 (0.6)
 small/med metro 11.0 (1.9) 17.2 (1.0) 35.8 (2.4) 14.9 (1.2) 18.3 (0.8) 16.7 (0.5)
 micropolitan 6.2 (1.5) 16.2 (1.0) 9.2 (1.4) 16.9 (1.3) 14.9 (0.8) 14.5 (0.5)
 rural 9.5 (2.0) 30.5 (1.3) 9.6 (1.3) 23.0 (1.4) 23.7 (0.9) 22.6 (0.5)
By Eligibility Type % (SE) <.0001a,b
 other 95.6 (1.4) 95.5 (0.5) 91.4 (1.3) 97.2 (0.4) 95.6 (0.3) 94.2 (0.3)
 pregnancy 4.4 (1.4) 4.5 (0.5) 8.6 (1.3) 2.8 (0.4) 4.4 (0.3) 5.8 (0.3)
By Psych Comorbidity % (SE) <.0001a,b
 none 44.0 (3.4) 22.2 (1.0) 29.4 (2.1) 41.9 (1.5) 31.8 (0.9) 35.6 (0.6)
 any 56.0 (3.4) 77.8 (1.0) 70.6 (2.1) 58.1 (1.5) 68.2 (0.9) 64.4 (0.6)
By SUD Comorbidity % (SE) <.0001a,b
 none 71.4 (2.8) 73.4 (1.1) 75.6 (1.9) 81.2 (1.1) 76.4 (0.7) 71.2 (0.5)
 any 28.6 (2.8) 26.6 (1.1) 24.4 (1.9) 18.8 (1.1) 23.6 (0.7) 28.8 (0.5)
Elixhauser <.0001a,b
 mean (SE) 1.14 (0.10) 1.12 (0.04) 1.08 (0.07) 0.77 (0.04) 0.98 (0.03) 1.00 (0.02)
Hours/week Psychosocial Treatment <.0001a,b
 mean (SE) 5.1 (0.4) 0.5 (0.1) 0.4 (0.0) 0.7 (0.1) 0.9 (0.0) 13.9 (0.2)
Buprenorphine Retention
 median days 22 126 59 30 58 59
Open in a new tab
a

B-PSY, buprenorphine treatment in a psychosocial program; B-OBOT, buprenorphine treatment in office-based setting; B-OPH, buprenorphine treatment in outpatient hospital; B-PHA, buprenorphine without billed treatment setting

b

Includes PSY group from Table 1 and corresponds to analyses reported in Table 5.

c

p-value for global comparison of buprenorphine groups.

d

p-value for global comparison of buprenorphine groups and PSY group in Table 1.

Table 1 describes demographic, comorbidity, and service characteristics of PSY (n=7,497) and B-PSY (n=1,019), episodes defined by continuous psychosocial SUD treatment. Non-white and younger patients were less likely than white and older patients to receive buprenorphine treatment in a psychosocial program. Patients who received buprenorphine during psychosocial treatment were more likely to have a psychiatric or SUD comorbidity, a higher Elixhauser index of physical comorbidity, and to live outside large central and fringe metro counties than patients who did not receive buprenorphine.

Table 2 describes the demographic, comorbidity, and service characteristics associated with B-PSY, B-OBOT, B-OPH, and B-PHA, episodes defined by continuous treatment with buprenorphine. B-PSY patients were more likely than other buprenorphine patients to be non-white, of central metro residence, pregnancy eligible for Medicaid, and to have a SUD comorbidity. B-OPH patients were more likely to be female and to live in fringe metro areas. B-OBOT patients were more likely to live in rural areas and to have a psychiatric comorbidity. B-PSY and B-OBOT patients had higher mean Elixhauser indices of physical comorbidity. White patients were more likely than non-white patients to receive buprenorphine treatment in any non-psychosocial medical setting (B-OBOT, B-OPH, or B-PHA).

The mean quantity of psychosocial treatment received in each treatment group lends support for our distinction between psychosocial and medical settings; while receiving treatment in medical settings (B-OPH, B-OBOT and B-PHA), patients received on average 0.9 hours of psychosocial treatment per week; patients in the PSY and PSY-B groups received an average of 21.2 and 17.3 hours of psychosocial treatment per week respectively.

3.3. The association of buprenorphine with psychosocial treatment attrition and engagement (Table 3)

After controlling for demographic and comorbidity characteristics, B-PSY patients had a 33% reduced hazard for treatment attrition compared to PSY patients. Additional risk factors for treatment attrition were male sex, ages 21 to 24, pregnancy eligibility for Medicaid, any SUD or psychiatric comorbidity, and a higher Elixhauser index of physical comorbidity. In the subset of patients retained in treatment more than 30 days, buprenorphine patients did not differ significantly (p = .054) from non-buprenorphine patients in the mean hours of psychosocial treatment they completed during the first 30 days after controlling for demographic and comorbidity characteristics. Younger age, non-white race, and the presence of a psychiatric or SUD comorbidity predicted increased hours of psychosocial treatment during the first 30 days. Central metro residence was associated with fewer hours of psychosocial treatment during the first 30 days.

3.4. The association of treatment setting with buprenorphine treatment attrition (Table 4)

Table 4.

Hazard of discontinuing buprenorphine treatment by setting (n=4180)

Hazard Ratio (95% CI)
Settinga (vs. B-PSY)
 B-OBOT 0.27 (0.24–0.31)b
 B-OPH 0.46 (0.39–0.54)b
 B-PHA 0.70 (0.61–0.81)b
Sex (vs. male)
 female 1.06 (0.98–1.14)
Race (vs. non-white)
 white 0.77 (0.69–0.85)b
Age (vs. 30+)
 <21 1.44 (1.14–1.82)b
 21–24 1.23 (1.09–1.38)b
 25–29 1.06 (0.98–1.15)
Urbanicity (vs. central metro)
 fringe metro 1.02 (0.92–1.14)
 small/med metro 0.65 (0.58–0.74)b
 micropolitan 0.83 (0.73–0.95)b
 rural 0.78 (0.69–0.88)b
Eligibility Type (vs. other)
 pregnancy 1.20 (1.03–1.41)b
Psych Comorbidity (vs. none)
 any 1.13 (1.05–1.22)b
SUD Comorbidity (vs. none)
 any 1.14 (1.05–1.24)b
Elixhauser 1.06 (1.03–1.08)b
Open in a new tab
a

B-PSY, buprenorphine treatment in psychosocial program; B-OBOT, buprenorphine treatment in office-based setting; B-OPH, buprenorphine treatment in outpatient hospital; B-PHA, buprenorphine without billed treatment setting

b

Statistically significant at conventional threshold (p<0.05).

Compared to B-PSY, B-OBOT, B-OPH, and B-PHA were associated with 73%, 54%, and 30% decreased hazards for buprenorphine treatment attrition, respectively. Non-white race, age less than 25, central or fringe metro residence, pregnancy eligibility for Medicaid, and SUD, psychiatric, and physical comorbidity were associated with increased hazards for buprenorphine treatment attrition.

3.5. The association of buprenorphine and treatment setting with patient hazards for ED visits and hospitalizations (Table 5)

Table 5.

Hazard for an SUD-related ED visit or hospitalization (n=11,677a)

Hazard Ratio (95% CI)
Treatment Typeb (vs. B-PSY)
 B-OBOT 0.59 (0.41–0.85)c
 B-OPH 0.77 (0.52–1.14)
 B-PHA 0.53 (0.36–0.78)c
 PSY 1.32 (0.92–1.88)
Sex (vs. male)
 female 0.92 (0.83–1.01)
Race (vs. non-white)
 white 0.91 (0.79–1.04)
Age (vs. 30+)
 <21 0.97 (0.77–1.22)
 21–24 1.53 (1.31–1.79)c
 25–29 1.50 (1.35–1.67)c
Urbanicity (vs. central metro)
 fringe metro 0.85 (0.74–0.98)c
 small/med metro 0.69 (0.59–0.81)c
 micropolitan 0.63 (0.53–0.76)c
 rural 0.63 (0.54–0.74)c
Eligibility Type (vs. other)
 pregnancy 0.97 (0.78–1.20)
Psych Comorbidity (vs. none)
 any 1.52 (1.36–1.70)c
SUD Comorbidity (vs. none)
 any 1.55 (1.41–1.71)c
Elixhauser 1.20 (1.17–1.23)c
Open in a new tab
a

14 observations excluded due to missing demographic characteristics

b

PSY, Psychosocial SUD treatment without medication; B-PSY, buprenorphine treatment in a psychosocial SUD program; B-OBOT, buprenorphine treatment in office-based setting; B-OPH, buprenorphine treatment in outpatient hospital; B-PHA, buprenorphine without billed treatment setting

c

Statistically significant at conventional threshold (p<0.05).

B-OBOT and B-PHA were associated with 41% and 47% decreased hazards for an ED visit or hospitalization after controlling for demographic and comorbidity characteristics. The hazard for an ED visit or hospitalization was not significantly different in B-PSY vs. B-OPH or PSY patients. Young adult age (21–29), central metro residence, and SUD, psychiatric, and physical comorbidity were associated with increased hazards for an ED visit or hospitalization.

4. Discussion

Our results suggest that for preventing treatment attrition and drug-related ED visits or hospitalizations, psychosocial treatment without buprenorphine is inferior to psychosocial treatment with buprenorphine, and, moreover, that buprenorphine treatment in psychosocial settings is inferior to buprenorphine treatment in office-based medical settings. Buprenorphine treatment attrition in psychosocial programs may result from more intensive psychosocial treatment requirements, implicit or explicit expectations that patients taper off buprenorphine, more stringent consequences for the continued use of opioid and non-opioid substances, and/or the influence of 12-step culture and its bias against agonist therapy (Aletraris et al., 2016; Gryczynski et al., 2014; Kornør, Waal, & Sandvik, 2007; Parran et al., 2010). Missouri’s Department of Mental Health should continue to evaluate buprenorphine treatment practices in psychosocial settings and work to ensure that Missouri’s publicly-funded treatment benefits from adherence to best practices such as high stable doses and low-barriers to buprenorphine treatment initiation and continuation (Faggiano, Vigna-Taglianti, Versino, & Lemma, 2003; Kourounis et al., 2016; Mattick et al., 2014). Consideration should also be given to funding OUD treatment in office-based settings such as federally qualified health centers, rural health clinics, and psychiatric practices. There may be positive aspects of buprenorphine treatment in these settings that do not easily generalize to psychosocial programs, such as the positive health behaviors associated with OUD treatment integration in primary care and chronic disease management (Haddad, Zelenev, & Altice, 2015; Korthuis et al., 2010). The inferior outcomes associated with buprenorphine treatment in psychosocial programs may result from group differences that we were unable to account for in our models. Compared to B-OBOT patients, B-PSY patients were less likely to have a diagnosed psychiatric comorbidity but more likely to be non-white, live in a central metro area, and have a higher index of physical comorbidity; the social determinants of health in these patient populations may differ in ways that impacted our measured outcomes. Although the psychosocial setting was associated with decreased buprenorphine retention, buprenorphine treatment was associated with increased patient retention during psychosocial treatment and did not appear to negatively impact psychosocial treatment engagement.

Our results revealed three important health disparities that should be the focus of public health policy. Consistent with past research, non-white patients were less likely than white patients to receive buprenorphine treatment in psychosocial treatment programs (Hansen, Siegel, Wanderling, & DiRocco, 2016); if non-white patients did receive buprenorphine, they were much less likely to receive it in the medical settings associated with the best clinical outcomes. Even after controlling for setting, there was a significant negative association of non-white race with buprenorphine but not psychosocial treatment retention. Efforts should be made ensure that non-white patients have equal access to high quality, evidence-based treatment.

Secondly, women with pregnancy eligibility for Medicaid had lower buprenorphine retention rates than patients with other types of eligibility. This result is unsurprising as pregnancy Medicaid in Missouri terminates 60 days after a newborn’s delivery. The temporary nature of pregnancy Medicaid naturally affects both provider and patient decisions about buprenorphine treatment. Missouri’s legislature passed a 2018 law extending pregnancy Medicaid for one year in mothers active in SUD treatment. If this law is implemented, it should positively impact buprenorphine treatment outcomes for postpartum women.

As the largest group of buprenorphine treatment episodes, the B-PHA group was of special interest. All the patients in the B-PHA group had an OUD diagnosis in their claims, but there were no billed prescriber visits associated with their buprenorphine therapy. There is a known shortage of buprenorphine prescribers due to the special training and certification required to obtain a buprenorphine waiver; as of 2015, only 16% of psychiatrists and 3% of primary care physicians had obtained the waiver (Rosenblatt, Andrilla, Catlin, & Larson, 2015). It is likely that the B-PHA group consists of buprenorphine episodes in which patients paid out-of-pocket for their prescriber visits and used Medicaid to pay for their medication. Compared to psychosocial treatment, B-PHA appeared to be protective against attrition and SUD-related ED visits or hospitalizations. However, compared to documented buprenorphine settings, B-PHA was associated with the highest hazard for treatment attrition. This finding supports the importance of insurance coverage for buprenorphine provider visits; the cost of out-of-pocket services is likely a significant deterrent to maintenance care even in patients who access buprenorphine treatment.

Finally, and consistent with past research, our results confirm that youth (below age 25) are less likely than older patients to receive buprenorphine (Hadland et al., 2017); they are also less likely to be retained in buprenorphine treatment. Many factors may contribute to this result, but it is likely that treatment providers are reluctant to start adolescents on a medication that causes physical dependence. To address this bias, the American Academy of Pediatrics has recommended increased use of OUD pharmacotherapy in teens with OUD and endorsed a buprenorphine waiver course for pediatricians (Committee on Substance Use and Prevention, 2016). As early SUD onset is positively associated with illness severity, withholding pharmacotherapy from adolescents and young adults with OUD is likely to increase their risk for OUD-related morbidity and death. This age-related disparity and the optimal clinical pathways for adolescents and young adults should be the focus of future research.

The Missouri Department of Mental Health acts as the Single State Agency administering block-grant and Medicaid funding for SUD treatment in Missouri. In 2016, department leaders determined that federal funding from the 2017–18 State Targeted Response to the Opioid Crisis grant should be used to increase access to pharmacotherapy by expanding capacity at existing OTPs and funding low-barrier buprenorphine treatment in traditionally psychosocial SUD programs (Knopf, 2018). This “Medication First” model has three key principles: OUD patients receive pharmacotherapy as quickly as possible; maintenance pharmacotherapy is delivered without arbitrary tapering or time limits; and individualized psychosocial services are offered but not required as a condition for pharmacotherapy. The Medication First initiative was implemented after the treatment analyzed in this research report.

Our results lend support for Missouri’s Medication First initiative. The model should increase the proportion of patients in psychosocial treatment who receive buprenorphine and methadone. The model should also improve retention and associated clinical outcomes by reducing barriers to maintenance treatment. Our results highlight the importance of funding buprenorphine treatment not only in psychosocial programs but also in office-based settings such as federally qualified health centers and rural health clinics. These community health centers are designed to provide chronic medical care, and their practitioners possess a core competency in pharmacotherapeutic disease management. Unlike Medicaid beneficiaries, uninsured patients in Missouri can generally access federal and state-funded treatment only in DMH-funded psychosocial treatment programs and geographically sparse OTPs; they do not generally have access to the expanding network of office-based buprenorphine providers that accept Medicaid. Expansion of the grant-funded OUD treatment network to office-based medical settings would give uninsured patients access to the growing network of office-based providers that serve Medicaid beneficiaries. This public policy priority is especially important in states like Missouri which have not expanded Medicaid.

Limitations

Our study had several limitations. Our Medicaid sample is not representative of Missouri’s treatment population since most SUD treatment in Missouri is funded by federal block grants and is not administered by MO Healthnet. Medicaid service data does not include measures of OUD severity, individual level social determinants, or detailed information about patient clinical outcomes. The data includes limited information from which to construct relevant covariates. Our linkage of buprenorphine treatment to prescriber setting was only approximate since pharmacy claims did not include prescriber information. We were also unable to compare the costs of various treatment modalities due to the high rates of managed Medicaid and the lack of reimbursement data. These limitations affect virtually all claims-based research and are offset to some extent by the high level of detail and accuracy in the patterns of service delivery. Studying treatment-as-usual, we were not able to identify the psychosocial practices or medical protocols that best support recovery. Nevertheless, a treatment-as-usual study is highly relevant to real-world patients, providers, and policymakers.

Acknowledgements:

We thank Mr. Glennon M. Floyd for editorial assistance while developing this manuscript. Ned Presnall is a paid consultant for the Missouri Department of Mental Health through the State Targeted Response to the Opioid Crisis and State Opioid Response grants and co-owner of CB Programs, LLC, a private SUD treatment program not funded by state or federal funds. The other authors have no financial relationships to disclose.

Funding: This work was supported by the Center for Health Economics and Policy, Washington University—St. Louis and the National Institute for Drug Abuse [R21 DA 044744-0].

Abbreviations

ASAM

American Society of Addiction Medicine

B-PHA

buprenorphine treatment with no billed treatment setting

B-OBOT

buprenorphine treatment in an office-based setting

B-OPH

buprenorphine treatment in an outpatient hospital setting

B-PSY

buprenorphine treatment in a psychosocial SUD program

CPT

Current Procedural Terminology

CSTAR

Comprehensive Substance Abuse Treatment and Rehabilitation program

DMH

Department of Mental Health

ED

emergency department

OTP

federally regulated opioid treatment program

OUD

opioid use disorder

PSY

psychosocial SUD treatment without buprenorphine

PSY-B

psychosocial SUD treatment with buprenorphine

SUD

substance use disorder

XR

extended-release

References

  1. Abraham AJ, Knudsen HK, Rieckmann T, & Roman PM (2013). Disparities in Access to Physicians and Medications for the Treatment of Substance Use Disorders Between Publicly and Privately Funded Treatment Programs in the United States. Journal of Studies on Alcohol and Drugs, 74(2), 258–265. 10.15288/jsad.2013.74.258 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ahmad F, Rossen L, Spencer M, Warner M, & Sutton P (2018). Provisional drug overdose death counts. Retrieved from https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm
  3. Aletraris L, Edmond MB, Paino M, Fields D, & Roman PM (2016). Counselor training and attitudes toward pharmacotherapies for opioid use disorder. Substance Abuse, 37(1), 47–53. 10.1080/08897077.2015.1062457 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Amato L, Minozzi S, Davoli M, & Vecchi S (2011). Psychosocial combined with agonist maintenance treatments versus agonist maintenance treatments alone for treatment of opioid dependence. Cochrane Database of Systematic Reviews, (10), CD004147. 10.1002/14651858.CD004147.pub4 [DOI] [Google Scholar]
  5. Bentzley BS, Barth KS, Back SE, & Book SW (2015). Discontinuation of Buprenorphine Maintenance Therapy: Perspectives and Outcomes. Journal of Substance Abuse Treatment, 52, 48–57. 10.1016/j.jsat.2014.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brooklyn JR, & Sigmon SC (2017). Vermont Hub-and-Spoke Model of Care for Opioid Use Disorder: Development, Implementation, and Impact. Journal of Addiction Medicine, 11(4), 286–292. 10.1097/ADM.0000000000000310 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chou R, Korthuis TP, Weimer M, Bougatsos C, Blazina I, Zakher B, … McCarty D (2016). Medication-Assisted Treatment Models of Care for Opioid Use Disorder in Primary Care Settings. Agency for Healthcare Research and Quality (AHRQ), (Publication No. 16(17)-EHC039-EF). 10.1021/acs.jafc.8b03125 [DOI] [PubMed] [Google Scholar]
  8. Clark RE, Baxter JD, Aweh G, O’Connell E, Fisher WH, & Barton BA (2015). Risk Factors for Relapse and Higher Costs Among Medicaid Members with Opioid Dependence or Abuse: Opioid Agonists, Comorbidities, and Treatment History. Journal of Substance Abuse Treatment, 57, 75–80. 10.1016/j.jsat.2015.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Committee on Substance Use and Prevention. (2016). Medication-Assisted Treatment of Adolescents With Opioid Use Disorders. PEDIATRICS, 138(3), e20161893–e20161893. 10.1542/peds.2016-1893 [DOI] [PubMed] [Google Scholar]
  10. Connery HS (2015). Medication-assisted treatment of opioid use disorder: review of the evidence and future directions. Harvard Review of Psychiatry, 23(2), 63–75. 10.1097/HRP.0000000000000075 [DOI] [PubMed] [Google Scholar]
  11. Elixhauser a, Steiner C, Harris DR, & Coffey RM (1998). Comorbidity measures for use with administrative data. Medical Care. 10.1097/00005650-199801000-00004 [DOI] [PubMed] [Google Scholar]
  12. Etheridge RM, Craddock SG, Dunteman GH, & Hubbard RL (1995). Treatment services in two national studies of community-based drug abuse treatment programs. Journal of Substance Abuse, 7(1), 9–26. 10.1016/0899-3289(95)90303-8 [DOI] [PubMed] [Google Scholar]
  13. Faggiano F, Vigna-Taglianti F, Versino E, & Lemma P (2003). Methadone maintenance at different dosages for opioid dependence. Cochrane Database of Systematic Reviews (Online), (3), CD002208 Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12917925 [DOI] [PubMed] [Google Scholar]
  14. Fiellin DA, Barry DT, Sullivan LE, Cutter CJ, Moore BA, O’Connor PG, & Schottenfeld RS (2013). A Randomized Trial of Cognitive Behavioral Therapy in Primary Care-based Buprenorphine. The American Journal of Medicine, 126(1), 74.e11–74.e17. 10.1016/j.amjmed.2012.07.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fiellin DA, Pantalon MV, Chawarski MC, Moore BA, Sullivan LE, O’Connor PG, & Schottenfeld RS (2006). Counseling plus Buprenorphine–Naloxone Maintenance Therapy for Opioid Dependence. New England Journal of Medicine, 355(4), 365–374. 10.1056/NEJMoa055255 [DOI] [PubMed] [Google Scholar]
  16. Friedmann PD, Jiang L, & Alexander JA (2010). Top manager effects on buprenorphine adoption in outpatient substance abuse treatment programs. The Journal of Behavioral Health Services & Research, 37(3), 322–337. 10.1007/s11414-009-9169-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gryczynski J, Mitchell SG, Jaffe JH, O’Grady KE, Olsen YK, & Schwartz RP (2014). Leaving Buprenorphine Treatment: Patients’ Reasons for Cessation of Care. Journal of Substance Abuse Treatment, 46(3), 356 10.1016/J.JSAT.2013.10.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Haddad MS, Zelenev A, & Altice FL (2015). Buprenorphine Maintenance Treatment Retention Improves Nationally Recommended Preventive Primary Care Screenings when Integrated into Urban Federally Qualified Health Centers. Journal of Urban Health : Bulletin of the New York Academy of Medicine, 92(1), 193 10.1007/S11524-014-9924-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hadland SE, Wharam JF, Schuster MA, Zhang F, Samet JH, & Larochelle MR (2017). Trends in Receipt of Buprenorphine and Naltrexone for Opioid Use Disorder Among Adolescents and Young Adults, 2001–2014. JAMA Pediatrics, 171(8), 747–755. 10.1001/jamapediatrics.2017.0745 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hansen H, Siegel C, Wanderling J, & DiRocco D (2016). Buprenorphine and methadone treatment for opioid dependence by income, ethnicity and race of neighborhoods in New York City. Drug and Alcohol Dependence. 10.1016/j.drugalcdep.2016.03.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hubbard RL, Craddock SG, & Anderson J (2003). Overview of 5-year followup outcomes in the drug abuse treatment outcome studies (DATOS). Journal of Substance Abuse Treatment, 25(3), 125–134. 10.1016/S0740-5472(03)00130-2 [DOI] [PubMed] [Google Scholar]
  22. Ingram DD, & Franco SJ (2014). 2013 NCHS Urban-Rural Classification Scheme for Counties. Vital and Health Statistics. Series 2, Data Evaluation and Methods Research, (166), 1–73. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/24776070 [PubMed] [Google Scholar]
  23. Knopf A (2018). Missouri dramatically increases buprenorphine/methadone with STR. Alcoholism & Drug Abuse Weekly, 30(24), 1–3. 10.1002/adaw.31999 [DOI] [Google Scholar]
  24. Komaromy M, Duhigg D, Metcalf A, Carlson C, Kalishman S, Hayes L, … Arora S (2016). Project ECHO (Extension for Community Healthcare Outcomes): A new model for educating primary care providers about treatment of substance use disorders. Substance Abuse, 37(1), 20–24. 10.1080/08897077.2015.1129388 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kornør H, Waal H, & Sandvik L (2007). Time-limited buprenorphine replacement therapy for opioid dependence: 2-year follow-up outcomes in relation to programme completion and current agonist therapy status. Drug and Alcohol Review, 26(2), 135–141. 10.1080/09595230601146603 [DOI] [PubMed] [Google Scholar]
  26. Korthuis PT, Gregg J, Rogers WE, McCarty D, Nicolaidis C, & Boverman J (2010). Patients’ Reasons for Choosing Office-based Buprenorphine: Preference for Patient-Centered Care. Journal of Addiction Medicine, 4(4), 204–210. 10.1097/ADM.0b013e3181cc9610 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Korthuis PT, McCarty D, Weimer M, Bougatsos C, Blazina I, Zakher B, … Chou R (2017). Primary Care-Based Models for the Treatment of Opioid Use Disorder: A Scoping Review. Annals of Internal Medicine, 166(4), 268–278. 10.7326/M16-2149 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kourounis G, Richards BDW, Kyprianou E, Symeonidou E, Malliori M-M, & Samartzis L (2016). Opioid substitution therapy: Lowering the treatment thresholds. Drug and Alcohol Dependence, 161, 1–8. 10.1016/j.drugalcdep.2015.12.021 [DOI] [PubMed] [Google Scholar]
  29. LaBelle CT, Han SC, Bergeron A, & Samet JH (2016). Office-Based Opioid Treatment with Buprenorphine (OBOT-B): Statewide Implementation of the Massachusetts Collaborative Care Model in Community Health Centers. Journal of Substance Abuse Treatment, 60, 6–13. 10.1016/j.jsat.2015.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Marsch LA (1998). The efficacy of methadone maintenance interventions in reducing illicit opiate use, HIV risk behavior and criminality: a meta-analysis. Addiction, 93(4), 515–532. [DOI] [PubMed] [Google Scholar]
  31. Mattick RP, Breen C, Kimber J, & Davoli M (2014). Buprenorphine maintenance versus placebo or methadone maintenance for opioid dependence. Cochrane Database of Systematic Reviews, (2), CD002207. 10.1002/14651858.CD002207.pub4 [DOI] [Google Scholar]
  32. Mee-Lee D, Shulman GD, Fishman MJ, Gastfriend DR, & M. MM (2013). The ASAM criteria: Treatment criteria for addictive, substance-related, and co-occurring conditions. The Change Companies. [Google Scholar]
  33. Mitchell SG, Gryczynski J, Schwartz RP, O’Grady KE, Olsen YK, & Jaffe JH (2013). A randomized trial of intensive outpatient (IOP) vs. standard outpatient (OP) buprenorphine treatment for African Americans. Drug and Alcohol Dependence, 128(3), 222–229. 10.1016/j.drugalcdep.2012.08.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Parran TV, Adelman CA, Merkin B, Pagano ME, Defranco R, Ionescu RA, & Mace AG (2010). Long-term outcomes of office-based buprenorphine/naloxone maintenance therapy. Drug and Alcohol Dependence, 106(1), 56–60. 10.1016/j.drugalcdep.2009.07.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Pierce M, Bird SM, Hickman M, Marsden J, Dunn G, Jones A, & Millar T (2016). Impact of treatment for opioid dependence on fatal drug-related poisoning: a national cohort study in England. Addiction, 111(2), 298–308. 10.1111/add.13193 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Rosenblatt RA, Andrilla CHA, Catlin M, & Larson EH (2015). Geographic and specialty distribution of US physicians trained to treat opioid use disorder. Annals of Family Medicine, 13(1), 23–26. 10.1370/afm.1735 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. SAMHSA. (2010). Treatment Episodes Data Set (TEDS). 1998–2008 National Admissions to Treatment Services. Rockville, MD. [Google Scholar]
  38. SAMHSA. (2015). Treatment episodes data set admissions (TEDS-A). Rockville, MD. [Google Scholar]
  39. Seppala MD, & Larson B (2015). Integrating the Twelve Steps with Medication-Assisted Treatment for Opioid Use Disorder. Center City: Hazelden. [Google Scholar]
  40. Seth P, Scholl L, Rudd RA, & Bacon S (2018). Overdose Deaths Involving Opioids, Cocaine, and Psychostimulants — United States, 2015–2016. MMWR. Morbidity and Mortality Weekly Report, 67(12), 349–358. 10.15585/mmwr.mm6712a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sordo L, Barrio G, Bravo MJ, Indave BI, Degenhardt L, Wiessing L, … Pastor-Barriuso R (2017). Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ, j1550. 10.1136/bmj.j1550 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Volkow ND, Frieden TR, Hyde PS, & Cha SS (2014). Medication-assisted therapies--tackling the opioid-overdose epidemic. The New England Journal of Medicine, 370(22), 2063–2066. 10.1056/NEJMp1402780 [DOI] [PubMed] [Google Scholar]
  43. Weiss RD, Potter JS, Fiellin DA, Byrne M, Connery HS, Dickinson W, … Ling W (2011). Adjunctive counseling during brief and extended buprenorphine-naloxone treatment for prescription opioid dependence: a 2-phase randomized controlled trial. Archives of General Psychiatry, 68(12), 1238–1246. 10.1001/archgenpsychiatry.2011.121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. World Services Board of Trustees. (1997). Bulletin #29: Regarding methadone and other drug replacement programs. Retrieved from https://na.org/?ID=bulletins-bull29

RESOURCES