Clearing the Fog: A Scoping Literature Review on the Ethical Issues Surrounding Artificial Intelligence-Based Medical Devices.

The use of AI in healthcare has sparked much debate among philosophers, ethicists, regulators and policymakers who raised concerns about the implications of such technologies. The presented scoping review captures the progression of the ethical and legal debate and the proposed ethical frameworks available concerning the use of AI-based medical technologies, capturing key themes across a wide range of medical contexts. The ethical dimensions are synthesised in order to produce a coherent ethical framework for AI-based medical technologies, highlighting how transparency, accountability, confidentiality, autonomy, trust and fairness are the top six recurrent ethical issues. The literature also highlighted how it is essential to increase ethical awareness through interdisciplinary research, such that researchers, AI developers and regulators have the necessary education/competence or networks and tools to ensure proper consideration of ethical matters in the conception and design of new AI technologies and their norms. Interdisciplinarity throughout research, regulation and implementation will help ensure AI-based medical devices are ethical, clinically effective and safe. Achieving these goals will facilitate successful translation of AI into healthcare systems, which currently is lagging behind other sectors, to ensure timely achievement of health benefits to patients and the public.

Treatment setting and buprenorphine discontinuation: an analysis of multi-state insurance claims.

BACKGROUND: Potential differences in buprenorphine treatment outcomes across various treatment settings are poorly characterized in multi-state administrative data. We thus evaluated the association of opioid use disorder (OUD) treatment setting and insurance type with risk of buprenorphine discontinuation among commercial insurance and Medicaid enrollees initiated on buprenorphine. METHODS: In this observational, retrospective cohort study using the Merative MarketScan databases (2006-2016), we analyzed buprenorphine retention in 58,200 US adults with OUD. Predictor variables included insurance status (Medicaid vs commercial) and treatment setting, operationalized as substance use disorder (SUD) specialty treatment facility versus outpatient primary care physicians (PCPs) versus outpatient psychiatry, ascertained by linking physician visit codes to buprenorphine prescriptions. Treatment setting was inferred based on timing of prescriber visit claims preceding prescription fills. We estimated time to buprenorphine discontinuation using multivariable cox regression. RESULTS: Among enrollees with OUD receiving buprenorphine, 26,168 (45.0%) had prescriptions from SUD facilities without outpatient buprenorphine treatment, with the remaining treated by outpatient PCPs (n = 23,899, 41.1%) and psychiatrists (n = 8133, 13.9%). Overall, 50.6% and 73.3% discontinued treatment at 180 and 365 days respectively. Buprenorphine discontinuation was higher among enrollees receiving prescriptions from SUD facilities (aHR = 1.03[1.01-1.06]) and PCPs (aHR = 1.07[1.05-1.10]). Medicaid enrollees had lower buprenorphine retention than those with commercial insurance, particularly those receiving buprenorphine from SUD facilities and PCPs (aHR = 1.24[1.20-1.29] and aHR = 1.39[1.34-1.45] respectively, relative to comparator group of commercial insurance enrollees receiving buprenorphine from outpatient psychiatry). CONCLUSION: Buprenorphine discontinuation is high across outpatient PCP, psychiatry, and SUD treatment facility settings, with potentially lower treatment retention among Medicaid enrollees receiving care from SUD facilities and PCPs.

Use of in-network insurance benefits is critical for improving retention in telehealth-based buprenorphine treatment.

An empiric evidence base is lacking regarding the relationship between insurance status, payment source, and outcomes among patients with opioid use disorder (OUD) on telehealth platforms. Such information gaps may lead to unintended impacts of policy changes. Following the phase-out of the COVID-19 Public Health Emergency, states were allowed to redetermine Medicaid eligibility and disenroll individuals. Yet, financial barriers remain a common and significant hurdle for patients with OUD and are associated with worse outcomes. We studied 3842 patients entering care in 2022 at Ophelia Health, one of the nation's largest OUD telehealth companies, to assess associations between insurance status and 6-month retention. In multivariable analyses, in-network patients who could use insurance benefits were more likely to be retained compared with cash-pay patients (adjusted risk ratio [aRR]: 1.50; 95% CI: 1.40-1.62; P < .001). Among a subsample of 882 patients for whom more detailed insurance data were available (due to phased-in electronic health record updates), in-network patients were also more likely to be retained at 6 months compared with insured, yet out-of-network patients (aRR: 1.86; 95% CI: 1.54-2.23; P < .001). Findings show that insurance status, and specifically the use of in-network benefits, is associated with superior retention and suggest that Medicaid disenrollment and insurance plan hesitation to engage with telehealth providers may undermine the nation's response to the opioid crisis.

Application of an opioid use disorder cascade of care in a large public health system.

Background: Over the past decade, hospitals and health systems have increasingly adopted interventions to address the needs of patients with substance use disorders. The Opioid Use Disorder (OUD) Cascade of Care provides a framework for organizing and tracking patient health milestones over time and can assist health systems in identifying areas of intervention to maximize the impact of evidence-based services. However, detailed protocols are needed to guide health systems in how to operationalize the OUD Cascade and track outcomes using electronic health records.Objective: In this paper, we describe the process of operationalizing and applying the OUD Cascade in a large, urban, public hospital system.Methods: Through this case example, we describe the technical processes around data mining, as well as the decision-making processes, challenges encountered, lessons learned from compiling preliminary patient data and defining stages and outcome measures for the OUD Cascade of Care, and preliminary dataResults: We identified 33,616 (26.17% female) individuals with an OUD diagnosis. Almost half (48%) engaged with addiction services, while only 10.7% initiated medication-based treatment in an outpatient setting, 6.7% had timely follow-up, and 3.5% were retained for a minimum of 6 months.Conclusion: The current paper serves as a primer for other health systems seeking to implement data-informed approaches to guide more efficient care and improved substance use-related outcomes. An OUD Cascade of Care must be tailored to local systems based on inherent data limitations and services design with an emphasis on early stages wherein drop-off is the greatest.

Referral to and engagement in substance use disorder treatment within opioid intervention courts in New York: a qualitative study of implementation barriers and facilitators.

BACKGROUND: People with opioid use disorder (OUD) are frequently in contact with the court system and have markedly higher rates of fatal opioid overdose. Opioid intervention courts (OIC) were developed to address increasing rates of opioid overdose among court defendants by engaging court staff in identification of treatment need and referral for opioid-related services and building collaborations between the court and OUD treatment systems. The study goal was to understand implementation barriers and facilitators in referring and engaging OIC clients in OUD treatment. METHODS: Semi-structured interviews were conducted with OIC stakeholders (n = 46) in 10 New York counties in the United States, including court coordinators, court case managers, and substance use disorder treatment clinic counselors, administrators, and peers. Interviews were recorded and transcribed and thematic analysis was conducted, guided by the Exploration, Preparation, Implementation, Sustainment (EPIS) framework, employing both inductive and deductive coding. RESULTS: Results were conceptualized using EPIS inner (i.e., courts) and outer (i.e., OUD treatment providers) implementation contexts and bridging factors that impacted referral and engagement to OUD treatment from the OIC. Inner factors that facilitated OIC implementation included OIC philosophy (e.g., non-punitive, access-oriented), court organizational structure (e.g., strong court staff connectedness), and OIC court staff and client characteristics (e.g., positive medications for OUD [MOUD] attitudes). The latter two also served as barriers (e.g., lack of formalized procedures; stigma toward MOUD). Two outer context entities impacted OIC implementation as both barriers and facilitators: substance use disorder treatment programs (e.g., attitudes toward the OIC and MOUD; operational characteristics) and community environments (e.g., attitudes toward the opioid epidemic). The COVID-19 pandemic and bail reform were macro-outer context factors that negatively impacted OIC implementation. Facilitating bridging factors included staffing practices that bridged court and treatment systems (e.g., peers); barriers included communication and cultural differences between systems (e.g., differing expectations about OIC client success). CONCLUSIONS: This study identified key barriers and facilitators that OICs may consider as this model expands in the United States. Referral to and engagement in OUD treatment within the OIC context requires ongoing efforts to bridge the treatment and court systems, and reduce stigma around MOUD.

Are you thinking what I'm thinking? Defining what we mean by "polysubstance use."

The rise in drug overdoses and harms associated with the use of more than one substance has led to increased use of the term "polysubstance use" among researchers, clinicians, and public health officials. However, the term retains no consistent definition across contexts. The current authors convened from disciplines including sociology, epidemiology, neuroscience, and addiction psychiatry to propose a recommended definition of polysubstance use. An iterative process considered authors' formal and informal conversations, insights from relevant symposia, talks, and conferences, as well as their own research and clinical experiences to propose the current definition. Three key concepts were identified as necessary to define polysubstance use: (1) substances involved, (2) timing, and (3) intent. Substances involved include clarifying either (1) the number and type of substances used, (2) presence of more than one substance use disorder, or (3) primary and secondary substance use. The concept of timing is recommended to use clear terms such as simultaneous, sequential, and same-day polysubstance use to describe short-term behaviors (e.g., 30-day windows). Finally, the concept of intent refers to clarifying unintentional use or exposure when possible, and greater attention to motivations of polysubstance use. These three components should be clearly defined in research on polysubstance use to improve consistency across disciplines. Consistent definitions of polysubstance use can aid in the synthesis of evidence to better address an overdose crisis that increasingly involves multiple substances.

Association of medications for opioid use disorder with reduced risk of repeat opioid overdose in Medicaid: A cohort study.

INTRODUCTION: Following a nonfatal opioid overdose, patients are at high risk for repeat overdose. The objective of this study was to examine the association of MOUD after nonfatal opioid overdose with risk of repeat overdose in the following year. METHODS: This retrospective cohort study analyzed Missouri Medicaid claims from July 2012 to December 2021. The study identified opioid overdoses occurring between 2013 and 2020 using diagnosis codes for opioid poisoning in an inpatient or emergency department setting. The study implemented Cox models with a time-varying covariate for post-overdose receipt of MOUD. RESULTS: During the study period, MOUD receipt after overdose more than tripled, from 4.8 % to 18.9 %. Overall, only 12.1 % of patients received MOUD in the year after index. MOUD during follow-up was associated with significantly lower risk of repeat overdose (HR = 0.34, 95 % CI = 0.14-0.82). Out of 3017 individuals meeting inclusion criteria, 13.6 % had a repeat opioid overdose within 1 year. Repeat overdose risk was higher for those whose index overdose involved heroin or synthetic opioids (HR = 1.71, 95 % CI = 1.35-2.15), but MOUD was associated with significantly reduced risk in this group (HR = 0.34, 95 % CI = 0.13-0.92). CONCLUSIONS: MOUD receipt was associated with reduced risk of repeat overdose. Those whose index overdoses involved heroin or synthetic opioids were at greater risk of repeat overdose, but MOUD was associated with reduced risk in this group.

Retention and critical outcomes among new methadone maintenance patients following extended take-home reforms: a retrospective observational cohort study.

Background: Approximately 1800 opioid treatment programs (OTPs) in the US dispense methadone to upwards of 400,000 patients with opioid use disorder (OUD) annually, operating under longstanding highly restrictive guidelines. OTPs were granted novel flexibilities beginning March 15, 2020, allowing for reduced visit frequency and extended take-home doses to minimize COVID exposure with great variation across states and sites. We sought to use electronic health records to compare retention in treatment, opioid use, and adverse events among patients newly entering methadone maintenance in the post-reform period in comparison with year-ago, unexposed, controls. Methods: Retrospective observational cohort study across 9 OTPs, geographically dispersed, in the National Institute of Drug Abuse (NIDA) Clinical Trials Network. Newly enrolled patients between April 15 and October 14, 2020 (post-COVID, reform period) v. March 15-September 14, 2019 (pre-COVID, control period) were assessed. The primary outcome was 6-month retention. Secondary outcomes were opioid use and adverse events including emergency department visits, hospitalizations, and overdose. Findings: 821 individuals were newly admitted in the post-COVID and year-ago control periods, average age of 38.3 (SD 11.1), 58.9% male. The only difference across pre- and post-reform groups was the prevalence of psychostimulant use disorder (25.7% vs 32.9%, p = 0.02). Retention was non-inferior (60.0% vs 60.1%) as were hazards of adverse events in the aggregate (X2 (1) = 0.55, p = 0.46) in the post-COVID period. However, rates of month-level opioid use were higher among post-COVID intakes compared to pre-COVID controls (64.8% vs 51.1%, p < 0.001). Moderator analyses accounting for stimulant use and site-level variation in take-home schedules did not change findings. Interpretation: Policies allowing for extended take-home schedules were not associated with worse retention or adverse events despite slightly elevated rates of measured opioid use while in care. Relaxed guidelines were not associated with measurable increased harms and findings could inform future studies with prospective trials. Funding: USDHHSNIDACTNUG1DA013035-15.

Co-occurring psychiatric disorders and disparities in buprenorphine utilization in opioid use disorder: An analysis of insurance claims.

Background: As the overdose crisis continues in the U.S. and Canada, opioid use disorder (OUD) treatment outcomes for people with co-occurring psychiatric disorders are not well characterized. Our objective was to examine the influence of co-occurring psychiatric disorders on buprenorphine initiation and discontinuation. Methods: This retrospective cohort study used multi-state administrative claims data in the U.S. to evaluate rates of buprenorphine initiation (relative to psychosocial treatment without medication) in a cohort of 236,198 people with OUD entering treatment, both with and without co-occurring psychiatric disorders, grouping by psychiatric disorder subtype (mood, psychotic, and anxiety-and-related disorders). Among people initiating buprenorphine, we assessed the influence of co-occurring psychiatric disorders on buprenorphine retention. We used multivariable Poisson regression to estimate buprenorphine initiation and Cox regression to estimate time to discontinuation, adjusting for all 3 classes of co-occurring disorders simultaneously and adjusting for baseline demographic and clinical characteristics. Results: Buprenorphine initiation occurred in 29.3 % of those with co-occurring anxiety-and-related disorders, compared to 25.9 % and 17.5 % in people with mood and psychotic disorders. Mood (adjusted-risk-ratio[aRR] = 0.82[95 % CI = 0.82-0.83]) and psychotic disorders (aRR = 0.95[0.94-0.96]) were associated with decreased initiation (versus psychosocial treatment), in contrast to greater initiation in the anxiety disorders cohort (aRR = 1.06[1.05-1.06]). We observed an increase in buprenorphine discontinuation associated with mood (adjusted-hazard-ratio[aHR] = 1.20[1.17-1.24]) and anxiety disorders (aHR = 1.12[1.09-1.14]), in contrast to no association between psychotic disorders and buprenorphine discontinuation. Conclusions: We observed underutilization of buprenorphine among people with co-occurring mood and psychotic disorders, as well as high buprenorphine discontinuation across anxiety, mood, and psychotic disorders.

Application of an Opioid Use Disorder Cascade of Care in a Large Public Health System.

Background: Over the past decade, hospitals and health systems have increasingly adopted interventions to address the needs of patients with substance use disorders. The Opioid Use Disorder (OUD) Cascade of Care provides a framework for organizing and tracking patient health milestones over time, and can assist health systems in identifying areas of intervention to prevent overdose and maximize the impact of evidence-based services for patients with OUD. However, detailed protocols are needed to guide health systems in how to operationalize the OUD Cascade and track outcomes using their systems' electronic medical records (EMR). Objective: In this paper, we describe the process of operationalizing and implementing the OUD Cascade in one large, urban, public hospital system. Methods: Through this case example, we describe the technical processes around data mining, as well as the decision-making processes, challenges encountered, and lessons learned from compiling patient data and defining stages and outcome measures for the OUD Cascade of Care. The current established framework and process will set the stage for subsequent research studies that quantify and evaluate patient progression through each stage of OUD treatment across the health system and identify target areas for quality improvement initiatives to better engage patients in care and improve health outcomes. Results: The current paper can therefore serve as a primer for other health systems seeking to implement a data-informed approach to guide more efficient care and improved substance use-related outcomes. Conclusion: An OUD Cascade of Care must be tailored to local systems based on inherent data limitations and services design.

Urine Drug Screening in a Telehealth Setting for the Treatment of Opioid Use Disorder.

Importance: Amid rapid and widespread adoption of telehealth-based opioid treatment (TBOT), there is an urgent need for rigorous studies exploring the feasibility and characteristics of urine drug screening (UDS). Objective: To investigate administration patterns and results of UDS to assess feasibility of UDS and patient outcomes in a TBOT setting. Design: This observational cohort study was conducted between January 1, 2021, and December 6, 2022, and included patients with opioid use disorder treated in Ophelia, a TBOT treatment platform in 14 states. Data analysis was performed from January to March 2023. Main Outcomes and Measures: Number and percentage of patients with UDS within 30, 90, and 180 days of intake, grouped by adherence to clinical protocols. Associations were assessed between baseline characteristics and UDS completion and opioid positivity in first 30 days using χ2 tests. Baseline and 180-day follow-up UDS results were compared using McNemar tests. Results: Among 3395 patients (mean [SD] age, 38.2 [9.3] years, mostly male [54.1%], non-Hispanic White [81.5%], urban-residing [80.3%], and cash-pay at intake [74.0%]), 2782 (83.3%) completed a UDS within 30 days (90.0% among protocol-adherent patients, 67.0% among protocol-nonadherent patients). A total of 2750 of 2817 (97.6%) patients retained more than 90 days completed 1 or more UDS, as did 2307 of 2314 (99.7%) patients retained more than 180 days. Younger patients, patients of a racial and ethnic minority group, those living in urban areas, and cash-pay patients were less likely to complete a UDS in the first 30 days. Buprenorphine positivity increased (from 96.9% to 98.4%, P = .004) and opioid positivity declined (from 7.9% to 3.3%, P < .001) over time. Conclusions and Relevance: In this cohort study of patients with opioid use disorder receiving buprenorphine in a remote care environment, UDS was highly feasible, though early UDS completion rates varied across demographic subgroups. The prevalence of unexpected UDS results was low and declined over time in treatment.

Buprenorphine After Nonfatal Opioid Overdose: Reduced Mortality Risk in Medicare Disability Beneficiaries.

INTRODUCTION: Opioid-involved overdose mortality is a persistent public health challenge, yet limited evidence exists on the relationship between opioid use disorder treatment after a nonfatal overdose and subsequent overdose death. METHODS: National Medicare data were used to identify adult (aged 18-64 years) disability beneficiaries who received inpatient or emergency treatment for nonfatal opioid-involved overdose in 2008-2016. Opioid use disorder treatment was defined as (1) buprenorphine, measured using medication days' supply, and (2) psychosocial services, measured as 30-day exposures from and including each service date. Opioid-involved overdose fatalities were identified in the year after nonfatal overdose using linked National Death Index data. Cox proportional hazards models estimated the associations between time-varying treatment exposures and overdose death. Analyses were conducted in 2022. RESULTS: The sample (N=81,616) was mostly female (57.3%), aged ≥50 years (58.8%), and White (80.9%), with a significantly elevated overdose mortality rate, compared with the general U.S. population (standardized mortality ratio=132.4, 95% CI=129.9, 135.0). Only 6.5% of the sample (n=5,329) had opioid use disorder treatment after the index overdose. Buprenorphine (n=3,774, 4.6%) was associated with a significantly lower risk of opioid-involved overdose death (adjusted hazard ratio=0.38, 95% CI=0.23, 0.64), but opioid use disorder-related psychosocial treatment (n=2,405, 2.9%) was not associated with risk of death (adjusted hazard ratio=1.18, 95% CI=0.71, 1.95). CONCLUSIONS: Buprenorphine treatment after nonfatal opioid-involved overdose was associated with a 62% reduction in the risk of opioid-involved overdose death. However, fewer than 1 in 20 individuals received buprenorphine in the subsequent year, highlighting a need to strengthen care connections after critical opioid-related events, particularly for vulnerable groups.

Telehealth for opioid use disorder: retention as a function of demographics and rurality.

Background: Despite lifesaving medications such as buprenorphine and methadone, the majority of individuals with opioid use disorder (OUD) face access barriers to evidence-based treatment. COVID-19 era regulatory reforms have shown that telehealth can improve access to care, although disparities in clinical outcomes are likely to persist.Objective: We aimed to analyze 180-day and 365-day retention in treatment with buprenorphine for OUD overall and by demographics, hypothesizing that retention would be lower among racial/ethnic minorities and rural patients.Methods: We analyzed data from a cohort of individuals with OUD enrolled in treatment from April 1, 2020 to September 30, 2021, in Pennsylvania and New York using a virtual-first telehealth OUD treatment platform to assess rates of 180-day and 365-day retention. Associations between demographic characteristics and retention were assessed using unadjusted and adjusted logistic regression models.Results: Among 1,378 patients (58.8% male), 180-day retention was 56.4%, and 365-day retention was 48.3%. Adjusted analyses found that only an association between older age and greater odds of 180-day retention was significant (aOR for patients aged 30-50 vs. <30: 1.83 [1.37-2.45]). There were no significant associations between sex, race/ethnicity, state, or rurality with retention.Conclusion: While we were unable to control for socioeconomic variables, we found retention within telehealth services for buprenorphine was high irrespective of geography or race/ethnicity, but disparities with age indicate a subset of patients who may benefit from more intensive services early in care.

Adult Medical Cannabinoid Use and Changes in Prescription Controlled Substance Use.

Introduction: Nonopioid-based strategies for managing chronic noncancer pain are needed to help reduce overdose deaths. Although lab studies and population-level data suggest that cannabinoids could provide opioid-sparing effects, among medical cannabis participants they may also impact overdose risk by modifying other controlled substance use such as sedative hypnotics. However, no study has combined observational data at the individual level to empirically address interactions between the use of cannabinoids and prescribed controlled substances. Methods: Electronic health records, including prescription drug monitoring program data, from a large multisite medical cannabis program in New York State were abstracted for all participants with noncancer pain and recently prescribed noncannabinoid controlled substances who completed a new intake visit from April 15, 2018-April 14, 2019 and who remained actively in treatment for >180 days. Participants were partitioned into two samples: those with recent opioid use and those with active opioid use and co-use of sedative hypnotics. A patient-month level analysis assessed total average equivalent milligrams by class of drug (i.e., cannabinoid distinguishing tetrahydrocannabinol [THC] vs. cannabidiol [CBD], opioids, and sedative-hypnotics) received as a time-varying outcome measure across each 30-day "month" period postintake for at least 6 months for all participants. Results: Sample 1 of 285 opioid users were 61.1 years of age (±13.5), 57.5% female, and using an average of 49.7 (±98.5) morphine equivalents daily at intake. Unadjusted analyses found a modest decline in morphine equivalents to 43.9 mg (±94.1 mg) from 49.7 (±98.5) in month 1 (p=0.047) while receiving relatively low doses of THC (2.93 mg/day) and CBD (2.15 mg/day). Sample 2 of 95 opioid and sedative-hypnotic users were 60.9 years of age (±13.1), 63.2% female, and using an average of 86.6 (±136.2) morphine equivalents daily, and an average of 4.3 (±5.6) lorazepam equivalents. Unadjusted analyses did not find significant changes in either morphine equivalents (p=0.81) or lorazepam equivalents (p=0.980), and patients similarly received relatively low doses of THC (2.32 mg/day) and CBD (2.24 mg/day). Conclusions: Findings demonstrated minimal to no change in either opioids or sedative hypnotics over the 6 months of medical cannabis use but may be limited by low retention rates, external generalizability, and an inability to account for nonprescribed substance use.

Opioid use disorder Cascade of care framework design: A roadmap.

Unintentional overdose deaths, most involving opioids, have eclipsed all other causes of US deaths for individuals less than 50 years of age. An estimated 2.4 to 5 million individuals have opioid use disorder (OUD) yet a minority receive treatment in a given year. Medications for OUD (MOUD) are the gold standard treatment for OUD however early dropout remains a major challenge for improving clinical outcomes. A Cascade of Care (CoC) framework, first popularized as a public health accountability strategy to stem the spread of HIV, has been adapted specifically for OUD. The CoC framework has been promoted by the NIH and several states and jurisdictions for organizing quality improvement efforts through clinical, policy, and administrative levers to improve OUD treatment initiation and retention. This roadmap details CoC design domains based on available data and potential linkages as individual state agencies and health systems typically rely on limited datasets subject to diverse legal and regulatory requirements constraining options for evaluations. Both graphical decision trees and catalogued studies are provided to help guide efforts by state agencies and health systems to improve data collection and monitoring efforts under the OUD CoC framework.

Non-prescribed buprenorphine preceding treatment intake and clinical outcomes for opioid use disorder.

OBJECTIVE: Successful retention on buprenorphine improves outcomes for opioid use disorder (OUD); however, we know little about associations between use of non-prescribed buprenorphine (NPB) preceding treatment intake and clinical outcomes. METHODS: The study conducted observational retrospective analysis of abstracted electronic health record (EHR) data from a multi-state nationwide office-based opioid treatment program. The study observed a random sample of 1000 newly admitted patients with OUD for buprenorphine maintenance (2015-2018) for up to 12 months following intake. We measured use of NPB by mandatory intake drug testing and manual EHR coding. Outcomes included hazards of treatment discontinuation and rates of opioid use. RESULTS: Compared to patients testing negative for buprenorphine at intake, those testing positive (59.6%) had lower hazards of treatment discontinuation (HR = 0.52, 95% CI: 0.44, 0.60, p < 0.01). Results were little changed following adjustment for baseline opioid use and other patient characteristics (aHR: 0.60, 95% CI: 0.51, 0.70, p < 0.01). Risk of discontinuation did not significantly differ between patients by buprenorphine source: prescribed v. NPB (reference) at admission (HR = 1.15, 95% CI: 0.90, 1.46). Opioid use was lower in the buprenorphine positive group at admission (25.0% vs. 53.1%, p < 0.0001) and throughout early months of treatment but converged after 7 months for those remaining in care (17.1% vs. 16.5%, p = 0.89). CONCLUSION: NPB preceding treatment intake was associated with decreased hazards of treatment discontinuation and lower opioid use. These findings suggest use of NPB may be a marker of treatment readiness and that buprenorphine testing at intake may have predictive value for clinical assessments regarding risk of early treatment discontinuation.