Impact of Potential Case Misclassification by Administrative Diagnostic Codes on Outcome Assessment of Observational Study for People Who Inject Drugs.

Introduction: Initiation of medications for opioid use disorder (MOUD) within the hospital setting may improve outcomes for people who inject drugs (PWID) hospitalized because of an infection. Many studies used International Classification of Diseases (ICD) codes to identify PWID, although these may be misclassified and thus, inaccurate. We hypothesized that bias from misclassification of PWID using ICD codes may impact analyses of MOUD outcomes. Methods: We analyzed a cohort of 36 868 cases of patients diagnosed with Staphylococcus aureus bacteremia at 124 US Veterans Health Administration hospitals between 2003 and 2014. To identify PWID, we implemented an ICD code-based algorithm and a natural language processing (NLP) algorithm for classification of admission notes. We analyzed outcomes of prescribing MOUD as an inpatient using both approaches. Our primary outcome was 365-day all-cause mortality. We fit mixed-effects Cox regression models with receipt or not of MOUD during the index hospitalization as the primary predictor and 365-day mortality as the outcome. Results: NLP identified 2389 cases as PWID, whereas ICD codes identified 6804 cases as PWID. In the cohort identified by NLP, receipt of inpatient MOUD was associated with a protective effect on 365-day survival (adjusted hazard ratio, 0.48; 95% confidence interval, .29-.81; P < .01) compared with those not receiving MOUD. There was no significant effect of MOUD receipt in the cohort identified by ICD codes (adjusted hazard ratio, 1.00; 95% confidence interval, .77-1.30; P = .99). Conclusions: MOUD was protective of all-cause mortality when NLP was used to identify PWID, but not significant when ICD codes were used to identify the analytic subjects.

Costs of implementing a multi-site facilitation intervention to increase access to medication treatment for opioid use disorder.

BACKGROUND: The United States has been grappling with the opioid epidemic, which has resulted in over 75,000 opioid-related deaths between April 2020 and 2021. Evidence-based pharmaceutical interventions (buprenorphine, methadone, and naltrexone) are available to reduce opioid-related overdoses and deaths. However, adoption of these medications for opioid use disorder has been stifled due to individual- and system-level barriers. External facilitation is an evidence-based implementation intervention that has been used to increase access to medication for opioid use disorder (MOUD), but the implementation costs of external facilitation have not been assessed. We sought to measure the facility-level direct costs of implementing an external facilitation intervention for MOUD to provide decision makers with estimates of the resources needed to implement this evidence-based program. METHODS: We performed a cost analysis of the pre-implementation and implementation phases, including an itemization of external facilitation team and local site labor costs. We used labor estimates from the Bureau of Labor and Statistics, and sensitivity analyses were performed using labor estimates from the Veterans Health Administration (VHA) Financial Management System general ledger data. RESULTS: The average total costs for implementing an external facilitation intervention for MOUD per site was $18,847 (SD 6717) and ranged between $11,320 and $31,592. This translates to approximately $48 per patient with OUD. Sites with more encounters and participants with higher salaries in attendance had higher costs. This was driven mostly by the labor involved in planning and implementation activities. The average total cost of the pre-implementation and implementation activities were $1031 and $17,816 per site, respectively. In the sensitivity analysis, costs for VHA were higher than BLS estimates likely due to higher wages. CONCLUSIONS: Implementing external facilitation to increase MOUD prescribing may be affordable depending on the payer's budget constraints. Our study reported that there were variations in the time invested at each phase of implementation and the number and type of participants involved with implementing an external facilitation intervention. Participant composition played an important role in total implementation costs, and decision makers will need to identify the most efficient and optimal number of stakeholders to involve in their implementation plans.

Urine drug testing among Medicaid enrollees initiating buprenorphine treatment for opioid use disorder within 9 MODRN states.

BACKGROUND: Treatment guidelines recommend regular urine drug testing (UDT) for persons initiating buprenorphine for opioid use disorder (OUD). However, little is known about UDT utilization. We describe state variation in UDT utilization and examine demographic, health, and health care utilization factors associated with UDT in Medicaid. METHODS: We used Medicaid claims and enrollment data from persons initiating buprenorphine treatment for OUD during 2016-2019 in 9 states (DE, KY, MD, ME, MI, NC, PA, WI, WV). The main outcome was at least 1 UDT within 180 days of buprenorphine initiation, the secondary outcome was at least 3. Logistic regression models included demographics, pre-initiation comorbidities, and health service use. State estimates were pooled using meta-analysis. RESULTS: The study cohort included 162,437 Medicaid enrollees initiating buprenorphine. The percent receiving ≥1 UDT varied from 62.1% to 89.8% by state. In the pooled analysis, enrollees with pre-initiation UDT had much higher odds of ≥1 UDT after initiation (aOR=3.83, 3.09-4.73); odds were also higher for enrollees with HIV, HCV, and/or HBV infection (aOR=1.25, 1.05-1.48) or who initiated in later years (2018 v 2016: aOR=1.39, 1.03-1.89; 2019 v 2016: aOR=1.67, 1.24-2.25). The odds of having ≥3 UDT were lower with pre-initiation opioid overdose (aOR=0.79, 0.64-0.96) and higher with pre-initiation UDT (aOR=2.63, 2.13-3.25) or OUD care (aOR=1.35, 1.04-1.74). The direction of associations with demographics varied by state. CONCLUSIONS: Rates of UDT increased over time and there was variability among states in UDT rates and demographic predictors of UDT. Pre-initiation conditions, UDT, and OUD care were associated with UDT.

National Trends in Buprenorphine Treatment for Opioid Use Disorder From 2007 to 2018.

BACKGROUND: Buprenorphine is a key medication to treat opioid use disorder (OUD). Since its approval in 2002, buprenorphine access has grown markedly, spurred by major federal and state policy changes. This study characterizes buprenorphine treatment episodes during 2007 to 2018 with respect to payer, provider specialty, and patient demographics. METHODS: In this observational cohort study, IQVIA Real World pharmacy claims data were used to characterize trends in buprenorphine treatment episodes across four time periods: 2007-2009, 2010-2012, 2013-2015, and 2016-2018. RESULTS: In total, we identified more than 4.1 million buprenorphine treatment episodes among 2 540 710 unique individuals. The number of episodes doubled from 652 994 in 2007-2009 to 1 331 980 in 2016-2018. Our findings indicate that the payer landscape changed dramatically, with the most pronounced growth observed for Medicaid (increased from 17% of episodes in 2007-2009 to 37% of episodes in 2016-2018), accompanied by relative declines for both commercial insurance (declined from 35 to 21%) and self-pay (declined from 27 to 11%). Adult primary care providers (PCPs) were the dominant prescribers throughout the study period. The number of episodes among adults older than 55 increased more than 3-fold from 2007-2009 to 2016-2018. In contrast, youth under age 18 experienced an absolute decline in buprenorphine treatment episodes. Buprenorphine episodes increased in length from 2007-2018, particularly among adults over age 45. CONCLUSIONS: Our findings demonstrate that the U.S. experienced clear growth in buprenorphine treatment-particularly for older adults and Medicaid beneficiaries-reflecting some key health policy and implementation success stories. Yet, since the prevalence of OUD and fatal overdose rate have also approximately doubled during this period, the observed growth in buprenorphine treatment did not demonstrably impact the pronounced treatment gap. To date, only a minority of individuals with OUD currently receive treatment, indicating continued need for systemic efforts to equitably improve treatment uptake.

Trends in Use of Medication to Treat Opioid Use Disorder During the COVID-19 Pandemic in 10 State Medicaid Programs.

Importance: Federal and state agencies granted temporary regulatory waivers to prevent disruptions in access to medication for opioid use disorder (MOUD) during the COVID-19 pandemic, including expanding access to telehealth for MOUD. Little is known about changes in MOUD receipt and initiation among Medicaid enrollees during the pandemic. Objectives: To examine changes in receipt of any MOUD, initiation of MOUD (in-person vs telehealth), and the proportion of days covered (PDC) with MOUD after initiation from before to after declaration of the COVID-19 public health emergency (PHE). Design, Setting, and Participants: This serial cross-sectional study included Medicaid enrollees aged 18 to 64 years in 10 states from May 2019 through December 2020. Analyses were conducted from January through March 2022. Exposures: Ten months before the COVID-19 PHE (May 2019 through February 2020) vs 10 months after the PHE was declared (March through December 2020). Main Outcomes and Measures: Primary outcomes included receipt of any MOUD and outpatient initiation of MOUD via prescriptions and office- or facility-based administrations. Secondary outcomes included in-person vs telehealth MOUD initiation and PDC with MOUD after initiation. Results: Among a total of 8 167 497 Medicaid enrollees before the PHE and 8 181 144 after the PHE, 58.6% were female in both periods and most enrollees were aged 21 to 34 years (40.1% before the PHE; 40.7% after the PHE). Monthly rates of MOUD initiation, representing 7% to 10% of all MOUD receipt, decreased immediately after the PHE primarily due to reductions in in-person initiations (from 231.3 per 100 000 enrollees in March 2020 to 171.8 per 100 000 enrollees in April 2020) that were partially offset by increases in telehealth initiations (from 5.6 per 100 000 enrollees in March 2020 to 21.1 per 100 000 enrollees in April 2020). Mean monthly PDC with MOUD in the 90 days after initiation decreased after the PHE (from 64.5% in March 2020 to 59.5% in September 2020). In adjusted analyses, there was no immediate change (odds ratio [OR], 1.01; 95% CI, 1.00-1.01) or change in the trend (OR, 1.00; 95% CI, 1.00-1.01) in the likelihood of receipt of any MOUD after the PHE compared with before the PHE. There was an immediate decrease in the likelihood of outpatient MOUD initiation (OR, 0.90; 95% CI, 0.85-0.96) and no change in the trend in the likelihood of outpatient MOUD initiation (OR, 0.99; 95% CI, 0.98-1.00) after the PHE compared with before the PHE. Conclusions and Relevance: In this cross-sectional study of Medicaid enrollees, the likelihood of receipt of any MOUD was stable from May 2019 through December 2020 despite concerns about potential COVID-19 pandemic-related disruptions in care. However, immediately after the PHE was declared, there was a reduction in overall MOUD initiations, including a reduction in in-person MOUD initiations that was only partially offset by increased use of telehealth.

Prior authorization restrictions on medications for opioid use disorder: trends in state laws from 2005 to 2019.

RESEARCH OBJECTIVE: Medications for opioid use disorder (MOUDs) - including methadone, buprenorphine, and naltrexone - are the most effective treatments for opioid use disorder (OUD). Historically, insurers have required prior authorization for MOUD, but prior authorization is often reported as a key barrier to MOUD prescribing. Some states have passed laws prohibiting MOUD prior authorization requirements. We sought to identify the frequency of MOUD prior authorization prohibitions in state laws and to categorize types of prohibitions. METHODS: We searched for regulations and statutes present in all U.S. states and Washington DC between 2005 and 2019 using MOUD-related terms in Westlaw legal software. In qualitative software, we coded laws discussing MOUD prior authorization using template analysis - a mixed deductive/inductive approach. Finally, we used coded laws to identify frequencies of states with prior authorization prohibitions, including changes over time. RESULTS: No states had laws prohibiting MOUD prior authorization between 2005 and 2015, with the first prohibition appearing in 2016. By 2019, fifteen states had MOUD prior authorization prohibitions. States varied significantly in their approach to prohibiting MOUD prior authorization. In 2019, it was more common for states to have MOUD prior authorization prohibitions applying to all insurers (n = 10 states) than to only Medicaid (n = 7 states) or only non-Medicaid insurers (n = 1 state). In 2019, general prior authorization prohibitions (n = 10 states) were more common than prohibitions only applicable to medications on the formulary, prohibitions only applicable to medications on the preferred drug list, prohibitions only applicable during the first 5 days of treatment, and prohibitions only applicable during the first 30 days of treatment. CONCLUSIONS: The number of states with an MOUD prior authorization law prohibition increased in recent years. Such laws could help expand access to life-saving OUD treatments by making it easier for clinicians to prescribe MOUD.KEY MESSAGESNo states had MOUD prior authorization prohibitions between 2005 and 2015 in state statutes or regulations, and only one state had such a prohibition in 2016.By 2019, fifteen states had an MOUD prior authorization prohibition law.States varied significantly in their approach to prohibiting MOUD prior authorization, including with respect to the insurer type, duration of the prohibition, and applicable medication.

Prevalence of Testing for Human Immunodeficiency Virus, Hepatitis B Virus, and Hepatitis C Virus Among Medicaid Enrollees Treated With Medications for Opioid Use Disorder in 11 States, 2016-2019.

BACKGROUND: Limited information exists about testing for human immunodeficiency virus (HIV), hepatitis B virus (HBV), and hepatitis C virus (HCV) among Medicaid enrollees after starting medication for opioid use disorder (MOUD), despite guidelines recommending such testing. Our objectives were to estimate testing prevalence and trends for HIV, HBV, and HCV among Medicaid enrollees initiating MOUD and examine enrollee characteristics associated with testing. METHODS: We conducted a serial cross-sectional study of 505 440 initiations of MOUD from 2016 to 2019 among 361 537 Medicaid enrollees in 11 states. Measures of MOUD initiation; HIV, HBV, and HCV testing; comorbidities; and demographics were based on enrollment and claims data. Each state used Poisson regression to estimate associations between enrollee characteristics and testing prevalence within 90 days of MOUD initiation. We pooled state-level estimates to generate global estimates using random effects meta-analyses. RESULTS: From 2016 to 2019, testing increased from 20% to 25% for HIV, from 22% to 25% for HBV, from 24% to 27% for HCV, and from 15% to 19% for all 3 conditions. Adjusted rates of testing for all 3 conditions were lower among enrollees who were male (vs nonpregnant females), living in a rural area (vs urban area), and initiating methadone or naltrexone (vs buprenorphine). Associations between enrollee characteristics and testing varied across states. CONCLUSIONS: Among Medicaid enrollees in 11 US states who initiated medications for opioid use disorder, testing for human immunodeficiency virus, hepatitis B virus, hepatitis C virus, and all 3 conditions increased between 2016 and 2019 but the majority were not tested.

Increasing Access to Buprenorphine for Opioid Use Disorder in Primary Care: an Assessment of Provider Incentives.

BACKGROUND: Primary care providers (PCPs) are essential to increasing access to office-based buprenorphine medication treatment for opioid use disorder (B-MOUD). Barriers to B-MOUD prescribing are well-documented, but there is little information regarding incentives to overcome these barriers. OBJECTIVE: To identify optimal incentives for PCPs to promote B-MOUD prescribing and compare incentive preferences across provider and practice characteristics. DESIGN: We surveyed PCPs using best-worst scaling (BWS) to prioritize seven potential incentives for B-MOUD prescribing (monetary compensation, paid vacation, protected time, professional development, reduced workload, service recognition, clinical resources). We then used a direct elicitation approach to determine preferred incentive levels (e.g., monetary thresholds) and types (e.g., specific clinical resources). PARTICIPANTS: Primary care physicians and advanced practice providers (APPs) at a large Department of Veterans Affairs healthcare system. MAIN MEASURES: B-MOUD prescribing incentive preferences and relative preference levels using descriptive statistics and conditional logistic regression with relative importance scale transformation (coefficients sum to 100, higher coefficient=greater importance). KEY RESULTS: Fifty-three PCPs responded (73% response), including 47% APPs and 36% from community-based clinics. Reduced workload (relative importance score=26.8), protected time (18.7), and clinical resources (16.8) were significantly more preferred (Ps < 0.001) than professional development (10.5), paid vacation (10.3), or service recognition (1.5). Relative importance of monetary compensation varied between physicians (12.6) and APPs (17.5) and between PCPs located at a medical center (11.4) versus community clinic (22.3). APPs were more responsive than physicians to compensation increases of $5000 and $12,000 but less responsive to $25,000; trends were similar for medical center versus community clinic PCPs. The most frequently requested clinical resource was on-demand consult access to an addiction specialist. CONCLUSIONS: Interventions promoting workload reductions, protected time, and clinical resources could increase access to B-MOUD in primary care. Monetary incentives may be additionally needed to improve B-MOUD prescribing among APPs and within community clinics.

Buprenorphine treatment episode duration, dosage, and concurrent prescribing of benzodiazepines and opioid analgesics: The effects of Medicaid prior authorization policies.

BACKGROUND: Buprenorphine is an effective medication for the treatment of opioid use disorder (OUD), but the association between prior authorization policies and quality of care for individuals receiving buprenorphine treatment is not well-understood. METHODS: Using 2006-2013 Medicaid Analytic eXtract (MAX) data from 34 states and the District of Columbia, we identified 294,031 episodes of buprenorphine treatment for OUD among individuals aged 14-64 years. We estimated generalized difference-in-differences models to examine the association between buprenorphine prior authorization policies and changes in buprenorphine treatment quality along four dimensions: (1) duration of at least 180 days, (2) dosage of at least 8 milligrams, and concurrent prescribing of (3) opioid analgesics and (4) benzodiazepines. RESULTS: Buprenorphine prior authorization policies were associated with an 11-percentage point reduction (p < 0.01) in the likelihood of episodes with a duration of at least 180 days in the first four years after policy implementation. The policy was not associated with changes in effective dosage or concurrent prescribing of opioid analgesics or benzodiazepines. CONCLUSIONS: Buprenorphine prior authorization policies were associated with a sizeable and significant reduction in episodes of at least 180 days duration, underscoring the importance of identifying and removing barriers to effective and appropriate OUD care.

Outpatient follow-up and use of medications for opioid use disorder after residential treatment among Medicaid enrollees in 10 states.

BACKGROUND: Follow-up after residential treatment is considered best practice in supporting patients with opioid use disorder (OUD) in their recovery. Yet, little is known about rates of follow-up after discharge. The objective of this analysis was to measure rates of follow-up and use of medications for OUD (MOUD) after residential treatment among Medicaid enrollees in 10 states, and to understand the enrollee and episode characteristics that are associated with both outcomes. METHODS: Using a distributed research network to analyze Medicaid claims data, we estimated the likelihood of 4 outcomes occurring within 7 and 30 days post-discharge from residential treatment for OUD using multinomial logit regression: no follow-up or MOUD, follow-up visit only, MOUD only, or both follow-up and MOUD. We used meta-analysis techniques to pool state-specific estimates into global estimates. RESULTS: We identified 90,639 episodes of residential treatment for OUD for 69,017 enrollees from 2018 to 2019. We found that 62.5% and 46.9% of episodes did not receive any follow-up or MOUD at 7 days and 30 days, respectively. In adjusted analyses, co-occurring mental health conditions, longer lengths of stay, prior receipt of MOUD or behavioral health counseling, and a recent ED visit for OUD were associated with a greater likelihood of receiving follow-up treatment including MOUD after discharge. CONCLUSIONS: Forty-seven percent of residential treatment episodes for Medicaid enrollees are not followed by an outpatient visit or MOUD, and thus are not following best practices.

Expert Panel Consensus on State-Level Policies to Improve Engagement and Retention in Treatment for Opioid Use Disorder.

Importance: In the US, recent legislation and regulations have been considered, proposed, and implemented to improve the quality of treatment for opioid use disorder (OUD). However, insufficient empirical evidence exists to identify which policies are feasible to implement and successfully improve patient and population-level outcomes. Objective: To examine expert consensus on the effectiveness and the ability to implement state-level OUD treatment policies. Evidence Review: This qualitative study used the ExpertLens online platform to conduct a 3-round modified Delphi process to convene 66 stakeholders (health care clinicians, social service practitioners, addiction researchers, health policy decision-makers, policy advocates, and persons with lived experience). Stakeholders participated in 1 of 2 expert panels on 14 hypothetical state-level policies targeting treatment engagement and linkage, evidence-based and integrated care, treatment flexibility, and monitoring or support services. Participants rated policies in round 1, discussed results in round 2, and provided final ratings in round 3. Participants used 4 criteria associated with either the effectiveness or implementability to rate and discuss each policy. The effectiveness panel (n = 29) considered policy effects on treatment engagement, treatment retention, OUD remission, and opioid overdose mortality. The implementation panel (n = 34) considered the acceptability, feasibility, affordability, and equitability of each policy. We measured consensus using the interpercentile range adjusted for symmetry analysis technique from the RAND/UCLA appropriateness method. Findings: Both panels reached consensus on all items. Experts viewed 2 policies (facilitated access to medications for OUD and automatic Medicaid enrollment for citizens returning from correctional settings) as highly implementable and highly effective in improving patient and population-level outcomes. Participants rated hub-and-spoke-type policies and provision of financial incentives to emergency departments for treatment linkage as effective; however, they also rated these policies as facing implementation barriers associated with feasibility and affordability. Coercive policies and policies levying additional requirements on individuals with OUD receiving treatment (eg, drug toxicology testing, counseling requirements) were viewed as low-value policies (ie, decreasing treatment engagement and retention, increasing overdose mortality, and increasing health inequities). Conclusions and Relevance: The findings of this study may provide urgently needed consensus on policies for states to consider either adopting or deimplementing in their efforts to address the opioid overdose crisis.

Design, Implementation, and Evolution of the Medicaid Outcomes Distributed Research Network (MODRN).

BACKGROUND: In the US, Medicaid covers over 80 million Americans. Comparing access, quality, and costs across Medicaid programs can provide policymakers with much-needed information. As each Medicaid agency collects its member data, multiple barriers prevent sharing Medicaid data between states. To address this gap, the Medicaid Outcomes Distributed Research Network (MODRN) developed a research network of states to conduct rapid multi-state analyses without sharing individual-level data across states. OBJECTIVE: To describe goals, design, implementation, and evolution of MODRN to inform other research networks. METHODS: MODRN implemented a distributed research network using a common data model, with each state analyzing its own data; developed standardized measure specifications and statistical software code to conduct analyses; and disseminated findings to state and federal Medicaid policymakers. Based on feedback on Medicaid agency priorities, MODRN first sought to inform Medicaid policy to improve opioid use disorder treatment, particularly medication treatment. RESULTS: Since its 2017 inception, MODRN created 21 opioid use disorder quality measures in 13 states. MODRN modified its common data model over time to include additional elements. Initial barriers included harmonizing utilization data from Medicaid billing codes across states and adapting statistical methods to combine state-level results. The network demonstrated its utility and addressed barriers to conducting multi-state analyses of Medicaid administrative data. CONCLUSIONS: MODRN created a new, scalable, successful model for conducting policy research while complying with federal and state regulations to protect beneficiary health information. Platforms like MODRN may prove useful for emerging health challenges to facilitate evidence-based policymaking in Medicaid programs.

Laws for expanding access to medications for opioid use disorder: a legal analysis of 16 states & Washington D.C.

Background: Medications for opioid use disorder (MOUDs) are the gold standard for OUD treatment but are underused. To our knowledge, no published study has systematically identified and categorized state policy innovations for expanding MOUD utilization.Objective: We sought to identify and categorize state MOUD policy innovations.Methods: Within a stratified random sample of 16 U.S. states and Washington D.C. we searched for 2019 state statutes and regulations related to MOUD in Westlaw legal database. We then identified laws that appeared designed to increase MOUD utilization and categorized them using a template analysis approach.Results: We found 82 laws with one or more MOUD expansion policies. We identified six high-level MOUD expansion policy categories: 1) policies expanding the availability of waivered buprenorphine providers; 2) needs assessments and policies increasing public MOUD awareness; 3) criminal justice system policies; 4) Substance use disorder (SUD) treatment and sober living facility policies; 5) insurance policies; and 6) hospital policies. SUD treatment and housing facility policies, as well as insurance policies, were most common.Conclusions: Multipronged approaches are being pursued by several states to increase MOUD access. Our results can inform policymakers of MOUD expansion approaches in other jurisdictions. Policy categories can serve as the basis for policy variables for future analyses of policy effects.

Duration of medication treatment for opioid-use disorder and risk of overdose among Medicaid enrollees in 11 states: a retrospective cohort study.

BACKGROUND AND AIMS: Medication for opioid use disorder (MOUD) reduces harms associated with opioid use disorder (OUD), including risk of overdose. Understanding how variation in MOUD duration influences overdose risk is important as health-care payers increasingly remove barriers to treatment continuation (e.g. prior authorization). This study measured the association between MOUD continuation, relative to discontinuation, and opioid-related overdose among Medicaid beneficiaries. DESIGN: Retrospective cohort study using landmark survival analysis. We estimated the association between treatment continuation and overdose risk at 5 points after the index, or first, MOUD claim. Censoring events included death and disenrollment. SETTING AND PARTICIPANTS: Medicaid programs in 11 US states: Delaware, Kentucky, Maryland, Maine, Michigan, North Carolina, Ohio, Pennsylvania, Virginia, West Virginia and Wisconsin. A total of 293 180 Medicaid beneficiaries aged 18-64 years with a diagnosis of OUD and had a first MOUD claim between 2016 and 2017. MEASUREMENTS: MOUD formulations included methadone, buprenorphine and naltrexone. We measured medically treated opioid-related overdose within claims within 12 months of the index MOUD claim. FINDINGS: Results were consistent across states. In pooled results, 5.1% of beneficiaries had an overdose, and 67% discontinued MOUD before an overdose or censoring event within 12 months. Beneficiaries who continued MOUD beyond 60 days had a lower relative overdose hazard ratio (HR) compared with those who discontinued by day 60 [HR = 0.39; 95% confidence interval (CI) = 0.36-0.42; P < 0.0001]. MOUD continuation was associated with lower overdose risk at 120 days (HR = 0.34; 95% CI = 0.31-0.37; P < 0.0001), 180 days (HR = 0.31; 95% CI = 0.29-0.34; P < 0.0001), 240 days (HR = 0.29; 95% CI = 0.26-0.31; P < 0.0001) and 300 days (HR = 0.28; 95% CI = 0.24-0.32; P < 0.0001). The hazard of overdose was 10% lower with each additional 60 days of MOUD (95% CI = 0.88-0.92; P < 0.0001). CONCLUSIONS: Continuation of medication for opioid use disorder (MOUD) in US Medicaid beneficiaries was associated with a substantial reduction in overdose risk up to 12 months after the first claim for MOUD.

Developing and validating a machine-learning algorithm to predict opioid overdose in Medicaid beneficiaries in two US states: a prognostic modelling study.

BACKGROUND: Little is known about whether machine-learning algorithms developed to predict opioid overdose using earlier years and from a single state will perform as well when applied to other populations. We aimed to develop a machine-learning algorithm to predict 3-month risk of opioid overdose using Pennsylvania Medicaid data and externally validated it in two data sources (ie, later years of Pennsylvania Medicaid data and data from a different state). METHODS: This prognostic modelling study developed and validated a machine-learning algorithm to predict overdose in Medicaid beneficiaries with one or more opioid prescription in Pennsylvania and Arizona, USA. To predict risk of hospital or emergency department visits for overdose in the subsequent 3 months, we measured 284 potential predictors from pharmaceutical and health-care encounter claims data in 3-month periods, starting 3 months before the first opioid prescription and continuing until loss to follow-up or study end. We developed and internally validated a gradient-boosting machine algorithm to predict overdose using 2013-16 Pennsylvania Medicaid data (n=639 693). We externally validated the model using (1) 2017-18 Pennsylvania Medicaid data (n=318 585) and (2) 2015-17 Arizona Medicaid data (n=391 959). We reported several prediction performance metrics (eg, C-statistic, positive predictive value). Beneficiaries were stratified into risk-score subgroups to support clinical use. FINDINGS: A total of 8641 (1·35%) 2013-16 Pennsylvania Medicaid beneficiaries, 2705 (0·85%) 2017-18 Pennsylvania Medicaid beneficiaries, and 2410 (0·61%) 2015-17 Arizona beneficiaries had one or more overdose during the study period. C-statistics for the algorithm predicting 3-month overdoses developed from the 2013-16 Pennsylvania training dataset and validated on the 2013-16 Pennsylvania internal validation dataset, 2017-18 Pennsylvania external validation dataset, and 2015-17 Arizona external validation dataset were 0·841 (95% CI 0·835-0·847), 0·828 (0·822-0·834), and 0·817 (0·807-0·826), respectively. In external validation datasets, 71 361 (22·4%) of 318 585 2017-18 Pennsylvania beneficiaries were in high-risk subgroups (positive predictive value of 0·38-4·08%; capturing 73% of overdoses in the subsequent 3 months) and 40 041 (10%) of 391 959 2015-17 Arizona beneficiaries were in high-risk subgroups (positive predictive value of 0·19-1·97%; capturing 55% of overdoses). Lower risk subgroups in both validation datasets had few individuals (≤0·2%) with an overdose. INTERPRETATION: A machine-learning algorithm predicting opioid overdose derived from Pennsylvania Medicaid data performed well in external validation with more recent Pennsylvania data and with Arizona Medicaid data. The algorithm might be valuable for overdose risk prediction and stratification in Medicaid beneficiaries. FUNDING: National Institute of Health, National Institute on Drug Abuse, National Institute on Aging.

Sociodemographic differences in quality of treatment to Medicaid enrollees receiving buprenorphine.

Background: Buprenorphine is a key medication to treat opioid use disorder, but little is known about how treatment quality varies across sociodemographic groups. Objective: We examined measures of treatment quality and explored variation by sociodemographic factors. Methods: We used Medicaid MAX data from 50 states from 2006 to 2014 to identify buprenorphine treatment episodes (N = 317,494). We used multivariable logistic regression to examine the quality of buprenorphine treatment along four dimensions: (1) sufficient duration, (2) effective dosage, and concurrent prescribing of (3) opioid analgesics and (4) benzodiazepines. We explored how quality varied by race/ethnicity, age, sex, and urbanicity. Results: In adjusted models, compared to non-Hispanic White individuals, non-Hispanic Black and Hispanic individuals had lower odds of receiving effective dosage (aORs = 0.79 and 0.89, respectively) and sufficient duration (aORs = 0.64 and 0.71, respectively), and lower odds of concurrent prescribing of opioid analgesics (aORs = 0.86 and 0.85, respectively) and benzodiazepines (aORs = 0.51 and 0.59, respectively). Older individuals had higher odds of sufficient duration (aORs from 1.21-1.33), but also had higher odds of concurrent opioid analgesics prescribing (aORs from 1.29-1.56) and benzodiazepines (aORs from 1.44-1.99). Females had higher odds of sufficient duration (aOR = 1.12), but lower odds of effective dosage (aOR = 0.77) and higher odds of concurrent prescribing of opioid analgesics (aOR = 1.25) and benzodiazepines (aOR = 1.16). Compared to individuals living in metropolitan areas, individuals living in non-metropolitan areas had higher odds of sufficient duration (aORs = 1.11 and 1.24) and effective dosage (aORs = 1.06 and 1.33), and lower odds of concurrent prescribing (aORs from 0.81-0.98). Conclusions: Black and Hispanic individuals were less likely to receive effective buprenorphine dosage and sufficient duration. Quality results were mixed for older and female individuals; although these individuals were more likely to receive treatment of sufficient duration, they were also more likely to be concurrently prescribed potentially contraindicated medications, and females were less likely to receive effective dosage. Findings raise concerns about adequacy of care for minority and other at-risk populations.

Subsequent Buprenorphine Treatment Following Emergency Physician Buprenorphine Prescription Fills: A National Assessment 2019 to 2020.

STUDY OBJECTIVE: Buprenorphine treatment for opioid use disorder provided in the emergency department with subsequent buprenorphine treatment by community prescribers is associated with improved outcomes, but the frequency with which this occurs is unknown. We examined the rates of subsequent buprenorphine treatment for buprenorphine-naïve individuals filling buprenorphine prescriptions from emergency physicians and initiated buprenorphine treatment and how such rates varied before and during the coronavirus disease 2019 (COVID-19) pandemic. METHODS: Using pharmacy claims capturing an estimated 92% of prescriptions filled at US retail pharmacies, we identified buprenorphine prescriptions filled between February 1, 2019, and November 30, 2020, written by emergency physicians. In this observational study, we calculated the rate at which patients subsequently filled buprenorphine prescriptions from other nonemergency clinicians, the frequency with which subsequent filled prescriptions were from different types of prescribers, and the changes in the rates of subsequent prescriptions following the declaration of the COVID-19 public health emergency. RESULTS: We identified 22,846 prescriptions written by emergency physicians and filled by buprenorphine-naïve patients. They were most commonly paid for by Medicaid and were in metropolitan counties; 28.5% of patients subsequently filled buprenorphine prescriptions written by other clinicians. Adult primary care physicians and advanced practice providers (eg, physician assistants and nurse practitioners) were responsible for most of the subsequent prescriptions. The rates of subsequent prescriptions were 3.5% lower after the COVID-19 public health emergency declaration. CONCLUSION: The majority of patients filling buprenorphine prescriptions written by emergency physicians do not subsequently fill prescriptions written by other clinicians, and the rates of subsequent prescriptions were lower after the declaration of the COVID-19 public health emergency. These findings highlight the need for a system of care that improves buprenorphine treatment continuity of care for patients with opioid use disorder from emergency settings to community treatment providers.

Changes in predicted opioid overdose risk over time in a state Medicaid program: a group-based trajectory modeling analysis.

BACKGROUND AND AIMS: The time lag encountered when accessing health-care data is one major barrier to implementing opioid overdose prediction measures in practice. Little is known regarding how one's opioid overdose risk changes over time. We aimed to identify longitudinal patterns of individual predicted overdose risks among Medicaid beneficiaries after initiation of opioid prescriptions. DESIGN, SETTING AND PARTICIPANTS: A retrospective cohort study in Pennsylvania, USA among Pennsylvania Medicaid beneficiaries aged 18-64 years who initiated opioid prescriptions between July 2017 and September 2018 (318 585 eligible beneficiaries (mean age = 39 ± 12 years, female = 65.7%, White = 62.2% and Black = 24.9%). MEASUREMENTS: We first applied a previously developed and validated machine-learning algorithm to obtain risk scores for opioid overdose emergency room or hospital visits in 3-month intervals for each beneficiary who initiated opioid therapy, until disenrollment from Medicaid, death or the end of observation (December 2018). We performed group-based trajectory modeling to identify trajectories of these predicted overdose risk scores over time. FINDINGS: Among eligible beneficiaries, 0.61% had one or more occurrences of opioid overdose in a median follow-up of 15 months. We identified five unique opioid overdose risk trajectories: three trajectories (accounting for 92% of the cohort) had consistent overdose risk over time, including consistent low-risk (63%), consistent medium-risk (25%) and consistent high-risk (4%) groups; another two trajectories (accounting for 8%) had overdose risks that substantially changed over time, including a group that transitioned from high- to medium-risk (3%) and another group that increased from medium- to high-risk over time (5%). CONCLUSIONS: More than 90% of Medicaid beneficiaries in Pennsylvania USA with one or more opioid prescriptions had consistent, predicted opioid overdose risks over 15 months. Applying opioid prediction algorithms developed from historical data may not be a major barrier to implementation in practice for the large majority of individuals.

Provider and Patient-panel Characteristics Associated With Initial Adoption and Sustained Prescribing of Medication for Opioid Use Disorder.

OBJECTIVES: Limited information is available regarding provider- and patient panel-level factors associated with primary care provider (PCP) adoption/prescribing of medication for opioid use disorder (MOUD). METHODS: We assessed a retrospective cohort from 2015 to 2018 within the Pennsylvania Medicaid Program. Participants included PCPs who were Medicaid providers, with no history of MOUD provision, and who treated ≥10 Medicaid enrollees annually. We assessed initial MOUD adoption, defined as an index buprenorphine/buprenorphine-naloxone or oral/extended release naltrexone fill and sustained prescribing, defined as ≥1 MOUD prescription(s) for 3 consecutive quarters from the PCP. Independent variables included provider- and patient panel-level characteristics. RESULTS: We identified 113 rural and 782 urban PCPs who engaged in initial adoption and 36 rural and 288 urban PCPs who engaged in sustained prescribing. Rural/urban PCPs who issued increasingly larger numbers of antidepressant and antipsychotic medication prescriptions had greater odds of initial adoption and sustained prescribing (P < 0.05) compared to those that did not prescribe these medications. Further, each additional patient out of 100 with opioid use disorder diagnosed before MOUD adoption increased the adjusted odds for initial adoption 2% to 4% (95% confidence interval [CI] = 1.01-1.08) and sustained prescribing by 4% to 7% (95% CI = 1.01-1.08). New Medicaid providers in rural areas were 2.52 (95% CI = 1.04-6.11) and in urban areas were 2.66 (95% CI = 1.94, 3.64) more likely to engage in initial MOUD adoption compared to established PCPs. CONCLUSIONS: MOUD prescribing adoption was concentrated among PCPs prescribing mental health medications, caring for those with OUD, and new Medicaid providers. These results should be leveraged to test/implement interventions targeting MOUD adoption among PCPs.

Accessibility to Medication for Opioid Use Disorder After Interventions to Improve Prescribing Among Nonaddiction Clinics in the US Veterans Health Care System.

Importance: With increasing rates of opioid use disorder (OUD) and overdose deaths in the US, increased access to medications for OUD (MOUD) is paramount. Rigorous effectiveness evaluations of large-scale implementation initiatives using quasi-experimental designs are needed to inform expansion efforts. Objective: To evaluate a US Department of Veterans Affairs (VA) initiative to increase MOUD use in nonaddiction clinics. Design, Setting, and Participants: This quality improvement initiative used interrupted time series design to compare trends in MOUD receipt. Primary care, pain, and mental health clinics in the VA health care system (n = 35) located at 18 intervention facilities and nonintervention comparison clinics (n = 35) were matched on preimplementation MOUD prescribing trends, clinic size, and facility complexity. The cohort of patients with OUD who received care in intervention or comparison clinics in the year after September 1, 2018, were evaluated. The preimplementation period extended from September 1, 2017, through August 31, 2018, and the postimplementation period from September 1, 2018, through August 31, 2019. Exposures: The multifaceted implementation intervention included education, external facilitation, and quarterly reports. Main Outcomes and Measures: The main outcomes were the proportion of patients receiving MOUD and the number of patients per clinician prescribing MOUD. Segmented logistic regression evaluated monthly proportions of MOUD receipt 1 year before and after initiative launch, adjusting for demographic and clinical covariates. Poisson regression models examined yearly changes in clinician prescribing over the same time frame. Results: Overall, 7488 patients were seen in intervention clinics (mean [SD] age, 53.3 [14.2] years; 6858 [91.6%] male; 1476 [19.7%] Black, 417 [5.6%] Hispanic; 5162 [68.9%] White; 239 [3.2%] other race [including American Indian or Alaska Native, Asian, Native Hawaiian or other Pacific Islander, and multiple races]; and 194 [2.6%] unknown) and 7558 in comparison clinics (mean [SD] age, 53.4 [14.0] years; 6943 [91.9%] male; 1463 [19.4%] Black; 405 [5.4%] Hispanic; 5196 [68.9%] White; 244 [3.2%] other race; 250 [3.3%] unknown). During the preimplementation year, the proportion of patients receiving MOUD in intervention clinics increased monthly by 5.0% (adjusted odds ratio [AOR], 1.05; 95% CI, 1.03-1.07). Accounting for this preimplementation trend, the proportion of patients receiving MOUD increased monthly by an additional 2.3% (AOR, 1.02; 95% CI, 1.00-1.04) during the implementation year. Comparison clinics increased by 2.6% monthly before implementation (AOR, 1.03; 95% CI, 1.01-1.04), with no changes detected after implementation. Although preimplementation-year trends in monthly MOUD receipt were similar in intervention and comparison clinics, greater increases were seen in intervention clinics after implementation (AOR, 1.04; 95% CI, 1.01-1.08). Patients treated with MOUD per clinician in intervention clinics saw greater increases from before to after implementation compared with comparison clinics (incidence rate ratio, 1.50; 95% CI, 1.28-1.77). Conclusions and Relevance: A multifaceted implementation initiative in nonaddiction clinics was associated with increased MOUD prescribing. Findings suggest that engagement of clinicians in general clinical settings may increase MOUD access.