This study tests ways to help opioid treatment programs (OTPs) keep patients in care. Staying on methadone or buprenorphine is linked to better outcomes, yet many people leave treatment early. The project will compare two approaches that provide clinics with retention/outcome quality measures and a quality-improvement (QI) toolkit-either alone or with added facilitation-against usual care.

Forty-five BayMark OTPs in multiple states will be randomly assigned to one of three groups: (1) quality measures + QI toolkit; (2) quality measures + QI toolkit + external QI facilitation; or (3) usual care.

The primary outcome is 90-day retention in treatment, measured from OTP electronic health records and Medicaid claims. Secondary outcomes include emergency department visits, hospitalizations, overdoses, and mortality. Findings will identify practical, scalable strategies to improve patient retention in OTPs.

This cluster-randomized trial is part of an NIH/NIDA-funded program to advance quality measurement and management for opioid treatment programs (OTPs). PROMOTE-QI (Project 2) tests whether providing OTPs with retention/outcome quality measures and a quality-improvement (QI) toolkit, with or without additional QI facilitation, improves patient retention compared with usual care. The study is conducted in partnership with BayMark Health Services and academic/industry collaborators.

Design and setting. Forty-five BayMark OTPs in multiple states will be randomized in equal groups to three arms (≈15 sites/arm) in a parallel-group cluster design. We anticipate ~4,500 adult MOUD initiations across the 45 sites during the 12-month post-implementation observation window. Patients are not individually assigned; outcomes are derived from EHR and Medicaid claims.

Interventions. Arm 1: Quality measures + QI toolkit. Sites receive claims-based, case-mix-adjusted retention and outcome quality measures with benchmarks, plus a toolkit (evidence summaries, case studies, and "how-to" materials) to guide retention-focused QI efforts, delivered via a secure portal.

Arm 2: Quality measures + QI toolkit + QI facilitation (NIATx). Sites receive all Arm-1 components plus structured NIATx facilitation, including establishing a change team and running Plan-Do-Study-Act (PDSA) cycles to implement and test retention strategies.

Arm 3: Usual care. Sites continue existing practices; at study end they will be offered the quality-measure portal and toolkit.

Primary and secondary outcomes. The primary endpoint is the OTP's 90-day treatment retention rate. Secondary endpoints are the OTP's 90-day retention rate of medications to treat OUD, the percentage of the OTP's patients with an ED visit or hospitalization for a substance use disorder (SUD), and the percentage with an ED visit or hospitalization for any cause.

Data sources and analysis. Outcomes will be drawn from BayMark electronic health records (EHR) and Medicaid T-MSIS claims. Analyses use generalized linear mixed models that account for clustering at the OTP level; retention/discontinuation rules (e.g., ≥31-day gap) follow the study's measurement specifications.

Implementation outcomes. The study will document strategies adopted, barriers/facilitators, and costs of QI facilitation via surveys/interviews and cost tracking.

Rationale. Prior work shows audit-and-feedback is more effective when paired with actionable guidance; the trial therefore compares quality measures + toolkit with and without facilitation to determine the incremental benefit of NIATx support.

Only BayMark OTPs are eligible for participation.

Individual participant data (IPD) will not be shared because the study relies on de-identified Medicaid claims and electronic health records obtained through data use agreements, as well as site-level information from opioid treatment programs (OTPs). These data are protected by federal and state privacy regulations and cannot be shared at the participant level. Only aggregated results will be disseminated.

This is an open-label, cluster-randomized implementation trial. OTPs, care providers, and investigators will be aware of arm assignment. Masking is not feasible because the interventions-provision of clinic-level quality measures and a QI toolkit, with or without external QI facilitation-are organizational changes visible to staff. Patient outcomes are obtained from EHR and Medicaid claims; the central analytic team may be masked to arm labels when feasible.