PRimary Care Opioid Use Disorders Treatment (PROUD) Trial (PROUD)

Effective treatment for opioid use disorders (OUDs) requires medications. Two medications for treating OUDs-buprenorphine and injectable naltrexone-can be prescribed in primary care (PC). However, despite the current opioid epidemic and expert recommendations that OUDs should be treated in PC, most PC clinics do not offer treatment for OUDs. This reflects a lack of consensus among health system leaders and clinicians that OUDs should be treated in PC. The PRimary care Opioid Use Disorders treatment (PROUD) Trial is a pragmatic cluster-randomized, quality improvement trial that evaluates implementation of a team-based approach to PC supported by a full time nurse (the "PROUD intervention"). This type of team-based PC is often referred to as "collaborative care" for management of OUDs in PC, and this type of trial is often referred to as a Hybrid Type III implementation trial. The trial is being conducted in 6 diverse health systems spanning 5 states (New York, Florida, Michigan, Texas, and Washington), with 2 PC clinics in each system randomized. One clinic is randomly selected to implement the PROUD intervention and the other continues usual PC (UPC). The overall objective of the PROUD trial is to provide information to guide health system leaders who are faced with the decision of whether or not to treat OUDs in PC, by evaluating the benefits of implementing the PROUD intervention that integrates high quality OUD treatment (i.e. buprenorphine or injectable naltrexone) into the normal flow of PC. The primary objective of the PROUD trial is to evaluate whether the PROUD intervention increases OUD treatment with buprenorphine or injectable naltrexone, documented in the electronic health records (EHRs) of PC patients, over a 2 year follow-up, as compared to UPC. The primary hypothesis is that there will be a significant increase in the number of patient-days of medication treatment for OUDs documented in the EHR of PC patients in the 2 years after clinics are randomized to the PROUD intervention compared to PC clinics randomized to UPC. This implementation objective reflects whether the PROUD intervention increases initiation of and/or retention in OUD treatment, documented in EHRs within medical settings. The main secondary objective is to test the hypothesis that PC patients with OUDs documented in their EHRs in the 3 years prior to randomization who receive care in PROUD intervention clinics, compared to those who receive care in UPC clinics, will have fewer days of acute care utilization (including urgent care, emergency department [ED] and hospital care) in the 2 years after randomization. This effectiveness objective assesses whether implementation of the MA Model improves patient outcomes.

Pragmatic Trial: Six health systems are participating in a pragmatic, cluster-randomized, quality improvement trial (hybrid Type III implementation trial). Each of the 6 health systems provided 2 PC clinics willing to be in the randomized controlled trial. Each health system will implement collaborative care for OUD treatment in one randomly selected PC clinic. All data used to evaluate primary and secondary outcomes are secondary data from the EHR or insurance claims.

Statistical analysts (biostatisticians) who analyze trial results will be masked for objectives 1 and 2. Main analyses of the primary outcome will be conducted by the National Drug Abuse Treatment Clinical Trials Network (NDAT CTN) Data and Statistics Center (DSC). The de-identified dataset provided to the DSC by the data management team at Kaiser Permanente Washington Health Research Institute will have the identity of the intervention and Usual Primary Care clinic at each site masked.

Primary care clinics randomized to the PROUD Intervention will implement the Massachusetts (MA) Model of collaborative care for opioids use disorders (OUDs). The PROUD trial provides financial support to cover the nurse case manager (NCM) salary and technical assistance for the duration of the study, but the health systems-not investigators-implement the MA Model as part of quality improvement, and the health system and its clinicians provide all clinical care.

Clinics randomized to usual primary care do not receive any resources or support from the study but are free to improve opioid use disorder (OUD) care in any way they choose.

The PROUD intervention includes 3 strategies used to implement the MA Model: 1. Clinic leadership receives funding for a 1.0 full time equivalent NCM for 2 years after randomization and technical support for recruiting and hiring the NCM. Once hired for the study, the NCM will receive technical assistance (TA) from experts at Boston Medical Center (BMC) supported by PROUD, but NCMs will be employed and supervised by the health system. 2. Experts at BMC who originally developed and disseminated the MA Model will: provide intervention clinics with a Manual; train PROUD NCMs at BMC; and provide the ongoing TA for 2 years after randomization. 3. At least 3 primary care providers in the PROUD intervention clinic will obtain DEA waivers to prescribe buprenorphine for OUDs, if not already waivered, and work closely with the NCM to offer high quality primary care for OUDs (e.g. medication treatment with buprenorphine or naltrexone with close follow-up to maximize retention in treatment).

Clinic-level number of patient-days of OUD treatment with buprenorphine and injectable naltrexone documented in the EHR during the period from randomization until two years after, reported per 10,000 primary care patients in the clinic in the 2 years post-randomization.

Patient-level number of days of acute care utilization during the period from randomization until two years after, among patients with an OUD diagnosis documented in the EHR in the three years prior to randomization.

Clinic-level number of patients with a new International Classification of Disease (ICD) code for OUD documented in the EHR during the period from randomization until two years after who did not have an OUD diagnosis documented in the EHR in the three years prior to randomization, reported per 10,000 primary care patients in the clinic in the 2 years post-randomization.

Clinic-level number of patients who initiate (1) buprenorphine or (2) injectable naltrexone with an OUD diagnosis as documented in the EHR during the period from randomization until two years after, reported per 10,000 primary care patients in the clinic in the 2 years post-randomization. Measure will be calculated for any initiation and separately for initiation of each type of medication.

Clinic-level number of patients initiating OUD treatment during the period from randomization until two years after randomization as documented in the EHR, who also receive OUD treatment on 80% of days available after initiation, reported per 10,000 primary care patients in the clinic in the 2 years post-randomization.

Patient-level number of visits to urgent care or EDs during the period from randomization until two years after, among patients with an OUD diagnosis documented in the EHR in the three years prior to randomization.

Patient-level number of days hospitalized during the period from randomization until two years after, among patients with an OUD diagnosis documented in the EHR in the three years prior to randomization.

Patient-level number of prescriptions of naloxone for overdose management in the period from randomization until two years after, among patients with an OUD diagnoses in the three years prior to randomization.

Inclusion Criteria: The 12 clinics were eligible if their health system leaders agreed they would participate prior to study start. Patients are eligible for inclusion in the sample for analyses of the PROUD trial if they visited one of the randomized clinics at any time in the 5 year study period. Specific inclusion criteria for the trial are: Age is 16 to 90 years at any time during the study; and Visited a PROUD trial primary care clinic in the 3 years prior to randomization or the 2 years after (note: one of the 6 study sites may only be able to provide EHR data for 2 years before randomization) Exclusion Criteria: Patients who have requested through their health systems to opt out of research will be excluded from this study.

The main analytic dataset for the primary aim will be de-identified and shared on the NIDA Data Share website. This dataset will have no data elements representing clinics or health systems. The NIDA Data Share website will explicitly indicate that data elements for site or clinic (which would be masked) may be obtained from the study principal investigator on a case-by-case basis, but access will be highly restricted and may require an additional data use agreement between the original sites and the requesting researcher.

De-identified public datasets for CTN protocols will be made available on NIDA Data Share after (1) The primary paper has been accepted for publication, or (2) The final trial database has been locked for 18 months, whichever comes first.

The data on NIDA Data Share website are accessible to anyone. Sharing of datasets with clinic and site included will be determined on a case by case basis by the study principal investigator.

Alford DP, LaBelle CT, Kretsch N, Bergeron A, Winter M, Botticelli M, Samet JH. Collaborative care of opioid-addicted patients in primary care using buprenorphine: five-year experience. Arch Intern Med. 2011 Mar 14;171(5):425-31. doi: 10.1001/archinternmed.2010.541.

LaBelle CT, Han SC, Bergeron A, Samet JH. Office-Based Opioid Treatment with Buprenorphine (OBOT-B): Statewide Implementation of the Massachusetts Collaborative Care Model in Community Health Centers. J Subst Abuse Treat. 2016 Jan;60:6-13. doi: 10.1016/j.jsat.2015.06.010. Epub 2015 Jun 26.

Campbell CI, Saxon AJ, Boudreau DM, Wartko PD, Bobb JF, Lee AK, Matthews AG, McCormack J, Liu DS, Addis M, Altschuler A, Samet JH, LaBelle CT, Arnsten J, Caldeiro RM, Borst DT, Stotts AL, Braciszewski JM, Szapocznik J, Bart G, Schwartz RP, McNeely J, Liebschutz JM, Tsui JI, Merrill JO, Glass JE, Lapham GT, Murphy SM, Weinstein ZM, Yarborough BJH, Bradley KA. PRimary Care Opioid Use Disorders treatment (PROUD) trial protocol: a pragmatic, cluster-randomized implementation trial in primary care for opioid use disorder treatment. Addict Sci Clin Pract. 2021 Jan 31;16(1):9. doi: 10.1186/s13722-021-00218-w.

Bobb JF, Qiu H, Matthews AG, McCormack J, Bradley KA. Addressing identification bias in the design and analysis of cluster-randomized pragmatic trials: a case study. Trials. 2020 Mar 23;21(1):289. doi: 10.1186/s13063-020-4148-z.