SUMMARY Background: The US is in the midst of an opioid epidemic and evidence-based treatment for opioid use disorders (OUD) requires medications. Two medications for OUD (MOUD)—buprenorphine and injectable naltrexone—can be provided in primary care (PC). Systems of team-based PC show promise for improving access and retention in OUD treatment. One such model, collaborative care (CC), includes a care manager, supervised by experts, who supports PC teams in providing evidence-based high-quality OUD care for OUD. However, many patients with OUDs have mental health and other substance use (MH/SU) disorders and chronic pain, which complicate OUD engagement and retention. While CC improves outcomes of depression, other MH/SU conditions and pain, it is unknown how to optimally integrate CC for OUD with CC for other MH/SU disorders. Objective: This pragmatic trial tests whether an innovative, scalable, telephonic model of CC for OUD and comorbid conditions increases engagement in MOUD treatment and improve depression symptoms in PC patients with OUD and depression. Study team: Senior investigators with expertise leading pragmatic PC trials of CC for OUD, depression, anxiety, PTSD, alcohol use disorders, and chronic pain, have partnered with clinical experts in PC management of OUD and comorbid conditions, to design and test a novel, scalable, telephonic model of CC. Intervention: This telephonic CC model includes the 6 key elements of CC, as well as shared decision-making regarding comorbid conditions of concern to patients, and will be refined in a developmental phase in year 1. Pragmatic trial sample and design: Years 2-4, the proposed effectiveness trial will test our CC model in at least 800 PC patients with OUD and depression, followed over 12 months (at least 400 each from Kaiser Permanente Washington and University of New Mexico PC clinics). Innovative pragmatic elements of this trial include: inclusion of all eligible patients with OUD and active depression (PHQ- 9 ≥ 10) seen in participating PC clinics during trial enrollment (both sites routinely screen for depression with the PHQ-9); 50% of the sample will be randomly recruited for and consented for the CC intervention; and main outcomes will be measured in the entire eligible sample using only secondary data from electronic health records, insurance claims, and state-wide prescription monitoring program and emergency department information exchanges. These pragmatic elements avoid studying only motivated (recruited) patients and avoids activating patients randomized to usual care. Main outcomes include 1) 6 months of sustained MOUD during 12-months follow-up (primary) and 2) resolution of depression symptoms based on the PHQ-9. A secondary outcome, any major adverse events, includes: suicide attempt, self-harm, overdose, hospitalization or death. Analyses compare the 50% of eligible patients offered the CC intervention (irrespective of enrollment or engagement) to those who are not, in an in...