Highlights
- The CLARO trial found no significant difference between collaborative care and enhanced usual care (EUC) regarding time to buprenorphine initiation or cumulative days of medication for opioid use disorder (MOUD).
- Mental health outcomes, measured by PHQ-9 (depression) and PCL-5 (PTSD) scores, improved in both the intervention and control groups, with no statistically significant advantage for the collaborative care arm.
- The intervention utilized community health workers (CHWs) as care managers in low-resource primary care settings, highlighting the difficulties of managing high-acuity patients with multi-faceted biopsychosocial needs.
- Spillover effects and spontaneous improvement within the control group may have narrowed the gap between the two study arms.
Introduction: The Intersection of Opioid Use Disorder and Mental Health
The United States continues to grapple with a dual crisis: the opioid epidemic and a burgeoning mental health challenge. Patients presenting with opioid use disorder (OUD) frequently exhibit co-occurring psychiatric conditions, most notably major depressive disorder and posttraumatic stress disorder (PTSD). This comorbidity, often referred to as a dual diagnosis, is associated with poorer treatment retention, higher rates of overdose, and decreased quality of life compared to OUD alone.
Collaborative Care (CoCM) has long been heralded as the gold standard for integrating behavioral health into primary care. By utilizing a care manager and a psychiatric consultant to support primary care practitioners (PCPs), the model has proven effective for depression and anxiety. However, its efficacy in the context of OUD—particularly when combined with severe trauma or depression in low-resource settings—remains an area of active investigation. The Collaborative Care for Opioid Use Disorder and Mental Illness (CLARO) trial was designed to address this evidence gap.
Study Design and Methodology
Population and Setting
CLARO was a two-group, single-masked, pragmatic randomized clinical trial conducted across 18 primary care clinics in California and New Mexico. These sites were specifically chosen for their status as low-resource settings, often serving marginalized populations. The trial enrolled 797 adult participants (397 to EUC; 400 to intervention) between January 2021 and December 2023. Participants were required to have probable OUD alongside major depression and/or PTSD.
The Intervention: A CHW-Led Collaborative Care Model
Participants in the intervention group received six months of collaborative care. A unique feature of this trial was the use of community health workers (CHWs) as care managers. These CHWs worked alongside addiction psychiatrists and PCPs to deliver evidence-based treatments. Their role included monitoring biopsychosocial symptoms, facilitating referrals for psychotherapy, and supporting MOUD adherence. This model was tailored for settings where licensed behavioral health practitioners are scarce.
Comparators and Endpoints
The control group received Enhanced Usual Care (EUC), which included standard primary care supplemented by access to resources and periodic screenings. The primary outcomes at six months were:
1. Time to first filled buprenorphine prescription (for those not on MOUD at baseline).
2. Cumulative days of prescribed buprenorphine (obtained via state Prescription Drug Monitoring Programs).
3. Changes in PHQ-9 (depression) and PCL-5 (PTSD) scores.
Secondary outcomes included self-reported days of opioid use, substance use severity (PROMIS short form), and general health status (VR-12).
Key Findings: Outcomes for OUD and Mental Health
Despite the robust design of the intervention, the CLARO trial did not meet its primary endpoints. The data revealed that both the intervention and control groups improved over the six-month period, but the delta between them was not statistically significant.
MOUD Outcomes
For participants entering the study not taking medication, the adjusted mean difference in days until the first buprenorphine prescription was 7.0 (95% CI, -3.4 to 17.4; P = .19) for collaborative care vs. EUC. Similarly, the difference in cumulative days of prescribed buprenorphine was 4.3 (95% CI, -7.4 to 16.0; P = .47). These figures indicate that the collaborative care model did not significantly accelerate or sustain MOUD initiation compared to usual care.
Psychiatric Symptom Reduction
The mental health findings followed a similar trend. The adjusted mean difference for the PHQ-9 score was -1.0 (95% CI, -2.3 to 0.3; P = .13), and for the PTSD Checklist (PCL-5), it was -0.9 (95% CI, -4.6 to 2.8; P = .63). While the intervention group showed slight numerical improvements, they failed to reach the threshold of statistical or clinical significance over the EUC group.
Secondary and Exploratory Analysis
Secondary outcomes, including days of opioid use and general health scores, showed no significant differences between the arms. However, exploratory “as-treated” analyses—which looked at participants who actually engaged deeply with the intervention—suggested some evidence for improvements in OUD outcomes. This implies that while the intent-to-treat analysis was null, the model might offer benefits to those who are able to consistently engage with the care manager.
Expert Commentary and Clinical Interpretation
The Challenge of Complexity
Why did a proven model like CoCM fail to show superiority in this trial? One primary consideration is the sheer complexity of the patient population. OUD combined with PTSD often involves significant social determinants of health (SDOH), including housing instability and legal issues. While CHWs are excellent at addressing SDOH, they may lack the specialized clinical training required to manage high-acuity psychiatric symptoms or complex medication adjustments compared to licensed social workers or psychologists traditionally used in CoCM.
The Problem of ‘Spillover’ and EUC
In pragmatic trials, the “usual care” arm is often more robust than in real-world practice. The clinics involved were already sensitized to OUD treatment, and the act of screening and enrolling patients may have prompted PCPs in the EUC arm to provide better-than-average care. Furthermore, “spillover” can occur when PCPs apply the lessons learned from intervention patients to their control-group patients within the same clinic.
Biological and Mechanistic Insights
From a biological perspective, the recovery of the brain’s reward circuitry in OUD is a slow process, often hindered by the hyper-arousal and emotional dysregulation seen in PTSD. A six-month intervention may simply be too short to observe a significant divergence in clinical trajectories for such a chronic and relapsing condition.
Conclusion and Summary
The CLARO trial serves as a sobering reminder that evidence-based models must be carefully adapted to the specific needs of the population and the limitations of the setting. While collaborative care remains a powerful tool for general depression and anxiety, its application in OUD with high-acuity psychiatric comorbidity requires further refinement. The study suggests that simply adding a care manager may not be enough for clinically complex patients in low-resource environments.
Future research should explore more intensive or longer-duration interventions and investigate whether higher-tier clinical staffing (e.g., nurse care managers) yields better results for this specific subgroup. For now, clinicians should continue to integrate OUD and mental health services but remain vigilant about the high level of support these patients require.
Funding and ClinicalTrials.gov
This research was supported by the National Institute on Drug Abuse (NIDA) and the National Institute of Mental Health (NIMH). The trial is registered at ClinicalTrials.gov (NCT04559893 and NCT04634279).
References
1. Watkins KE, Osilla KC, McCullough CM, et al. Collaborative Care for Opioid Use Disorder and Mental Illness: The CLARO Randomized Clinical Trial. JAMA Intern Med. 2025;e257036. doi:10.1001/jamainternmed.2025.7036.
2. Archer J, Bower P, Gilbody S, et al. Collaborative care for depression and anxiety problems. Cochrane Database Syst Rev. 2012;10(10):CD006525.
3. Pincus HA, Scholle SH, Spaeth-Rublee B, et al. Quality measures for mental health and substance use: gaps, opportunities, and challenges. Health Aff (Millwood). 2016;35(6):1000-1008.
