What is an appropriate design for a randomized controlled trial evaluating integrated interagency care versus usual care in adults aged 18‑65 with a DSM‑5 diagnosis of substance use disorder?

Randomized Controlled Trial Design for Integrated Interagency Care in Substance Use Disorder Treatment

Design your RCT using a cluster randomization approach with integrated collaborative care as the intervention arm versus treatment-as-usual control, measuring medication initiation, linkage to follow-up care, and infection-related readmissions as primary outcomes over 6-12 months.

Core Design Framework

Intervention Model

The intervention should be based on the collaborative care model that integrates SUD treatment with medical services, following the evidence-based approach validated in multiple settings 1. The most recent and highest-quality evidence demonstrates that integrated care significantly improves medication initiation (OR 6.26) and linkage to follow-up care (OR 5.76) compared to usual care 2.

Structure your integrated intervention to include:

• Addiction medicine specialist and care manager team working collaboratively 2
• Motivational interviewing and discharge planning as core components 2
• Co-located services providing SUD treatment, medical care, and harm reduction in single appointments 3
• Weekly care coordination meetings between providers 3

Randomization Strategy

Use cluster randomization rather than individual randomization 4, 5. This approach:

• Randomizes entire clinics or practices rather than individual patients
• Prevents contamination between intervention and control arms
• Reflects real-world implementation where organizational changes affect entire sites
• Requires accounting for intracluster correlation coefficient (ICC) in sample size calculations 4

Consider a stepped-wedge design where all sites begin in usual care and sequentially cross over to the intervention at randomized time points 4. This design allows all sites to eventually receive the intervention while maintaining scientific rigor.

Primary Outcomes (Prioritize These)

Based on the strongest evidence for morbidity and mortality reduction:

1. Medication initiation for OUD/AUD during index treatment - demonstrated 6-fold improvement in integrated care 2
2. Linkage to follow-up care within 30 days - nearly 6-fold improvement shown 2
3. Infection-related readmissions at 6 months - critical for PWID with injection-related infections 3

Secondary Outcomes

• Substance-free days at 6 and 12 months
• Treatment retention rates
• Quality of life measures
• Healthcare utilization and costs 6
• Implementation outcomes: acceptability, appropriateness, feasibility 5

Critical Design Specifications

Population Inclusion

• Adults aged 18-65 with DSM-5 SUD diagnosis
• Focus on patients with substance abuse-related medical conditions (SAMCs) - this subgroup shows the strongest benefit from integrated care (OR 1.90 for abstinence) 6
• Include both opioid and stimulant use disorders, particularly those with injection-related complications 3

Sample Size Considerations

• Account for cluster effects and ICC when calculating sample size 4
• Plan for approximately 30-40% dropout rates typical in SUD populations 7
• The pilot study by 2 used 88 patients across two arms; a definitive trial requires 300+ participants to detect clinically meaningful differences 3

Intervention Duration

• Minimum 6-month active intervention period with monthly appointments 3
• Include 6-month post-intervention follow-up to assess sustainability 3
• Total study duration of 12 months aligns with evidence showing divergent outcomes emerge over time 7

Implementation Science Components

Integrate the EPIS Framework (Exploration, Preparation, Implementation, Sustainment) to assess implementation outcomes alongside effectiveness 3. This dual focus is critical because:

• Organizational readiness significantly affects adoption 5
• Provider acceptability and feasibility predict sustainability 5
• Real-world implementation barriers must be identified early 8

Monitor throughout the trial:

• Enrollment rates (patient and non-patient participants like clinicians) 4
• Intervention fidelity and component delivery 2
• Between-cluster variability 4
• Qualitative data on implementation barriers 4

Data Safety Monitoring

Establish an independent Data and Safety Monitoring Board (DSMB) for trials of this scope 4. The DSMB should review:

• Enrollment adequacy against targets
• Comparability of intervention arms at baseline
• Primary outcome ascertainment from electronic health records
• Interim analyses for futility or early effectiveness
• Adverse events, particularly overdose deaths

Critical Pitfalls to Avoid

Do not design a brief intervention study - the negative trial by 7 showed that a 6-week intensive program without sustained follow-up actually resulted in worse outcomes than usual care at 12 months. The control group had significantly more substance-free days (265 vs 139 days, p=0.021). This demonstrates that inadequate intervention duration or intensity can be harmful.

Do not treat SUD and medical conditions separately - the evidence consistently shows integrated care outperforms parallel uncoordinated services 1, 6, 9. Patients with SAMCs specifically benefit from integration, with an incremental cost-effectiveness ratio of only $1,581 per additional abstinent patient 6.

Do not rely solely on abstinence outcomes - while important, focus on medication initiation and linkage to care as primary outcomes since these show the strongest effect sizes and directly impact mortality 2. Substance use outcomes may not differ significantly between groups in the short term 2, 10.

Do not ignore organizational factors - facility type, accreditation, insurance acceptance, and state Medicaid policies significantly predict integration success 8. Stratify randomization by these factors or include them as covariates.

Measurement Strategy

Use validated instruments embedded in clinical workflow:

• Electronic health record data for medication initiation, readmissions, and healthcare utilization 4, 3
• Direct-to-patient electronic surveys for patient-reported outcomes 4
• Administrative data for cost-effectiveness analyses 6
• Implementation outcome measures following Proctor's taxonomy 5

Conduct intent-to-treat analysis as the primary analysis, with per-protocol analysis as sensitivity analysis 7. Use conservative imputation methods for missing data, carrying forward last observation or using baseline values 7.