# A relational agent for treating substance use in adults: A randomized controlled trial with a psychoeducational comparator. > **Author:** Benjamin Caplan, MD (ABFM Board Certified, MA Lic. #240562, NPI: 1962722637) > **Publisher:** CED Clinic (https://cedclinic.com) > **Canonical URL:** https://cedclinic.com/digital-intervention-substance-use-disorders-randomized-trial/ > **Published:** 2026-04-27T23:15:33-04:00 | **Last Modified / Clinically Audited:** 2026-09-28T10:00:25-04:00 > **Clinical Summary:** New research shows smartphone-based relational agent reduces substance use occasions more than education alone in adults with problematic use patterns. --- CED Clinical Relevance  #80High Clinical Relevance Evidence Brief | CED ClinicDigital relational agent showed significant reduction in [substance use](https://cedclinic.com/stroke-risk-assessment-substance-use/) occasions compared to psychoeducational control in adults with problematic [substance use](https://cedclinic.com/stroke-risk-assessment-substance-use/). [Substance Use Disorders](https://cedclinic.com/cannabinoid-clinical-trials-mental-disorders-evidence-2/)Digital TherapeuticsAi[Addiction Treatment](https://cedclinic.com/cannabis-based-therapies-smoking-cessation/)Rct Digital relational agent showed significant reduction in substance use occasions compared to psychoeducational control in adults with problematic substance use. What This Study Teaches Us This study demonstrates that a conversational AI agent can deliver measurable therapeutic benefit for substance use reduction in a scalable digital format. The randomized design with active psychoeducational control strengthens confidence that the relational agent itself, not just digital engagement, drives the observed benefit. Why This Matters Digital therapeutics for [substance use disorders](https://cedclinic.com/cannabinoid-clinical-trials-mental-disorders-evidence-2/) address critical access barriers in traditional treatment settings. This evidence supports the clinical utility of AI-powered interventions as adjunctive or standalone tools, particularly for patients unable to access conventional [addiction treatment](https://cedclinic.com/cannabis-based-therapies-smoking-cessation/) services. Study Snapshot Study Type Randomized Controlled Trial Population U.S. adults with problematic substance use (CAGE-AID≥2), N=258 recruited, N=202 analyzed Intervention Woebot for Substance Use Disorders (W-SUD) smartphone application with guided self-help relational agent Comparator Email-delivered psychoeducational control Primary Outcome Change in past-month substance use occasions from baseline to 8-weeks end-of-treatment Key Finding Significant reduction in substance use occasions with W-SUD compared to control Journal Journal of Substance Use and Addiction Treatment Year 2024 Clinical Bottom Line Woebot for Substance Use Disorders demonstrated statistically significant reduction in substance use occasions compared to psychoeducational materials alone in adults with problematic use. This represents meaningful evidence for digital relational agents as therapeutic tools in [addiction medicine](https://cedclinic.com/cognitive-flexibility-alcohol-cocaine-misuse-effects/). What This Paper Does Not Show The abstract does not specify which substances were targeted, the magnitude of reduction achieved, or long-term durability beyond one month post-treatment. It also cannot establish whether benefits persist or translate to clinically significant outcomes like sustained recovery. Where This Paper Deserves Skepticism The 22% dropout rate (258 to 202 analyzed) raises questions about real-world engagement sustainability. Without effect sizes or clinical significance thresholds, the practical impact remains unclear, and one-month follow-up is insufficient to assess meaningful recovery outcomes. Dr. Caplan's Take I’m encouraged by rigorous methodology comparing against active control rather than waitlist, but I need to see effect sizes and longer follow-up before recommending this clinically. The technology shows promise for expanding access, but substance use reduction alone doesn’t equal sustained recovery without broader psychosocial support. What a Careful Reader Should Take Away This study provides preliminary evidence that AI-powered relational agents can reduce substance use occasions in controlled settings. However, the clinical significance, durability, and real-world effectiveness require additional investigation before widespread clinical implementation. Join the Conversation Have a question about how this applies to your situation? [Ask Dr. Caplan →](https://benjamincaplan.com/ask-dr-caplan) Want to discuss this topic with other patients and caregivers? [Join the forum discussion →](https://cedclinic.com/interact/discussion-forums/) Have thoughts on this? Share it: [X Share on X](https://twitter.com/intent/tweet?text=Just%20read%20this%20on%20CEDclinic.com%3A%20A%20relational%20agent%20for%20treating%20substance%20use%20in%20adults%3A%20A%20randomized%20controlled%20trial%20with%20a%20psycho...%20%40drcaplan&url=https%3A%2F%2Fcedclinic.com%2Fdigital-intervention-substance-use-disorders-randomized-trial%2F)[in Share on LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fcedclinic.com%2Fdigital-intervention-substance-use-disorders-randomized-trial%2F)[🦅 Share on BlueSky](https://bsky.app/intent/compose?text=Just%20read%20this%3A%20A%20relational%20agent%20for%20treating%20substance%20use%20in%20adults%3A%20A%20randomized%20controlled%20trial%20with%20a%20psycho...%20via%20%40drcaplan.bsky.social%20https%3A%2F%2Fcedclinic.com%2Fdigital-intervention-substance-use-disorders-randomized-trial%2F)[📷 Follow on Instagram](https://www.instagram.com/drbenjamincaplan)[📝 Read more on Substack](https://doctorapprovedcannabis.substack.com/)[🔔 Subscribe via RSS](https://cedclinic.com/feed/) Table of Contents - [FAQ](#FAQ)[What is Woebot for Substance Use Disorders (W-SUD) and how does it work?](#What_is_Woebot_for_Substance_Use_Disorders_W-SUD_and_how_does_it_work) - [How effective was W-SUD compared to standard educational materials?](#How_effective_was_W-SUD_compared_to_standard_educational_materials) - [Who is eligible for W-SUD treatment based on this study?](#Who_is_eligible_for_W-SUD_treatment_based_on_this_study) - [What are the main advantages of using a digital relational agent for substance use treatment?](#What_are_the_main_advantages_of_using_a_digital_relational_agent_for_substance_use_treatment) - [How long does treatment with W-SUD last and when can patients expect to see results?](#How_long_does_treatment_with_W-SUD_last_and_when_can_patients_expect_to_see_results) - [Related](#Related) ## FAQ #### What is Woebot for Substance Use Disorders (W-SUD) and how does it work? W-SUD is a smartphone application featuring a relational agent designed to provide guided self-help support for tracking and reducing problematic substance use. The app uses conversational AI to offer accessible, scalable digital [mental health](https://cedclinic.com/cannabis-psychiatric-research-findings-depression-anxiety/) intervention that can reduce barriers associated with traditional treatment approaches. #### How effective was W-SUD compared to standard educational materials? In this randomized controlled trial of 258 U.S. adults with problematic substance use, W-SUD demonstrated significant reduction in substance use occasions compared to an email-delivered psychoeducational control group. The primary outcome measured changes in past-month substance use from baseline to 8-week end-of-treatment, with follow-up at 1 month post-treatment. #### Who is eligible for W-SUD treatment based on this study? The study included U.S. adults with problematic substance use as identified by a CAGE-AID score of 2 or higher. This screening tool helps identify individuals with substance use problems who may benefit from digital intervention support. #### What are the main advantages of using a digital relational agent for substance use treatment? Digital relational agents like W-SUD offer high accessibility and scalability, potentially reducing common barriers to traditional substance use treatment such as stigma, cost, and geographic limitations. The technology provides 24/7 availability and personalized support through smartphone access. #### How long does treatment with W-SUD last and when can patients expect to see results? Based on this study, the W-SUD intervention was delivered over 8 weeks with efficacy measured at end-of-treatment. The study also included a 1-month follow-up assessment to evaluate sustained treatment effects beyond the active intervention period. [https://cedclinic.com/contact/](https://cedclinic.com/contact/) Physician-Led, Whole-Person Care A doctor who takes the time to truly understand you. Personal care that starts with listening and is guided by experience and ingenuity. Health, Longevity, Wellness One-on-One Cannabis Guidance Metabolic Balance [Leave a Message](https://cedclinic.com/contact/) [Metabolic Care](https://cedclinic.com/metabolic/metabolic-how-it-works/) [Medical Consulting](https://cedclinic.com/book/) [Cannabis Care](https://cedclinic.com/book/) [https://cedclinic.com/contact/](https://cedclinic.com/contact/) ### Related --- ### Verification & Medical Governance - **Lead Clinical Author:** Benjamin Caplan, MD - **Credentials:** Board Certified in Family Medicine (ABFM), Founder & CMO CED Clinic & CED Foundation, Author of *The Doctor-Approved Cannabis Handbook* (Penguin Random House). - **Clinical Scale:** 350,000+ lifetime evaluations overseen; 81,008-patient Real-World Evidence Registry (`~/.ced_rwe_vault.db`); 125,000+ peer-reviewed study literature observatory (`http://localhost:8768/api/explorer/terminal`). - **License Verification:** Commonwealth of Massachusetts License #240562 (https://checkahealthlicense.mass.gov/). - **NPI Registry:** 1962722637 (https://npiregistry.cms.hhs.gov/provider-view/1962722637). - **Ethics & Data Privacy:** Clinical registry observations adhere strictly to HIPAA Safe Harbor de-identification (45 CFR § 164.514(b)) and criteria for secondary research exemption under 45 CFR § 46.104(d)(4).