Using AI and Peer Coaching to Address Racial Disparities Among People Who Use Opioids

Friends Research Institute

Status: Enrolling

Conditions: Opioid Use Disorder

Treatments: Behavioral: AI driven SDH-enhanced text only; Behavioral: PRC supported text + AI driven SDH-enhanced text; Behavioral: Treatment as Usual (TAU)

Study type: Interventional

Funder types: Other

Identifiers

NCT06569667

R18HS029783 (U.S. AHRQ Grant/Contract)

Details and patient eligibility

About

Black and Latinx people who use opioids are disproportionately impacted by opioid overdose deaths. The proposed study assesses the efficacy of an open source, multimodal artificial intelligence-driven texting tool combined with peer recovery coach-supported text contact that delivers social services, stigma reduction, health habitus, and patient navigation content addressing social determinants of health to enhance receipt of buprenorphine in primary care among emergency department-enrolled Black / Latinx people who use opioids.

Full description

The investigators have validated a theory-driven, artificial intelligence (AI)-driven texting tool using natural language processing to facilitate real-time text responses to patient queries combined with automated texts facilitating receipt of buprenorphine in office-based opioid treatment (OBOT) and social services that address social determinants of health (SDH). This open-source texting tool offers passive reminders, informational content, and interactive two-way response algorithms without personal staff contact. In addition, the investigators have adapted an efficacious cultural and structural humility training for PRCs that goes beyond SDH to also address stigma reduction, discrimination, health habitus, and patient navigation to enhance uptake of primary care and social services for PWUO.

Using a three-arm, comparative effectiveness trial design, the specific aims are to: (1) Assess the efficacy of PRC supported text-based care/services coordination with PWUO + AI-driven SDH-enhanced text messaging (intervention arm-1) vs. AI-driven SDH-enhanced text messaging only (intervention arm-2) vs. Treatment as Usual (TAU) or printed social/medical services referrals (control) to enhance the receipt of buprenorphine in OBOT among community and ED-enrolled Black / Latinx PWUO (N=252); (2) Evaluate the implementation of the multimodal intervention (arm-1) guided by the RE-AIM and CFIR frameworks using in-depth interviews among 3 stakeholder groups: (a) frontline providers (n=10); (b) administrators (n=10); and (c) a subset of the Black and Latinx PWUO from the multimodal intervention arm-1 (n=30); and (3) Identify the resources and estimate the associated cost of implementing and sustaining the multimodal intervention and incorporate this information into a customizable budget-impact tool and conduct a comprehensive economic evaluation to calculate the relative economic value (e.g., cost-per quality-adjusted life years, cost-per OUD treatment days) of each study arm from the healthcare sector, state policymaker, and societal perspectives which will also inform implementation framed by RE-AIM.

Enrollment: 292 estimated patients

Sex: All

Ages: 18 to 99 years old

Volunteers: No Healthy Volunteers

Inclusion criteria

1. ≥18 years of age;
2. fluent in English and/or Spanish;
3. self-reported non-prescription opioid use <30 days prior to consent;
4. provision of informed consent;
5. planned stay in NYC ≥12 month;
6. self-identified Black and/or Latinx race/ethnicity;
7. positive urine toxicology for opioids per EMR records;
8. diagnosis of OUD per the Diagnostic and Statistical Manual of Mental Disorders-5
9. self-reported interest in initiating buprenorphine in primary care
10. must have a mobile phone data plan.

Exclusion criteria

1. inability to comprehend text content written at a 3rd grade reading level;
2. physical or visual disability preventing mobile phone use;
3. self-reported receipt of medications for OUD in past 30 days.

Trial design

Primary purpose: Health Services Research

Allocation: Randomized

Interventional model: Parallel Assignment

Masking: None (Open label)

---

292 participants in 3 patient groups, including a placebo group

Intervention Arm-1: PRC supported text+ AI driven SDH-enhanced text

Experimental group

Description:

Participants in this arm will receive a combination of Peer Recovery Coaches (PRCs) supported text-based care/services coordination alongside AI-driven SDH-enhanced text messages to enhance the receipt of buprenorphine in Black/ Latin people who use opioids.

Treatment:

Behavioral: PRC supported text + AI driven SDH-enhanced text

Intervention Arm-2: AI driven SDH-enhanced text only

Experimental group

Description:

Participants in this arm will receive AI-driven SDH-enhanced text messages to enhance the receipt of buprenorphine in Black/ Latin people who use opioids. Unlike Arm-1, this intervention does not include the additional support and coordination provided by Peer Recovery Coaches.

Treatment:

Behavioral: AI driven SDH-enhanced text only

Control Arm 3- Treatment as Usual

Placebo Comparator group

Description:

Control Arm-3 will receive treatment as usual (i.e., verbal instructions, NYC Dept of Health pamphlets detailing access to OUD and social services, health system smartphone application EMR patient portal).

Treatment:

Behavioral: Treatment as Usual (TAU)