1 - (SYM 37) Leveraging Relational Agents to Improve Substance Use Screening, Brief Intervention, and Referral to Treatment in Trauma Centers

Up to 50% of the 3 million adults hospitalized after traumatic injury annually have positive toxicology tests at admission. Trauma centers are required to screen and address substance use but often have limited resources to do so while they prioritize survival and physical recovery. As such, the structure and quality of screening, brief intervention, and referral to treatment (SBIRT) is highly variable in trauma centers, which contributes to poor uptake of SBIRT and inequitable care for underserved patients. Relational agents (RAs) are animated computer agents that simulate face-to-face counseling, build rapport, and establish a therapeutic alliance in a standardized way. RAs improved fidelity to SBIRT and problems associated with alcohol use in a primary care setting. However, these benefits may not generalize to trauma centers due to a range of implementation challenges. We sought to establish feasibility of RA-delivered SBIRT in a trauma center and characterize preliminary outcomes relative to treatment as usual (TAU). TAU patients who received the CAGE-AID (substance use screener; n=100) are compared to n=100 patients who received RA-delivered SBIRT. TAU patients’ data were prospectively extracted from the electronic medical record during their hospital admission. The CAGE-AID was administered by a nurse or resident who alerted social work to address positive screens. Patients in the RA condition were handed a tablet with the RA, which administers the CAGE-AID and immediately offers a brief intervention (BI) and treatment referral for patients with positive screens. RA data collection is ongoing with n=27 enrolled; the full sample will be presented. Participants (N=127) averaged 44.5 (SD = 18.9) years old and were predominantly male (63.8%), White (55.1%), and non-Hispanic/Latino (100%); 37.8% were uninsured and 22.0% had private insurance. Motor vehicle collision (45.7%) and gunshot wound (14.2%) were primary mechanisms of injury. In TAU, 10% (n=10) screened positive and only 30% (n=3/10) received BI; 2 of 3 patients who needed a referral were offered one. In the RA group, 25.9% (n=7) screened positive, 71.4% (n=5/7) received BI, and 100% of patients who needed a referral (n=2) were offered one. The RA outperformed provider-delivered SBIRT on all accounts. RAs may also yield more positive screens than TAU screening, likely due to increased comfort and decreased fear of stigma in disclosing substance use. RA-automated technology may be a viable solution to improve quality and equity of SBIRT while reducing provider burden.