AI for Substance Use Care in the Hospital and Emergency Department: From Screening Models to Clinical Agents

Majid Afshar, MD, MS
University of Wisconsin, School of Medicine and Public Health

ABSTRACT:
Dr. Majid Afshar will explore how artificial intelligence tools that process clinical notes—such as natural language processing and large language models—are being implemented in real-world hospital settings. He will share lessons from deploying AI systems to identify opioid misuse and support addiction care, highlighting both opportunities and risks. Drawing from the Learning Health System at the University of Wisconsin, he will offer pragmatic strategies for evaluating, monitoring, and continuously improving AI tools embedded in clinical workflows.

Integrating Effective Substance Use Disorder Treatment in the Psychiatry Treatment Setting

Melanie Bennett, PhD
University of Maryland, School of Medicine

ABSTRACT:
Patients with mental illness are diagnosed with tobacco and alcohol use disorders at rates 2-4 times higher than the general population and experience worse physical and mental health outcomes because of these comorbidities. FDA-approved medications and effective psychosocial interventions are available but underutilized in these patients, who are often referred to specialty SUD programs despite recommendations for integrated SUD/mental health care. In this presentation, Dr. Bennett will describe two research studies designed to increase the implementation of substance use disorder treatment in psychiatric treatment settings. Study 1 is a completed pilot clinical trial providing tobacco treatment to patients with mental illness who smoke during an inpatient psychiatric hospitalization. Study 2 is an ongoing exploratory project of Implementation Facilitation to increase prescribing of medications for alcohol use disorder in three community psychiatry treatment programs. Lessons learned and next steps for wider-scale implementation will be discussed.  

Peer Recovery Specialists as EBI Providers: What Does Successful Implementation and Scaling Require?

Justin Bell, PhD
Chestnut Health Systems

ABSTRACT:
Peer recovery specialists (PRS) are increasingly trained to deliver evidence-based interventions (EBIs) such as motivational interviewing, behavioral activation, and SBIRT. This shift expands access to care but also raises questions: Where do EBIs fit in peers’ role scope? What makes a successful implementation? How can policy ensure peer providers can deliver EBIs sustainably?

This presentation shares findings from Bell et al.’s recent scoping review of factors influencing the delivery of EBIs by PRS. Twelve studies were synthesized using Proctor's implementation taxonomy, with determinants classified as facilitators or barriers across outcomes such as acceptability and sustainability. Results explore facilitators of implementation success that often happened on an individual- or organizational-level, and barriers that were often systemic and tied to occupational conditions.

The presentation situates these findings within the current policy landscape, including recent federal initiatives to expand the peer workforce. Drawing on the history of peer support as a social movement, practical recommendations will be shared for aligning implementation and workforce policy so that scaling peer-delivered EBIs strengthens rather than erodes the values that make peer support effective.