The objective of the project is to evaluate the implementation and impact of the Acute Pain Services Expansion Program (APSEP), an expansion of the independent and formal set of services that provide comprehensive inpatient pain management consultative services and is designed to meet perioperative care needs of veterans receiving inpatient pain management.
Patient Population: Adults
Complementary and Integrative Health Stepped Care for Co-occurring Chronic Pain and PTSD
The aim of the project is to conduct a pragmatic pilot trial of a CIH-based stepped care approach v. treatment as usual in two primary care settings (one rural and one urban). The pilot trial will focus on feasibility, acceptability, and appropriateness for providers and patients (e.g., randomization, retention, and treatment satisfaction) of the stepped care approach versus usual care (n=30 per site, N=60 total). Primary clinical outcomes are pain interference and PTSD symptoms at 6-months.
Improving treatment strategies and clinical outcomes in patients with first-episode psychosis and substance use disorders
Our project will seek to identify factors associated with gaps in transitions of care for psychiatric inpatients who presented with substance-induced psychosis (SIP) for the first time. We will analyze historical electronic health record data of patients who were treated for psychosis at Harborview Medical Center. We will test the hypotheses that (1) treatment with long-acting injectable antipsychotics (LAI) and referrals to outpatient behavioral health are lower for patients diagnosed with first-episode SIP compared to those diagnosed with first-episode psychosis and that (2) patients diagnosed with first-episode SIP will have worse post-discharge outcomes (rehospitalization, ED utilization), in part due to lower use of LAI.
Strengthening financial literacy for people living with serious mental illness
Improved financial literacy among people living with serious mental illness (SMI) is associated with a higher quality of life, fewer hospitalizations, and better treatment adherence. Yet people living with SMI frequently express how their lack of financial knowledge has negative personal consequences and that they don’t know where to turn for assistance. This project will gather qualitative and quantitative data from people admitted to the Center for Behavioral Health and Learning, a psychiatric hospital, to understand the need and desire for a financial skills intervention and its role in discharge planning. The assessment will also seek input from family members/caregivers, representative payees/fiduciaries and experts in the community. Ultimately, we hope to create a replicable, standardized intervention that can be evaluated and implemented in inpatient settings and modified as necessary for outpatient settings.
Using Large Language Models to identify video platform interactions indicating suicide risk
This project will identify interaction patterns with online video platforms that are indicative of suicide risk, focusing on YouTube and TikTok. Leveraging archival data including over 5 million interaction events collected from participants in previous research, we will use combinations of neural language models to identify suicide-related “like”, “search” and “watch” events. We will then assess the temporal relationships between suicide-related interaction events and suicidal ideation, behavior and mental health challenges reported by these participants. Building on these analyses, we will proceed to model patterns of interaction, differentiating between user-initiated (e.g. search) and algorithm-prompted (e.g. recommended content without a preceding search) content to characterize the ways in which intentional and algorithmically-driven behavior drive exposure to suicide-related content. In addition, we will develop a prototype of a privacy-preserving risk monitoring tool, which will detect interactions with concerning content and leverage light-touch intervention strategies to mitigate its impact.
Using Natural Language Processing to risk-stratify inpatient psychiatry conflict and violence
This QI project aims to expand from general medical wards to inpatient psychiatry the use of predictive risk-modeling for violence or restraint, using Natural Language Processing of clinical notes. We will also assess whether NLP paired with generative AI can accurately summarize a wider range of clinical notes relevant to behavioral emergencies
Making generative AI safe for people with mental health conditions
Hundreds of millions of people are already using Large Language Models (LLMs), including for mental health purposes, which has led to inadvertent harms. Critically, people with mental health conditions may be especially vulnerable to such harms.
In this project, we will develop the first computational framework to systematically quantify and benchmark the risks that LLMs present to people with mental health conditions. Our approach will simulate interactions of hundreds of users and LLMs to evaluate safety across a variety of mental health conditions, demographics, and AI failure modes.
Optimizing telemental health with live artificial intelligence clinical scaffolding and feedback
This project aims to develop a clinical scaffolding system to enhance telemental health care by providing real-time coaching and actionable suggestions during video-based sessions. Modeled after live supervision methodologies, it supports clinicians by identifying intervention targets and offering text-based coaching prompts to guide care. Unlike automated chatbots, this approach enables clinicians to adapt suggestions to patient needs, balancing automation with oversight for safer AI-supported mental healthcare. The proposed in-session support will facilitate efficient implementation of strategies and clinician skill development. This project seeks to enhance data privacy by processing all data on-device and avoiding external data transfers.
Developing a hospital-based treatment engagement program for Alcohol Use Disorder
Alcohol use disorder (AUD) frequently results in serious illness, injuries, and hospitalizations. Surviving illness or injuries related to alcohol use can motivate behavior change that could be harnessed through treatment engagement for AUD in the hospital; however, in general hospital settings, patients are rarely presented with more than a piece of paper with phone numbers to call for help with their drinking. This project is focused on designing and evaluating a shared decision making approach for AUD treatment in hospitalized patients. We are interviewing people who are hospitalized with complications of AUD to better understand their unique needs and preferences. We are using the knowledge gained from interviews with patients to adapt a paper-based decision aid that was originally designed to help clinic patients think about changing their drinking, and creating an interactive web-based interface, tailored to hospitalized patients. We will then evaluate the use and effectiveness of this new online decision aid for engaging hospitalized patients in AUD treatment. The overarching goal of this research is to more effectively use hospitalizations to bridge individuals to long-term, potentially lifesaving AUD care.
Partnering with patients to re-envision psychiatric hospitalization and discharge
We will analyze people’s stories about psychiatric hospitalization, interview people with experiences surrounding psychiatric hospitalization, and co-design with them to identify alternative approaches that would help people care for themselves as they transition out of the hospital. We will build upon our prior work on understanding patients’ challenges and co-designing new systems that help patients transition from psychiatric hospitalizationto self-management. In particular, we will focus on how we could redesign psychiatric hospital systems with the people who have experienced them, identifying patient insights on the knowledge, resources, and self-efficacy they need to help them return to the community.
