AI-powered predictions to prevent deaths and rehospitalizations from substance misuse
Official title: Using Multimodal Data Science To Gain Actionable Insights For Substance Misuse Prevention
Funded at University of Wisconsin-Madison · NIH-11556682 · Data updated 2026-11-12
A multimodal AI model will forecast who is at risk of dying or needing another hospital visit after discharge among people who misuse substances, using a large, integrated data platform.
Quick facts
| Grant type | NIH-funded research |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of Wisconsin-Madison NIH-funded |
| Lab location | 1 site (Madison, United States) |
| Project start | September 1, 2025 |
| Project end | August 31, 2027 |
| Funding year | 2026 |
| Award (this year) | $1,408,536 |
| NIH project number | 3OT2OD038002-01S1 |
| Project ID | NIH-11556682 on NIH RePORTER |
Who could benefit from this research
✓Could help
Individuals with opioid, non-opioid illicit, or alcohol misuse whose care involves hospital discharge and follow-up planning.
✕Less likely to help
Patients without a history of substance misuse or those not undergoing discharge planning may not directly benefit.
What this research studies
This project builds on a data-rich system that brings together electronic health records, claims, census data, and other sources to understand substance misuse patterns. Researchers will develop a multimodal AI tool to predict post-discharge death or rehospitalization and test it in diverse urban and rural populations. The work also includes engaging stakeholders to shape how the model is used in real-world settings to improve prevention and care access. The goal is to translate complex data into actionable insights that help clinicians and care teams intervene earlier for high-risk individuals.
Why it matters
Potential benefit: If successful, clinicians could identify high-risk patients sooner and tailor interventions to prevent deaths and unnecessary hospitalizations.
How similar studies have gone: Previous studies using multimodal data and AI for health-risk prediction have shown promise, but applying this approach to substance misuse outcomes is emerging and novel.
Where this research is happening
-
University of Wisconsin-Madison
Madison, United States Active
Researchers
- Principal investigator: Afshar, Majid — University of Wisconsin-Madison
- Study coordinator: Afshar, Majid
Related research
Other NIH-funded projects on the same topics.
-
BRAVE Innovation Center for Viral Resistance in Chickens
A center at UW–Madison aims to create chickens genetically resistant to avian viruses like H5N1 by identifying host genes involved in virus replication using genome-wide CRISPR screening.
-
Center to Fight Vector-Borne Diseases in the Midwest
A Midwest program brings researchers, public health teams, and vector-control groups together to prevent diseases spread by ticks and mosquitoes.
-
Globally useful predictor for cervical cancer treatment response
This project aims to develop a simple predictive signature to personalize treatment for locally advanced cervical cancer using easy tests and imaging.
-
Neutrophil development with stem cells to understand their behavior
Researchers are studying how neutrophils, a type of immune cell, develop from stem cells to better understand why they can help or harm in infections and inflammatory diseases.
-
WREN: Expanding Research Capacity in Rural Primary Care Clinics
A nationwide network in Wisconsin will strengthen rural primary care clinics to participate in real-world research and bring more clinical trials to communities.
-
Targeting early oral GVHD to improve mouth health after stem cell transplant
A study at UW–Madison will explore why some transplant patients develop painful oral GVHD and test ideas to prevent and lessen mouth-related problems.
About this research
- This is an active NIH-funded research project — typically early-stage science, not a clinical trial accepting patient enrollment.
- Some NIH-funded labs run parallel clinical studies or seek volunteers for related work. To check, contact the principal investigator or institution listed above.
- For full project details, budget, and progress reports, visit the official NIH RePORTER page.