Predicting outcomes after teens use crisis text support
Predicting point-of-care outcomes for text message crisis interventions in teens
This project uses computer programs to predict how teens who use a crisis text app will do, like whether they stay engaged with a counselor or are referred to emergency services.
Quick facts
| Grant type | R21 grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | Utah State Higher Education System--University of Utah NIH-funded |
| Lab location | 1 site (Salt Lake City, United States) |
| Project ID | NIH-11124650 on NIH RePORTER |
What this research studies
From your point of view, researchers will analyze a large, anonymized set of SafeUT text-message conversations (over 130,000 encounters and 2.3 million messages) to learn patterns linked to different outcomes. They will apply machine learning and natural language processing to the messages to build models that predict events such as referral to emergency services, whether users stay connected with counselors, and whether a full risk assessment is completed. The goal is to turn those models into real-time feedback tools that help counselors respond faster and more appropriately during a crisis. All message data will be rigorously anonymized and the work is meant to improve access to quality crisis support, including in rural areas.
Who could benefit from this research
Good fit: Teens and young adults who use the SafeUT crisis text app or other text-based crisis services, especially those experiencing suicidal thoughts or an acute mental health crisis, are the primary population tied to this work.
Not a fit: People who do not use text-based crisis services, lack access to a phone, or require immediate in-person emergency care may not benefit directly from these predictive tools.
Why it matters
Potential benefit: If successful, the tools could help counselors spot high-risk teens sooner and improve crisis triage and follow-up through real-time alerts and guidance.
How similar studies have performed: Some prior work has used AI to flag suicide risk from medical records or social media, but applying these methods to crisis text-message conversations is relatively new.
Where this research is happening
Salt Lake City, United States
- Utah State Higher Education System--University of Utah — Salt Lake City, United States (Active)
Researchers
- Principal investigator: Kious, Brent Michael — Utah State Higher Education System--University of Utah
- Study coordinator: Kious, Brent Michael
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 below.