Related Experiment Video
Updated: Aug 5, 2026
07:31
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Machine Learning Model for Predicting Suicide Risks Among Patients With Posttraumatic Stress Disorder Who Received
Shu Huang1, Amie J Goodin1,2, Jill A Star3
1Department of Pharmaceutical Outcomes and Policy, College of Pharmacy, University of Florida, Gainesville 32611, Florida, USA, ufl.edu.
Depression and Anxiety
|July 31, 2026
Summary
A machine learning suicide prediction model (SPM) effectively identifies posttraumatic stress disorder (PTSD) patients using opioids who are at risk of suicide. This tool aids in prioritizing interventions for high-risk individuals.
Area of Science:
- Clinical Informatics
- Psychiatry
- Machine Learning
Background:
- Co-occurring posttraumatic stress disorder (PTSD) and opioid use significantly elevate suicide risk.
- There is a critical need for accurate tools to identify at-risk individuals within this vulnerable population.
Purpose of the Study:
- To develop and validate a machine learning-based suicide prediction model (SPM).
- To identify patients with PTSD prescribed opioids who are at risk of suicide within 6-month intervals.
Main Methods:
- Utilized 2016-2021 OneFlorida+ data, comparing LASSO, GBM, RF, and DNN models.
- Developed an SPM to predict 6-month suicide risk in adult PTSD patients on opioids.
- Validated model performance on internal (2016-18) and external (2019-21) cohorts.
Main Results:
- The Random Forest (RF) model demonstrated strong predictive performance.
- Internal validation (2016-18) showed a C-statistic of 83.6% and accuracy of 86.2%.
- External validation (2019-21) yielded similar results with a C-statistic of 83.6% and accuracy of 87.0%.
Conclusions:
- The developed SPM is a feasible tool for identifying patients at high risk of suicide.
- This model can aid in prioritizing preventive interventions and optimizing resource allocation.
- Effective identification allows for more efficient provider follow-up for at-risk patients.