Original Research J Clin Psychiatry June 2026

Predictive Modeling of Suicidal Ideation in US Veterans: A Head-to-Head Comparison of the Patient Health Questionnaire-2 Versus Complex Machine Learning Approaches

Full Article Read the complete peer-reviewed article in J Clin Psychiatry. Clinical Summary Veterans face disproportionately high rates of suicidal ideation and behavior, yet screening tools must balance missing high-risk patients against overwhelming systems with false positives. This study asks a practical question for clinicians: do complex machine learning models meaningfully outperform a simple PHQ-2 sum score when predicting past-year suicidal ideation? FAQ Did complex machine learning models predict suicidal ideation in veterans better than the PHQ-2? 10 questions
Key Takeaways In the 32-variable models, discrimination improved to 0.85–0.88 versus 0.79 for the PHQ-2, but this came with low PPVs of 0.23 to 0.27, indicating that better AUC did not translate into efficient identification of true positives. 6 takeaways Clinical Guide How should clinicians choose between a simple PHQ-2 approach and more complex prediction models when screening veterans for suicidal ideation? 8 steps
Physicians Postgraduate Press, Inc. (PPP) makes no warranties about the accuracy or completeness of any information published in The Journal of Clinical Psychiatry or other PPP materials, and disclaims liability for any use or non-use of that information. Clinicians should not rely solely on these materials and should exercise their own professional judgment when making patient care decisions on an individualized basis.

Learn More About Psychedelics in Psychiatry

Explore the Collection
Neural network illustration symbolizing psychedelic research advancements

Subscribe to read
more articles

Subscribe