Integrating predictive modeling into mental health care: An example in suicide prevention

Abstract

Recent advances in statistical methods and computing power have improved the ability to predict risks associated with mental illness with more efficiency and accuracy. However, integrating statistical prediction into a clinical setting poses new challenges that need creative solutions. A case example explores the challenges and innovations that emerged at a Department of Veterans Affairs hospital while implementing REACH VET (Recovery Engagement and Coordination for Health - Veterans Enhanced Treatment), a suicide prevention program that is based on a predictive model that identifies veterans at statistical risk for suicide.