Presentation Type
Oral Presentation
Category
Social Sciences/Humanities
Abstract/Artist Statement
In the past decade, Veteran suicide rates increased by 34% in Oklahoma, a state with substantial medical provider shortages and related care barriers. Existing approaches to suicide prevention mainly consider “universal” risk factors identified through review of aggregated medical and death records. While useful, the universal approach fails to account for the unique presentations of risk factors that characterize individual Veterans’ experience. Approaches that examine individual risk profiles are needed for tailored safety and risk management. To meet this need, we developed PREVEnT: Personalized Veterans Protection, a mobile app Veterans can use to track their mental health. PREVEnT is equipped with a preliminary personalized machine learning algorithm to detect idiographic risk and deploy individualized suicide safety planning. This personal assistant system uses a smartphone to passively collect data, which prompts user interaction for safety promotion and engagement in acute mental health care, as appropriate. We are currently testing the PREVEnT app’s feasibility and its algorithm’s ability to accurately detect risk, trigger just-in-time interventions, and facilitate timely communication between Veterans’ mental and physical health providers during acute risk. Ultimately, we aim to harness individual-level data to avert deaths by suicide through streamlining collaborative healthcare. If our aim is realized, PREVEnT’s individualized and automated interventions will addresses the unique and varied clinical needs of Oklahoma Veterans, while also supporting the state’s overburdened healthcare system.
Mentor Name
Duncan Campbell
PREVEnT: Personalized Veterans Protection – Integrating Technology and Behavioral Health
UC 327
In the past decade, Veteran suicide rates increased by 34% in Oklahoma, a state with substantial medical provider shortages and related care barriers. Existing approaches to suicide prevention mainly consider “universal” risk factors identified through review of aggregated medical and death records. While useful, the universal approach fails to account for the unique presentations of risk factors that characterize individual Veterans’ experience. Approaches that examine individual risk profiles are needed for tailored safety and risk management. To meet this need, we developed PREVEnT: Personalized Veterans Protection, a mobile app Veterans can use to track their mental health. PREVEnT is equipped with a preliminary personalized machine learning algorithm to detect idiographic risk and deploy individualized suicide safety planning. This personal assistant system uses a smartphone to passively collect data, which prompts user interaction for safety promotion and engagement in acute mental health care, as appropriate. We are currently testing the PREVEnT app’s feasibility and its algorithm’s ability to accurately detect risk, trigger just-in-time interventions, and facilitate timely communication between Veterans’ mental and physical health providers during acute risk. Ultimately, we aim to harness individual-level data to avert deaths by suicide through streamlining collaborative healthcare. If our aim is realized, PREVEnT’s individualized and automated interventions will addresses the unique and varied clinical needs of Oklahoma Veterans, while also supporting the state’s overburdened healthcare system.