Using the Personalized Advantage Index to determine which veterans may benefit from more vs. less comprehensive...

Area of Science:

• Psychiatry and Behavioral Science
• Machine Learning Applications in Healthcare
• Clinical Psychology

Background:

• Intensive PTSD treatment programs (ITPs) are effective but vary in duration and adjunctive services.
• Understanding individual treatment response to varying ITP comprehensiveness is crucial.
• Current knowledge gaps exist regarding differential outcomes based on ITP intensity.

Purpose of the Study:

• To develop and validate a machine learning model, the Personalized Advantage Index (PAI).
• To predict individual treatment response to more versus less comprehensive ITPs.
• To utilize clinical and demographic factors for personalized ITP recommendations.

Main Methods:

• Developed and tested the PAI on 747 veterans diagnosed with PTSD.
• Compared outcomes between a 3-week (more comprehensive) and a 2-week (less comprehensive) ITP.
• Utilized PTSD Checklist for DSM-5 for symptom change assessment.

Main Results:

• Approximately 12.32% of participants showed a predicted differential response to ITP intensity.
• For the top 25% of PAI scores, the effect size for symptom change between optimal and non-optimal programs was d=0.35.
• No substantial difference in predicted outcomes for the majority of individuals.

Conclusions:

• A minority of individuals may benefit more from a specific ITP intensity.
• Less comprehensive, more financially sustainable ITPs appear effective for most individuals with PTSD.
• PAI model shows potential for guiding personalized treatment selection in PTSD care.