Predicting treatment outcome for complex posttraumatic stress disorder using the personalized advantage index

Area of Science:

• Psychiatry
• Clinical Psychology
• Psychotherapy Research

Background:

• Psychotherapy is effective for posttraumatic stress disorder (PTSD), but individual outcomes vary significantly.
• Treatment options for complex posttraumatic stress disorder (CPTSD) require further investigation.
• Personalized treatment approaches are needed to optimize therapeutic benefits for CPTSD patients.

Purpose of the Study:

• To develop a Personalized Advantage Index (PAI) for CPTSD.
• To identify clinical and demographic predictors of treatment outcome in exposure therapy and skills training for CPTSD.
• To personalize psychological treatment selection for improved CPTSD outcomes.

Main Methods:

• Utilized data from a randomized controlled trial (RCT) of 92 CPTSD patients comparing Prolonged Exposure (PE) and Skills Training in Affective and Interpersonal Regulation (STAIR).
• Employed random forest and bootstrap methods to identify outcome predictors from 61 candidate variables.
• Calculated PAI to retrospectively determine optimal versus suboptimal treatment allocation using a leave-one-out cross-validation approach.

Main Results:

• Specific predictors of worse outcome in PE included somatoform dissociation, depression, suicidal ideation, and reduced physical health.
• Predictors of worse outcome in STAIR included interpersonal problems, PTSD symptom severity, intrusions, guilt, and psychoticism; witnessing trauma predicted better outcome.
• PAI-guided optimal treatment allocation yielded significantly larger improvements in clinician-assessed (Cohen's d=0.83) and moderate improvements in self-rated (Cohen's d=0.60) PTSD symptoms compared to suboptimal allocation.

Conclusions:

• The Personalized Advantage Index (PAI) shows promise for personalizing psychological treatment in CPTSD.
• PAI-guided treatment selection can enhance therapeutic benefits for individuals with CPTSD.
• Further research with larger samples and external validation is necessary to confirm PAI's efficacy.