Predicting Change in an Integrated Dual Diagnosis Substance Abuse Intensive Outpatient Program.

Area of Science:

• Clinical Psychology
• Addiction Medicine
• Psychiatry

Background:

• Routine outcome monitoring is underutilized in substance use treatment settings.
• Barriers exist in developing clinical decision support tools for substance use patients.
• There is a need for reliable methods to track patient progress in dual diagnosis treatment.

Purpose of the Study:

• To develop expected trajectories of change for substance use and depression.
• To demonstrate the treatment effectiveness of a dual diagnosis intensive outpatient program.
• To create and validate predictive equations for monitoring patient outcomes.

Main Methods:

• Developed predictive equation models for substance use and depression scores using derivation samples.
• Applied models to cross-validation samples to assess accuracy.
• Validated equations using Cohen's kappa, ROC curves, reliable change index, and percentage improvement.

Main Results:

• Large effect sizes reported for reductions in substance use (Cohen's d = .76) and depressive symptoms (d = 1.10).
• Developed predictive models achieved absolute accuracy rates from 95% to 100%.
• Validated predictive equations for monitoring substance use and depression severity.

Conclusions:

• Predictive equations for depressive symptoms and substance use days can be successfully derived and validated in dual diagnosis populations.
• These validated models can aid in real-time outcome monitoring within substance use treatment.
• The findings support the effectiveness of intensive outpatient programs for dual diagnosis patients.