Self-help groups and opioid use disorder treatment: An investigation using a machine learning-assisted robust causal...

Area of Science:

• Addiction Medicine
• Public Health
• Behavioral Science

Background:

• Opioid Use Disorder (OUD) presents significant public health challenges.
• Medication for Opioid Use Disorder (MOUD) treatment has suboptimal adherence and retention rates.
• Patient-level factors influencing treatment completion require further investigation.

Purpose of the Study:

• To investigate the impact of self-help group participation on MOUD treatment completion.
• To evaluate the causal relationship between self-help group involvement and treatment completion in OUD patients.
• To identify patient-level factors associated with improved MOUD treatment outcomes.

Main Methods:

• Utilized the 2015-2019 Treatment Episode Data Set: Discharges (TEDS-D).
• Employed machine learning models (Lasso, Decision Trees, Random Forest, XGBoost) for prediction.
• Applied Outcome Adaptive Elastic Net (OAENet) for confounder selection and robust McNemars test for causal inference.

Main Results:

• Machine learning models indicated a strong association between self-help group participation and treatment completion.
• Causal analysis revealed a significant average treatment effect on treated (ATT) of 0.260 (p < 0.0001).
• Self-help group participation positively impacts MOUD treatment completion rates.

Conclusions:

• Participation in self-help groups is crucial for MOUD treatment recipients.
• Combining MOUD with self-help groups increases the likelihood of treatment completion compared to MOUD alone.
• Policymakers should integrate self-help groups into OUD treatment strategies to boost adherence and completion.