A longitudinal observational study with ecological momentary assessment and deep learning to predict non-prescribed opioid use, treatment retention, and medication nonadherence among persons receiving medication treatment for opioid use disorder

Abstract

Background

Despite effective treatments for opioid use disorder (OUD), relapse and treatment drop-out diminish their efficacy, increasing the risks of adverse outcomes, including death. Predicting important outcomes, including non-prescribed opioid use (NPOU) and treatment discontinuation among persons receiving medications for OUD (MOUD) can provide a proactive approach to these challenges. Our study uses ecological momentary assessment (EMA) and deep learning to predict momentary NPOU, medication nonadherence, and treatment retention in MOUD patients.

Methods

Study participants included adults receiving MOUD at a large outpatient treatment program. We predicted NPOU (EMA-based), medication nonadherence (Electronic Health Record [EHR]- and EMA-based), and treatment retention (EHR-based) using context-sensitive EMAs (e.g., stress, pain, social setting). We used recurrent deep learning models with 7-day sliding windows to predict the next-day outcomes, using Area Under the ROC Curve (AUC) for assessment. We employed SHapley additive ExPlanations (SHAP) to understand feature latency and importance.

Results

Participants comprised 62 adults with 14,322 observations. Model performance varied across EMA subtypes and outcomes with AUCs spanning 0.58–0.97. Recent substance use was the best performing predictor for EMA-based NPOU (AUC = 0.97). Life-contextual factors were best performers for EMA-based medication nonadherence (AUC = 0.68) and retention (AUC = 0.89), and substance use risk factors (e.g., nicotine and alcohol use) and self-reported MOUD adherence performed best for predicting EHR-based medication nonadherence (AUC = 0.79). SHAP revealed varying latencies between predictors and outcomes.

Conclusions

Findings support the effectiveness of EMA and deep learning for forecasting actionable outcomes in persons receiving MOUD. These insights will enable the development of personalized dynamic risk profiles and just-in-time adaptive interventions (JITAIs) to mitigate high-risk OUD outcomes.