A deep learning method to detect opioid prescription and opioid use disorder from electronic health records. (March 2023)

• Record Type: Journal Article
• Authors: Kashyap, Aditya; Callison-Burch, Chris; Boland, Mary Regina
• Is Part Of: International journal of medical informatics. Volume 171(2023)
• Publication Date: 2023-03
• Subjects: Opioid -- Machine learning -- Electronic health records -- Data mining -- Natural language processing; Medical informatics -- Periodicals; Information science -- Periodicals; Computers -- Periodicals; Medical technology -- Periodicals; Medical Informatics -- Periodicals; Technology, Medical -- Periodicals; Computers; Information science; Medical informatics; Medical technology; Electronic journals; Periodicals; Electronic journals; 610.285
• DOI: 10.1016/j.ijmedinf.2022.104979
• Languages: English
• ISSNs: 1386-5056

Abstract

Highlights: We are the first to predict opioid prescribing behavior using both structured and unstructured EHR data in a deep learning model. Our deep learning model to predict Opioid Use Disorder (OUD) out-performed all prior models in the literature with an AUC-ROC of 0.94 ± 0.008. Modeling both opioid prescribing and OUD will be important to tackle the opioid epidemic effectively. Abstract: Objective: As the opioid epidemic continues across the United States, methods are needed to accurately and quickly identify patients at risk for opioid use disorder (OUD). The purpose of this study is to develop two predictive algorithms: one to predict opioid prescription and one to predict OUD. Materials and Methods: We developed an informatics algorithm that trains two deep learning models over patient Electronic Health Records (EHRs) using the MIMIC-III database. We utilize both the structured and unstructured parts of the EHR and show that it is possible to predict both challenging outcomes. Results: Our deep learning models incorporate elements from EHRs to predict opioid prescription with an F1-score of 0.88 ± 0.003 and an AUC-ROC of 0.93 ± 0.002. We also constructed a model to predict OUD diagnosis achieving an F1-score of 0.82 ± 0.05 and AUC-ROC of 0.94 ± 0.008. Discussion: Our model for OUD prediction outperformed prior algorithms for specificity, F1 score and AUC-ROC while achieving equivalent sensitivity. This demonstrates the importance of a) deep learning approaches in predicting OUD and b) incorporating both structured and unstructured data for this prediction task. No prediction models for opioid prescription as an outcome were found in the literature and therefore our model is the first to predict opioid prescribing behavior. Conclusion: Algorithms such as those described in this paper will become increasingly important to understand the drivers underlying this national epidemic.