ADAPT: Adaptive Decision support for Addiction Treatment

Yale University, New Haven CT

Investigators

Abstract

The opioid crisis continues to evolve, creating ongoing challenges for timely and effective care delivery. To support more consistent treatment practices, implementation research must adapt quickly to emerging evidence. Clinical decision support (CDS) offers a scalable strategy to accelerate the adoption of evidence-based practices across healthcare systems. Patients often become more receptive to treatment following key clinical encounters, such as an emergency department (ED) visit for opioid overdose. However, treatment initiation in these settings remains inconsistent. To address this gap, we conducted the EMBED pragmatic, cluster-randomized trial. This study evaluated a non-interruptive, electronic health record (EHR)-based CDS to support patient assessment and automate EHR workflows related to initiating buprenorphine during routine ED care for individuals diagnosed with opioid use disorder (OUD). The CDS increased the proportion of physicians who initiated buprenorphine, resulting in its national dissemination. Post-trial analyses revealed variation in treatment patterns and highlighted opportunities to expand reach and increase adoption. However, most CDS tools remain static during evaluation, and traditional assessment methods are resource-intensive and slow, limiting the ability to rapidly improve and implement effective interventions. These challenges must be addressed to accelerate the delivery of data-driven solutions for the opioid crisis, including the continued optimization of the EMBED CDS. The EHR environment not only enables CDS delivery but also provides a scalable, low-burden method for evaluating care processes using automated log data. Current CDS performance metrics, such as alert dismissal rates, are insufficient for assessing interface design, workflow integration, and real-time uptake. To address this gap, we will apply a Multiphase Optimization Strategy (MOST) framework using rapid, iterative randomized testing and pragmatic EHR-based metrics to pursue the following aims: (1) Refine and validate reproducible, scalable outcome measures for CDS uptake and usability related to ED-initiated buprenorphine for OUD, and (2) Refine and test a multicomponent CDS intervention to improve ED-initiation of buprenorphine through enhanced CDS adoption, interface usability, and system reach. Achieving these aims will establish a path toward scalable, adaptive interventions for improving treatment delivery in the context of the opioid crisis. Dr. Melnick and the ADAPT team bring extensive experience in emergency medicine, clinical decision support, implementation research, biostatistics, and health systems evaluation, positioning them to execute this work successfully.