This study refines and optimizes the EMBED* clinical decision support (CDS; see NCT03658642) to increase number of ED physicians following standard of care for the administration of buprenorphine to appropriate patients with opioid use disorder.

This study does not have open enrollment. Investigators will use a Multiphase Optimization STrategy (MOST) framework study with preparation, optimization, and confirmatory phases.

Optimization Phase: This phase has two stages. In stage 1, investigators will conduct a 2x2x2 factorial trial to expand EMBED is expanded to include sustainable implementation strategies: nurse prompt for withdrawal assessment, and targeted clinician prompts to use the CDS along with individualized patient resources to promote equity and motivate readiness to start treatment. In Stage 2, investigators will improve CDS usability via serial randomized testing to inform iterative refinement of the CDS interface and workflow to minimize user errors, task disruption, and abandonment through identification of specific targets for improvement via application of novel CDS outcome measures in serial randomized tests.

Evaluation phase: Investigators will conduct a randomized trial to compare the efficacy effectiveness of the optimized package compared to the original on ED-initiation of buprenorphine rates in patients with OUD.

*EMBED is a user user-centered, clinician clinician-facing clinical decision support system integrated into the electronic health record workflow to facilitate initiating buprenorphine in the emergency department by: diagnosing opioid use disorder with a checklist based on the diagnostic criteria of the DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, fifth edition), assessing the severity of withdrawal with the Clinical Opioid Withdrawal Scale (COWS), motivating patients to accept treatment with a scripted brief negotiation interview, and automating the electronic health record workflow, including clinical and after visit documentation, order entry, prescribing, and referral for ongoing treatment in the community

Computerized clinical decision support (CDS) is a scalable strategy to standardize evidence-

based practices and respond efficiently to the evolving operational demands of the opioid crisis.

Effective practice change requires CDS tools that are usable (easy to learn, efficient to navigate, support task completion with minimal errors, and provide a satisfactory user experience), well-

integrated into clinical workflows and supported by coordinated implementation strategies. We

will refine and evaluate the uptake, usability, and reach of a nationally disseminated multicomponent CDS intervention to improve treatment initiation during emergency care for individuals with opioid use disorder.

This study refines and optimizes the EMBED* clinical decision support (CDS; see NCT03658642) to increase the number of ED physicians following standard of care for the administration of buprenorphine to appropriate patients with opioid use disorder. This study does not have open enrollment.

Investigators will use a Multiphase Optimization Strategy (MOST) framework study with preparation, optimization, and confirmatory phases. Optimization Phases: This phase has two stages. In stage 1, investigators will conduct a 2x2x2 factorial trial to expand EMBED to include sustainable implementation strategies: nurse prompts for withdrawal assessment, and targeted clinician prompts to use the CDS along with individualized patient resources to promote equity and motivate readiness to start treatment. We will be randomizing these 3 strategies at the encounter level, resulting in these 8 arms:

Original EMBED

EMBED + nurse prompt

EMBED + clinician prompt

EMBED + patient facing materials

EMBED + nurse prompt + clinician prompt

EMBED + nurse prompt + patient facing materials

EMBED + clinician prompt + patient facing materials

EMBED + nurse prompt + clinician prompt + patient facing materials

In Stage 2, investigators will improve CDS usability via serial randomized testing to inform iterative refinement of the CDS interface and workflow to minimize user errors, task disruption, and abandonment. This will be done by identifying specific targets for improvement and applying novel CDS outcome measures in serial randomized tests.

Evaluation phase: In this confirmatory phase, investigators will conduct a randomized trial to compare the efficacy and effectiveness of the optimized package compared to the original on ED-initiation of buprenorphine rates in patients with OUD. During this phase, the optimized treatment package will be tested in a fully powered 2-arm randomized trial comparing the efficacy of the optimized, multicomponent intervention package to the original EMBED CDS on rates of ED initiation of buprenorphine in patients with OUD. Based on the results of the optimization phase factorial trial in conjunction with findings from the rapid, serial randomized testing, the optimized treatment package will include individual components found to increase CDS uptake and all usability and workflow improvements found to decrease errors and increase efficiency in serial randomized testing.

*EMBED is a user user-centered, clinician clinician-facing clinical decision support system integrated into the electronic health record workflow to facilitate initiating buprenorphine in the emergency department by: diagnosing opioid use disorder with a checklist based on the diagnostic criteria of the DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, fifth edition), assessing the severity of withdrawal with the Clinical Opioid Withdrawal Scale (COWS), motivating patients to accept treatment with a scripted brief negotiation interview, and automating the electronic health record workflow, including clinical and after visit documentation, order entry, prescribing, and referral for ongoing treatment in the community.

Inclusion Criteria:

• Emergency department patient
• 18 years of age or older
• Moderate to severe opioid use disorder

Exclusion Criteria:

• Under 18 years of age
• Pregnant
• Currently receiving medication for opioid use disorder

De-identified data related to EHR use measurements in various levels of aggregation. e.g., time spent on the CDS or threshold outcomes such as dismissal of the CDS without action under a specific amount of time or the time spent on CDS that exceeds what was expected based on the patient's clinical opioid withdrawal (COWS) score, time in the ED, or level of acuity, patient acuity, chief complaints or diagnoses, clinician, interruptions (such as orders started but not finished), task switching, seasonality, and variation in shift timing (such as nights and weekends). Data shared will include variables relating to patient demographics, medical history, prescription information, clinician facing clinical decision support (CDS) firing and clinician action information, and clinician prescribing decisions and historical prescribing

De-identified data will be shared with NAHDAP as soon as possible upon completion of quality control procedures, analysis, and at the time of associated publication or end of performance period. Data will be released by NAHDAP at the time of receipt and will be available indefinitely.

De-identified data sets used for analysis will be submitted to NAHDAP through the NIH HEAL Initiative and will be available for public use according to NAHDAP data sharing policies and restrictions set by our DUA with NAHDAP and our IRB. All data submissions to NAHDAP undergo confidentiality review to protect respondents' data from being re-identified.

Portions of the ADAPT intervention will be built for national distribution on the Epic EHR platform. Distribution of code that is proprietary to Epic may limit distribution of certain portions of software or code. In this situation, investigators will provide dashboards and visualizations of workflow diagrams as appropriate to disclose interface and workflow concepts to the broader scientific community These portions of the data will also be subject to restricted use access outside of organizations using Epic.