Creating a national registry of patients presenting to emergency departments with opioid poisoning to inform clinical practice: the CEDRN opioid registry

Background Opioid poisoning is a leading cause of preventable mortality in North America. Emergency departments are open 24/7 and frequently provide care to people who use illicit opioids. While emergency departments offer critical opportunities to reverse drug toxicity, mitigate harm, and initiate longer-term treatment, there is limited high-quality evidence to guide emergency care for people presenting with opioid poisoning and/or use disorder. Our objective is to establish a national registry of patients presenting to emergency departments with opioid poisoning and produce timely evidence to inform clinical care, public health surveillance and policy. Methods We will use the Canadian Emergency Department Research Network (CEDRN) data sharing and membership agreements, infrastructure and expertise to create the CEDRN Opioid Registry to collect data on consecutive patients of all ages presenting to participating emergency departments with suspected opioid poisoning. A waiver of informed consent for registry inclusion will ensure absence of volunteer bias. We will automate screening and data extraction from electronic health records. We will capture data from emergency medical services, emergency departments, and in-hospital care, including demographics (e.g., age, sex at birth, housing), presentation characteristics (e.g., vital signs, chief complaint), investigations (e.g., urine toxicology), treatments (e.g., medications), harm-reduction interventions (e.g., take-home naloxone), and consultations (e.g., physician, social work, peer support). We will collect short-term outcomes (e.g., disposition) and link data with administrative sources to ascertain retention on opioid agonist therapy, revisits, admissions and mortality. We will verify data quality and validate definitions by comparing automatically extracted electronic health record data with manually extracted data. We will measure the inter-rater reliability of data elements and the diagnostic performance of discharge diagnoses. We anticipate using mixed-effects models with splines to examine associations between treatments and outcomes, and regression models, internal bootstrap validation, and nomograms to derive clinical decision rules to predict mortality. Our data can also be used for public health surveillance. Conclusions The CEDRN Opioid Registry will capture harmonized clinical data describing emergency care for patients presenting with opioid poisoning and allow us to create real-world timely evidence to inform clinical care, public health surveillance and policy. Trial registration NCT07466784 (20 February 2026).