Abstract Body (Do not enter title and authors here): INTRODUCTION: Atrial fibrillation (AF), often diagnosed in the Emergency Department (ED), increases the risk of stroke by 64% but can be mitigated by guideline-directed oral anticoagulant (OAC) treatment.
RESEARCH QUESTION: Does clinician education of an AF clinical decision support (CDS) tool increase OAC prescription rates for patients with new-onset AF or paroxysmal AF (pAF)?
GOALS: To increase guideline recommended care practice of OAC prescribing for appropriate candidates in the ED.
METHODS: This multi-center, multi-phase, stepped-wedge cluster-randomized clinical trial includes a retrospective phase (Phase 0) and implementation of an external browser link to the CDS in the EHR (Phase 1) which included patient-specific stroke and bleeding risk stratification scores with recommendation for OAC based on professional society guidelines, along with targeted provider education on how to use the resource. Data was collected from ED visits of patients aged > 18 with a primary AF/pAF diagnosis from Jan 1, 2020–May 25, 2024. The three sites, one academic and two community hospitals, implemented Phase 1 in January ‘22, April ‘22, and April ‘23. Eligibility for OAC treatment was based on clinical judgment weighing the benefit of stroke prevention against the risk of bleeding. A logistic GEE model assessed the intervention's impact on OAC prescribing and treated ED providers as a random effect to account for clustering. The model included the CDS intervention and covariates for site, cardiology consultation, guideline citation, and patient factors such as sex, race, ethnicity, and age.
RESULTS: Of 4397 patients analyzed, 655 met inclusion criteria, among which 296 (45.19%, median age 68) were prescribed OAC, 175 in Phase 0, and 121 in Phase 1. The odds ratio estimates and their corresponding 95% confidence intervals are as follows: CDS intervention 1.75 [1.20, 2.54], sex female vs male 2.31 [1.57, 3.38], age 0.27 [0.12, 0.61], site 0.55 [0.29, 1.04] and 0.45 [0.25, 0.83] for each community site vs academic, cardiology consultation 4.56 [2.80, 7.42], and guideline citation 2.23 [1.44, 3.43]. Race and ethnicity did not show significant associations.
CONCLUSION(S): We demonstrated that clinician education and limited CDS tool integration into the EHR are linked to an almost two-fold increase in OAC prescribing for suitable AF/pAF patients. Future studies should evaluate the influence of further CDS integration on OAC prescribing, long-term adherence, patient outcomes.