Study objective: The high volume of patients treated in an emergency department (ED) for atrial fibrillation is predicted to increase significantly in the next few decades. Currently, 11% of these patients die within a year. We sought to derive and validate a complex model and a simplified model that predicts mortality in ED patients with atrial fibrillation.
Methods: This population-based, retrospective cohort study included 3,510 adult patients with a primary diagnosis of atrial fibrillation who were treated at 24 hospital EDs in Ontario, Canada, between April 2008 and March 2009. The main outcome was 30-day all-cause mortality.
Results: In the derivation cohort (n=2,343; mean age 68.8 years), 2.6% of patients died within 30 days of the ED visit versus 2.7% in the validation cohort (n=1,167; mean age 68.3 years). Variables associated with mortality in the complex model included age, presenting pulse rate and systolic blood pressure, presence of chest pain, 2 laboratory results (positive troponin result and creatinine level greater than 200 μmol [2.26 mg/dL]), 4 comorbidities (smoking, chronic obstructive pulmonary disease, cancer, and dementia), an increased bleeding risk, and a second acute ED diagnosis (in addition to atrial fibrillation). Observed 30-day mortality in the 5 risk strata that were defined by the predicted probability of death were 0.44%, 0.41%, 0.23%, 1.61%, and 10.3%. The c statistics were 0.88 and 0.87 in the derivation and validation cohorts, respectively. The a priori-selected 6-variable model, TrOPs-BAC, included a positive Troponin result, Other acute ED diagnosis, Pulmonary disease (chronic obstructive pulmonary disease), Bleeding risk, Aged 75 years or older, and Congestive heart failure. The c statistic for the simplified model was 0.81 in both the derivation and validation cohorts.
Conclusion: Using a population-based sample, we derived and validated both a complex and a simplified instrument that predicts mortality after an emergency visit for atrial fibrillation. These may aid clinicians in identifying high-risk patients for hospitalization while safely discharging more patients home.
Copyright © 2015 American College of Emergency Physicians. Published by Elsevier Inc. All rights reserved.