A Clinical Prediction Model to Estimate Risk for 30-Day Adverse Events in Emergency Department Patients With Symptomatic Atrial Fibrillation

Study objective

Atrial fibrillation affects more than 2 million people in the United States and accounts for nearly 1% of emergency department (ED) visits. Physicians have little information to guide risk stratification of patients with symptomatic atrial fibrillation and admit more than 65%. Our aim is to assess whether data available in the ED management of symptomatic atrial fibrillation can estimate a patient's risk of experiencing a 30-day adverse event.

Methods

We systematically reviewed the electronic medical records of all ED patients presenting with symptomatic atrial fibrillation between August 2005 and July 2008. Predefined adverse outcomes included 30-day ED return visit, unscheduled hospitalization, cardiovascular complication, or death. We performed multivariable logistic regression to identify predictors of 30-day adverse events. The model was validated with 300 bootstrap replications.

Results

During the 3-year study period, 914 patients accounted for 1,228 ED visits. Eighty patients were excluded for non–atrial-fibrillation-related complaints and 2 patients had no follow-up recorded. Of 832 eligible patients, 216 (25.9%) experienced at least 1 of the 30-day adverse events. Increasing age (odds ratio [OR] 1.20 per decade; 95% confidence interval [CI] 1.06 to 1.36 per decade), complaint of dyspnea (OR 1.57; 95% CI 1.12 to 2.20), smokers (OR 2.35; 95% CI 1.47 to 3.76), inadequate ventricular rate control (OR 1.58; 95% CI 1.13 to 2.21), and patients receiving β-blockers (OR 1.44; 95% CI 1.02 to 2.04) were independently associated with higher risk for adverse events. C-index was 0.67.

Conclusion

In ED patients with symptomatic atrial fibrillation, increased age, inadequate ED ventricular rate control, dyspnea, smoking, and β-blocker treatment were associated with an increased risk of a 30-day adverse event.