Predictors of regional variations in hospitalizations following emergency department visits for atrial fibrillation

Abstract

The emergency department (ED) is often where atrial fibrillation (AF) is first detected and acutely treated and affected patients dispositioned. We used the Nationwide Emergency Department Sample to estimate the percentage of visits resulting in hospitalization and investigate associations between patient and hospital characteristics with hospitalization at the national and regional levels. We conducted a cross-sectional study of adults with AF listed as the primary ED diagnosis in the 2007 to 2009 Nationwide Emergency Department Sample. We performed multivariate logistic regression analyses investigating the associations between prespecified patient and hospital characteristics with hospitalization. From 2007 to 2009, there were 1,320,123 weighted ED visits for AF, with 69% hospitalized nationally. Mean regional hospitalization proportions were: Northeast (74%), Midwest (68%), South (74%), and West (57%). The highest odds ratios for predicting hospitalization were heart failure (3.85, 95% confidence interval [CI] 3.66 to 4.02), chronic obstructive pulmonary disease (2.47, 95% CI 2.34 to 2.61), and coronary artery disease (1.65, 95% CI 1.58 to 1.73). After adjusting for age, privately insured (0.77, 95% CI 0.73 to 0.81) and self-pay (0.77 95% CI 0.66 to 0.90) patients had lower odds compared with Medicare recipients, whereas Medicaid (1.21, 95% CI 1.11 to 1.32) patients tended to have higher odds. Patients living in low-income zip codes (1.18, 95% CI 1.12 to 1.25) and patients treated at large metropolitan hospitals (1.75, 95% CI 1.59 to 1.93) had higher odds. In conclusion, our analysis showed considerable regional variation in the management of patients with AF in the ED and in associations between patient socioeconomic and hospital characteristics with ED disposition; adapting best practices from among these variations in management could reduce hospitalizations and health-care expenses.

Figure 1

Hospitalizations by region and adult census age categories Figures are stratified by region with each dot representing a year and each line type an age group. Overall estimates for each region are also included.

Figure 2

ED cardioversions by adult census age category and region Figures are stratified by region and illustrate percentages calculated from weighted frequencies. Each dot represents a year and each line type an age group, including the overall estimate for each region ignoring age.