R Proietti1, C Labos2, A AlTurki2, V Essebag2, T V Glotzer3, A Verma4. 1. a Cardiology Department , Luigi Sacco Hospital , Milan , Italy. 2. b McGill University Health Centre , Montreal , Quebec , Canada. 3. c Hackensack University Medical Center , Hackensack , NJ , USA. 4. d Heart Rhythm Program , Southlake Regional Health Centre , Newmarket , Canada.
Abstract
BACKGROUND: Contributory evidence on a direct association between asymptomatic atrial fibrillation (AF) burden and thromboembolic events is conflicting and contradictory. The aim of the article is to gather evidence available for a direct correlation between burden and stroke. METHODS: A literature search was performed to capture studies reporting data on the impact of asymptomatic AF burden on the risk of stroke. Data was then extracted from each included study including burden of AF, hazard ratio (HR) for stroke, and CHADS2 score. A random effects meta-analysis was carried out on the log-transformed HRs for different subgroups of AF burden. A meta-regression was performed on the two variables: burden of asymptomatic AF and CHADS2 score. RESULTS: The random-effect pooled analysis performed on a single subgroup of the six studies reporting data on HR, showed a HR of 2.150 (95% CI 1.523-3.003) for stroke during asymptomatic AF compared to sinus rhythm. At univariate meta-regression, no correlation was detected between burden of asymptomatic AF and HR for stroke (p-value 0,874). When CHADS2 score was included in the regression model as a covariate, no significant association was detected (p-value 0,939). CONCLUSION: A direct correlation between burden of asymptomatic AF and HR for stroke cannot be detected in our pooled analysis. However, due to the limitations acknowledged in the analysis, our findings need to be confirmed in large cohort studies.
BACKGROUND: Contributory evidence on a direct association between asymptomatic atrial fibrillation (AF) burden and thromboembolic events is conflicting and contradictory. The aim of the article is to gather evidence available for a direct correlation between burden and stroke. METHODS: A literature search was performed to capture studies reporting data on the impact of asymptomatic AF burden on the risk of stroke. Data was then extracted from each included study including burden of AF, hazard ratio (HR) for stroke, and CHADS2 score. A random effects meta-analysis was carried out on the log-transformed HRs for different subgroups of AF burden. A meta-regression was performed on the two variables: burden of asymptomatic AF and CHADS2 score. RESULTS: The random-effect pooled analysis performed on a single subgroup of the six studies reporting data on HR, showed a HR of 2.150 (95% CI 1.523-3.003) for stroke during asymptomatic AF compared to sinus rhythm. At univariate meta-regression, no correlation was detected between burden of asymptomatic AF and HR for stroke (p-value 0,874). When CHADS2 score was included in the regression model as a covariate, no significant association was detected (p-value 0,939). CONCLUSION: A direct correlation between burden of asymptomatic AF and HR for stroke cannot be detected in our pooled analysis. However, due to the limitations acknowledged in the analysis, our findings need to be confirmed in large cohort studies.