Mean corpuscular volume and red cell distribution width as predictors of left atrial stasis in patients with non-valvular atrial fibrillation.

BACKGROUND The role of erythrocyte indexes for the prediction of left atrial stasis, assessed by transesophageal echocardiography in patients with non-valvular atrial fibrillation, has not been previously clarified. METHODS Single center cross-sectional study comprising 247 consecutive patients admitted to the emergency department due to symptomatic atrial fibrillation and undergoing transesophageal echocardiogram evaluation for exclusion of left atrial appendage thrombus (LAAT) before cardioversion. All patients had a complete blood count performed up to 12 hours prior to the transesophageal echocardiogram. Markers of left atrial stasis were sought: LAAT, dense spontaneous echocardiographic contrast (DSEC) and low flow velocities (LFV) in the left atrial appendage. Erythrocyte indexes' accuracy for detecting transesophageal echocardiogram changes was evaluated through receiver operating curve analysis. Binary logistic multivariate analysis, using solely erythrocyte indexes and in combination with other variables (i.e. CHADS2, CHA2DS2VASc classifications and left ventricle ejection fraction), was used for transesophageal echocardiogram endpoints prediction. RESULTS LAAT was found in 8.5%, DSEC in 26.1% and LFV in 12.1%. Mean corpuscular volume and red cell distribution width were independent predictors of LAAT and DSEC. Despite adding incremental predictive value to each other, when clinical risk factors from CHADS2 and CHA2DS2VASc classifications and left ventricle ejection fraction were added to the models, only mean corpuscular volume remained an independent predictor of LAAT and DSEC. CONCLUSIONS These findings suggest that mean corpuscular volume and red cell distribution width may be linked to left atrial stasis markers.