Risk Scores for Prediction of Postoperative Atrial Fibrillation After Cardiac Surgery: A Systematic Review and Meta-Analysis

Abstract

Postoperative atrial fibrillation (POAF) is a common complication after cardiac surgery and is associated with poor clinical outcomes. The objective of this systematic review and meta-analysis was to assess the performance of risk scores to predict POAF in cardiac surgery patients. We searched MEDLINE, Embase, and Cochrane CENTRAL for studies that developed/evaluated a POAF risk prediction model. Pairs of reviewers independently screened studies and extracted data. We pooled area under the receiver operating curves (AUCs), sensitivity and specificity, and adjusted odds ratios from multivariable regression analyses using the generic inverse variance method and random effects models. Forty-three studies (n=63,847) were included in the quantitative synthesis. Most scores were originally developed for other purposes but evaluated for predicting POAF. Pooled AUC revealed moderate POAF discrimination for the EuroSCORE II (AUC 0.59, 95% confidence interval [CI] 0.54 to 0.65), Society of Thoracic Surgeons (AUC 0.60, 95% CI 0.56 to 0.63), EuroSCORE (AUC 0.63, 95% CI 0.58 to 0.68), CHADS2 (AUC 0.66, 95% CI 0.57 to 0.75), POAF Score (AUC 0.66, 95% CI 0.63 to 0.68), HATCH (AUC 0.67, 95% CI 0.57 to 0.75), CHA2DS2-VASc (AUC 0.68, 95% CI 0.60 to 0.75) and SYNTAX scores (AUC 0.74, 95% CI 0.71 to 0.78). Pooled analyses at specific cutoffs of the CHA2DS2-VASc, CHADS2, HATCH, and POAF scores demonstrated moderate-to-high sensitivity (range 46% to 87%) and low-to-moderate specificity (range 31% to 70%) for POAF prediction. In conclusion, existing clinical risk scores offer at best moderate prediction for POAF after cardiac surgery. Better models are needed to guide POAF risk stratification in cardiac surgery patients.