Abstract
OBJECTIVES: Preoperative left atrial (LA) strain parameters measured by 2-dimensional speckle tracking echocardiography have been used to predict postoperative atrial fibrillation (POAF) after cardiac surgery. The aim of this meta-analysis was to determine whether preoperative LA strain parameters predict POAF after cardiac surgery. METHODS: PubMed, Embase, Cochrane database, and Google Scholar were searched manually until January 31, 2025. Studies where preoperative LA strain was used to predict POAF following cardiac surgery in adults were considered. Reviews, case series, case reports, and studies where patients were in preoperative atrial fibrillation were excluded. RESULTS: Twenty-four observational studies involving 2242 patients were included. Preoperative LA reservoir strain was significantly lower in patients with POAF vs those without POAF (standardized mean difference [SMD] -2.37; 95% confidence interval [CI] -3.87 to -0.88; I2 = 94.5%). Preoperative LA conduit (SMD -0.73; 95% CI, -1.06 to -0.39; I2 = 41.5%) and contraction (SMD -1.04; 95% CI, -1.81 to -0.27; I2 = 92.2%) strain were significantly lower in patients with POAF while preoperative LA reservoir, conduit, and contraction strain rates were not different in patients with POAF vs no POAF. Meta regression for heterogeneity in reservoir strain was significant for gender, vendor platform, and filling pressures (E/e'). The cut-off value of LA reservoir strain for predicting POAF was 22 to 25% (area under curve 0.69, specificity 0.679 [95% CI, 0.645 to 0.711], sensitivity 0.713 [95% CI, 0.675 to 0.743]). CONCLUSIONS: Preoperative LA reservoir, conduit, and contraction strain predict POAF in adults undergoing cardiac surgery. PROSPERO REGISTRATION NO: CRD42024606011.