The use of preoperative B-type natriuretic peptide as a predictor of atrial fibrillation after thoracic surgery: systematic review and meta-analysis.

OBJECTIVE To determine whether elevated preoperative B-type natriuretic peptide (NP) measurements are an independent predictor of atrial fibrillation (AF) in patients having thoracic surgery. DESIGN Systematic review and meta-analysis. SETTING In-hospital and 30 days after thoracic surgery. PARTICIPANTS The 742 patients who participated in the 5 observational studies. INTERVENTIONS None. MEASUREMENTS AND MAIN RESULTS EMBASE, OVID Health Star, Ovid Medline, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and ProQuest Dissertations and Theses A&I databases were searched for all studies of noncardiac thoracic surgery patients in whom a preoperative NP was measured up to 1 month before surgery, and that measured the incidence of postoperative AF. Studies were included regardless of their language, sample size, publication status, or study design. Study quality was evaluated using the Newcastle Ottowa Scale. The combined incidence of postoperative AF was 14.5% (n=108/742), and the NP thresholds used to predict AF varied among studies. An elevated preoperative NP measurement was associated with an OR of 3.13 (95% CI 1.38-7.12; I2=87%) for postoperative AF, with the sensitivity analysis reporting an OR of 9.51 (95% CI 4.66-19.40; I2=0). CONCLUSION Patients with an elevated preoperative NP measurement are at an increased risk of postoperative AF. There may be value in incorporating NP measurement into existing AF risk prediction models.