Objectives: Prediction of atrial fibrillation (AF) after coronary artery bypass grafting (CABG) may lead to preventive or early treatment and improved outcome. We investigated the association of serial perioperative cardiac troponin T (cTNT) measurements with postoperative AF in patients undergoing CABG.
Methods: In a retrospective analysis of prospectively collected data, 3148 patients undergoing elective CABG were evaluated. cTNT values were routinely determined before the start of surgery (cTNT0), at arrival on the intensive care unit (cTNT1) and 8-12 h later (cTNT2). Measurement of cTNT was continued until the peak value was reached. The development of AF during hospital stay was scored. The association between cTNT (cTNT0, cTNT1, cTNT2 and cTNTmax in first 48 h) and AF was calculated in univariable and multivariable analysis.
Results: AF occurred in 1080 (34%) patients. cTNT0, cTNT2 and cTNTmax were significantly and positively associated with postoperative AF (P < 0.001) in a univariable analysis, whereas a trend was seen for cTNT1 (P = 0.051). Advanced age, inotropic support and postoperative infection were independently associated with postoperative AF after logistic regression analysis, but cTNT was not. Categorizing patients by inotropic support into categories of inotropic support duration (none, <48 h, >48 h), the mean cTNT values were significantly higher among patients with AF in each category (all P < 0.001). Perioperative cTNT was significantly higher in patients with postoperative complications, longer hospital stay and reduced in-hospital survival.
Conclusions: Perioperative cTNT is univariably associated with postoperative AF after CABG, but not independently. Further, no clinically useful cut-off point for preventive or early treatment could be identified. Both perioperative cTNT and postoperative AF are associated with negative outcome and prolonged hospital stay.
Keywords: Arrhythmia; Atrial fibrillation; Coronary artery bypass graft; Outcome; Troponin.