ABSTRACT
BACKGROUND: Preoperative N-terminal pro-BNP (NT-proBNP) is independently associated with adverse cardiac outcome but does not anticipate the dynamic consequences of anesthesia and surgery. The authors hypothesized that a single postoperative NT-proBNP level provides additional prognostic information for in-hospital and late cardiac events. METHODS: Two hundred eighteen patients scheduled to undergo vascular surgery were enrolled and followed up for 24-30 months. Logistic regression and Cox proportional hazards model were performed to evaluate predictors of in-hospital and long-term cardiac outcome. The optimal discriminatory level of preoperative and postoperative NT-proBNP was determined by receiver operating characteristic analysis. RESULTS: During a median follow-up of 826 days, 44 patients (20%) experienced 51 cardiac events. Perioperatively, median NT-proBNP increased from 215 to 557 pg/ml (interquartile range, 83/457 to 221/1178 pg/ml; P<0.001). The optimum discriminate threshold for preoperative and postoperative NT-proBNP was 280 pg/ml (95% confidence interval, 123-400) and 860 pg/ml (95% confidence interval, 556-1,054), respectively. Adjusted for age, previous myocardial infarction, preoperative fibrinogen, creatinine, high-sensitivity C-reactive protein, type, duration, and surgical complications, only postoperative NT-proBNP remained significantly associated with in-hospital (adjusted hazard ratio, 19.8; 95% confidence interval, 3.4-115) and long-term cardiac outcome (adjusted hazard ratio, 4.88; 95% confidence interval, 2.43-9.81). CONCLUSION: A single postoperative NT-proBNP determination provides important additional prognostic information to preoperative levels and may support therapeutic decisions to prevent subsequent structural myocardial damage.
