Background: The objective of this study was to determine whether measurement of B-type natriuretic peptide (BNP) concentration before operation could be used to predict perioperative cardiac morbidity.
Methods: A prospective derivation study was performed in high-risk patients undergoing major non-cardiac surgery, with a subsequent validation study. A venous blood sample was taken the day before surgery for measurement of plasma BNP concentration. Screening for cardiac events (non-fatal myocardial infarction and cardiac death) was performed using clinical criteria, cardiac troponin I analysis and serial electrocardiography.
Results: Forty-one patients were recruited to the derivation cohort and 149 to the validation cohort. In the derivation cohort, the median (interquartile range) BNP concentration in the 11 patients who had a postoperative cardiac event was 210 (165-380) pg/ml, compared with 34.5 (14-70) pg/ml in those with no cardiac complications (P < 0.001). In the validation cohort, the median BNP concentration in the 15 patients who had a cardiac event was 351 (127-1034) pg/ml, compared with 30.5 (11-79.5) pg/ml in the remainder (P < 0.001). BNP concentration remained a significant outcome predictor in multivariable analysis (P < 0.001). Using receiver-operator curve analysis it was calculated that a BNP concentration of 108.5 pg/ml best predicted the likelihood of cardiac events, with a sensitivity and specificity of 87 per cent each.
Conclusion: Preoperative serum BNP concentration predicted postoperative cardiac events in patients undergoing major non-cardiac surgery independently of other risk factors.
Copyright (c) 2007 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.