Abstract

Objective: To evaluate the predictive value of brain natriuretic peptide (BNP) for assessment of cardiac risk before non-cardiac surgery.

Methods: Consecutively treated patients (947 men, 643 women) whose BNP was measured before non-cardiac surgery were studied. Clinical and ECG variables were evaluated to identify predictors of postoperative cardiac events.

Results: Events occurred in 6% of patients: 21 cardiac deaths, 20 non-fatal myocardial infarctions, 41 episodes of pulmonary oedema and 14 patients with ventricular tachycardia. All of these patients had raised plasma BNP concentrations (best cut-off point 189 pg/ml). The only independent predictor of postoperative events was BNP (odds ratio 34.52, 95% confidence interval (CI) 17.08 to 68.62, p < 0.0001). Clinical variables of Goldman’s multifactorial index identified 18% of patients in class I, 40% in class II, 24% in class III and 18% in class IV preoperatively; postoperative event rates were 2%, 3%, 7% and 14%, respectively. BNP identified 60% of patients as having zero risk (BNP 0–100 pg/ml), 22% low risk (101–200 pg/ml), 14% intermediate risk (201–300 pg/ml) and 4% high risk (> 300 pg/ml); postoperative event rates were 0%, 5%, 12% and 81%, respectively.

Conclusions: In this population of patients evaluated before non-cardiac surgery, BNP is an independent predictor of postoperative cardiac events. BNP > 189 pg/ml identified patients at highest risk.