ABSTRACT
BACKGROUND: Single clinical parameters are inaccurate for diagnostic and prognostic estimation in patients with syncope. The cardiac marker NT-pro-BNP has not thoroughly been evaluated for this application. METHODS: NT-pro-BNP was assessed in 161 consecutive patients (median age 69 years, 58% male) hospitalized for syncope in a cardiological university department and association (odds ratio: OR, 95% confidence interval: CI) with diagnosis of cardiac cause and 6-months outcome was analyzed. RESULTS: NT-pro-BNP levels were significantly higher in patients with cardiac (n=78) compared to non-cardiac syncope (n=83). At a cutoff of 156 pg/ml, NT-pro-BNP showed a sensitivity of 89.7%, a specificity of 51.8% and a negative predictive value of 84.3% for the diagnosis of cardiac syncope. Increasing NT-pro-BNP was a significant predictor of cardiac syncope (OR 3.7, 95% CI 2.3-5.8 per standard deviation of Log NT-pro-BNP, p<0.001) and addition of NT-pro-BNP significantly improved a predictive model including heart rate, history of structural heart disease and abnormal ECG. Adding left-ventricular ejection fraction to the model did not change results. Sixty-three patients had an adverse event during hospitalization or 6-months follow-up. NT-pro-BNP>156 pg/ml significantly predicted an adverse outcome (OR 2.7, 95% CI 1.04-6.9, p=0.04) after multivariate adjustment. CONCLUSIONS: In patients hospitalized for syncope, NT-pro-BNP was a strong and independent diagnostic and prognostic marker and addition to conventional criteria of history and examination improved the discriminatory performance. Randomized trials must clarify the benefit and position of NT-pro-BNP in the management algorithm of patients with syncope.
