RESUMEN
BACKGROUND AND OBJECTIVE: Postoperative cardiac complications pose a substantial risk to patients undergoing orthopaedic surgery. METHODS: B-type natriuretic peptide (BNP) was determined preoperatively in 270 patients undergoing scheduled orthopaedic surgery. The accuracy of BNP to predict the occurrence of in-hospital cardiac events was evaluated as the primary endpoint. Cardiac events at 1 year of follow-up were considered secondary endpoints. RESULTS: Preoperative BNP levels were significantly higher in the four patients experiencing in-hospital cardiac events than in patients without events [median 306 pg ml(-1) (range 123-3958) vs. 35 pg ml(-1) (range 14-2074), P = 0.01]. In a receiver operating characteristic analysis for the prediction of in-hospital cardiac events, the area under the receiver operating characteristic curve for BNP was 0.86 (95% confidence interval 0.74-0.99). The optimal predictive accuracy was achieved with a BNP threshold of 174 pg ml(-1). Importantly, the combination of BNP and the American Society of Anesthesiologists score further improved this accuracy. Additionally, BNP retained a high predictive accuracy in the subgroup of patients with known cardiac diseases [area under the receiver operating characteristic curve 0.85 (95% confidence interval from 0.65 to <1)]. The area under the receiver operating characteristic curve for the prediction of long-term cardiac events by BNP was 0.71 (95% confidence interval 0.57-0.84). CONCLUSION: In patients undergoing orthopaedic surgery, preoperative BNP levels can predict short-term and long-term postoperative cardiac events. Despite the paucity of endpoint events observed in this study, our results are in agreement with all prior investigations. BNP used in addition to an American College of Cardiology/American Heart Association guideline-based risk assessment might, therefore, be a useful tool in the preoperative evaluation of patients undergoing orthopaedic surgery.