Assessment of the accuracy of procalcitonin to diagnose postoperative infection after cardiac surgery.

BACKGROUND Cardiopulmonary bypass induces a nonspecific inflammatory response. Procalcitonin has been advocated as a specific biomarker for infection. The authors studied the accuracy of procalcitonin to diagnose postoperative infection after cardiac surgery and compared it with those of C-reactive protein, white blood cell count, and interleukins 6 and 8. METHODS The authors prospectively included 100 patients scheduled to undergo elective cardiac procedures with cardiopulmonary bypass. Blood samples were taken before surgery and each day over the 7-day postoperative period, and measurement of procalcitonin, C-reactive protein, white blood cell count, and interleukins 6 and 8 were performed. Diagnosis of infection was performed by a blinded expert panel. Data are expressed as value [95% confidence interval]. RESULTS Infection was diagnosed in 16 patients. Procalcitonin was significantly higher in infected patients, with a peak reached on the third postoperative day. Only the areas under the receiver operating curve of procalcitonin (0.88 [0.71-0.95]) and C-reactive protein (0.72 [0.58-0.82]) were significantly different from the no-discrimination curve, and that of procalcitonin was significantly different from those of C-reactive protein, white blood cell count, and interleukins 6 and 8. A procalcitonin value greater than 1.5 ng/ml beyond the second day diagnosed postoperative infection with a sensitivity of 0.93 [0.70-0.99] and a specificity of 0.80 [0.70-0.87]. Procalcitonin was significantly higher in patients who died (27.5 [1.65-40.5] vs. 1.2 [0.7-1.5] ng/ml; P < 0.001). CONCLUSION Procalcitonin is a valuable marker of bacterial infections after cardiac surgery.