Diagnostic and prognostic value of procalcitonin in patients with septic shock

OBJECTIVE: To determine whether procalcitonin is a reliable diagnostic and prognostic marker in septic shock compared with nonseptic shock. DESIGN: Prospective controlled trial. SETTING: Intensive care unit of the Avicenne Teaching Hospital, Bobigny, France. PATIENTS: All patients admitted to our intensive care unit over a 12-month period with clinical evidence of shock. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Echocardiography or pulmonary artery flotation catheter measurements were used to assess hemodynamics, and multiple specimens were obtained for microbiological studies. Standard criteria were used to diagnose septic shock. Serum concentrations of procalcitonin, C-reactive protein, and lactate were determined on the day of shock onset (day 1) and on days 3, 7, and 10. Seventy-five patients were included, 62 in the septic shock group and 13 in the cardiogenic shock group. Serum procalcitonin on day 1 was significantly higher in patients with than without septic shock (median, 14 [0.3-767] ng/mL vs. 1 [0.5-36] ng/mL, p < .01). A cutoff value of 1 ng/mL had 95% sensitivity and 54% specificity for separating patients with and without sepsis. C-reactive protein failed to discriminate between these two groups. Among patients with sepsis, procalcitonin concentrations were significantly higher in those who died than in the survivors, at all four measurement time points (median, 16 [0.15-767] ng/mL vs. 6 [0.2-123] ng/mL, p = .045 on day 1; 6.5 [0.3-135] ng/mL vs. 1.05 [0.11-53] ng/mL, p = .02 on day 10). A cutoff value of 6 ng/mL on day 1 separated patients who died from those who survived with 87.5% sensitivity and 45% specificity. C-reactive protein was not helpful for predicting mortality. Serum lactate was a nonspecific prognostic marker. CONCLUSIONS: These data indicate that procalcitonin may be a valuable early diagnostic and prognostic marker in patients with septic shock.