Predictive value of procalcitonin, interleukin-6, and C-reactive protein for survival in postoperative patients with severe sepsis

Purpose: To prospectively evaluate the performance of procalcitonin (PCT), interleukin-6 (IL-6), and C-reactive protein (CRP) as percentage of baseline (FOB) in predicting hospital survival, we studied 64 consecutive, postoperative patients with severe sepsis. Materials and Methods: Plasma PCT, IL-6, and CRP were serially measured from day 1 (onset of sepsis) to day 14 in parallel with clinical data until day 28. Multivariate logistic regression and univariate analysis of predictive accuracy of PCT-, IL-6-, and CRP-POB were performed. Newly derived binary prediction rules were evaluated by calculating sensitivity, specificity, positive predictive value, and negative predictive value. Results: In survivors, PCT and IL-6 significantly decreased from days 1 to 14, whereas CRP did not. In nonsurvivors, the inflammation markers mostly increased within the second week. At day 7, logistic regression analysis revealed PCT-POB as an independent determinant for survival. Especially, PCT-POB not exceeding 50% and PCT-POB not exceeding 25% with CRP-POB not exceeding 75% on day 7 indicated a favorable outcome with a positive predictive value/sensitivity of 75%/97% and 92%/67%, respectively. In comparison, pretest likelihood to survive by day 28 and observed survival rate were 60% and 67%, respectively. Conclusions: Prediction rules of decrease in PCT-POB on day 7 in combination with CRP-POB may serve to monitor efficacy and guide duration of therapy in critically ill patients. (C) 2011 Elsevier Inc. All rights reserved.