Procalcitonin in the diagnosis of inflammation in intensive care units

Objectives: To assess the effectiveness of different procalcitonin cutoff values to distinguish non-infected (negative+STRS) from infected (sepsis+severe sepsis+septic shock) medical and surgical patients. Design and methods: PCT plasma concentration was measured using an automated chemiluminescence analyzer in 10 13 samples collected in 103 patients within 24 h of admission in ICU and daily during the ICU stay. We compared PCT levels in medical and surgical patients. We also compared PCT plasma levels in non-infected versus infected patients and in SIRS versus infected patients both in medical and in surgical groups. Results: Median values of PCT plasma concentrations were significantly higher in infected than in non-infected groups, both in medical (3.18 vs. 0.45 mu g/L) (p < 0.0001) and in surgical (10.45 vs. 3.89 mu g/L; p < 0.0001) patients. At the cutoff of 1 mu g/L, the LR+ was 4.78, at the cutoff of 6 mu g/L was 12.53, and at the cutoff of 10 mu g/L was 18.4. Conclusions: This study highlights the need of different PCT cutoff values in medical and surgical critically ill patients, not only at the ICU admission but also in the entire ICU stay. (c) 2006 The Canadian Society of Clinical Chemists. All rights reserved.