Background: Diagnosis/grading of infection and the systemic response to infection may be difficult on admission to the intensive care unit, but it is even more complicated for severely ill patients with long intensive care stays. The ACCP-SCCM criteria are difficult to apply for such patients, and objective, validated biomarkers would be of great use in this setting.
Methods: Long-term (>6 days) critically ill patients in the general ICU of University Hospital were prospectively enrolled in the study. All patients were assessed daily by the attending physician using the ACCP-SCCM classification. C-reactive protein (CRP, mg/dL), procalcitonin (PCT, ng/mL), and interleukin-6 (IL-6, pg/mL) of daily stored sera were measured after each patient's discharge. After discharge, an independent, overall clinical evaluation and an a posteriori ACCP-SCCM classification were chosen as the reference standard for all comparisons. The assessor was aware of the patient's clinical course but was blinded to levels of biomarkers.
Results: We studied clinical variables and biomarkers of 26 patients over a total of 592 patient days. The day-by-day ACCP-SCCM classification of the attending physician overestimated the severity of the inflammatory response to infection. The diagnostic discriminative ability of severe-sepsis/septic-shock for PCT was high (ROC area 0.952 [0.931-0.973]) and had a best threshold value of 1.58 (83.7% sensitivity, 94.6 % specificity). IL-6 had better discriminative ability than CRP, but both were worse than PCT.
Conclusion: PCT > 0.43 ng/mL could add to the clinical propensity for sepsis vs. SIRS not related to infection. Values higher than 1.58 ng/mL may support the bedside clinical diagnosis of severe-sepsis. PCT between 0.5 and 1.0 suggest tight daily monitoring of clinical conditions and re-evaluation of PCT.