Diagnostic and prognostic value of procalcitonin in patients with sepsis

Aim To investigate predictive value of procalcitonin in diagnosis of sepsis in predicting positive blood culture, and possibility to predict final outcome in septic patients. Method This prospective study involved 106 hospitalized patients who met two or more criteria for systemic inflammatory response syndrome (SIRS). In comparison to Sepsis Related Organ Failure Assessment score (SOFA) and Acute Physiology and Chronic Health Evaluation (APACHE) II score procalcitonin (PCT), C-reactive protein and lactate levels were used to predict final outcome in septic patients (recorded as 28-day survival or non-survival). Using Receiver operating characteristic (ROC) curve the area under the curve (AUC) was calculated for diagnostic value and accuracy of different parameters with the best sensitivity and specificity for given cut-off values. Result Fifty-two out of 82 patients with documented sepsis had positive blood culture. Procalcitonin showed the best predictive value for both diagnosis of sepsis and bacteraemia with the cut-off value of 0.57 ng/mL (AUC 0.99) and 4.68 ng/mL (AUC 0.94), respectively. Serum lactate level showed the best 28-day mortality predictive value with the cut-off value of 3.25 mmol/L (AUC 0.95), and procalcitonin with the cut-off value of 15.05 ng/mL (AUC 0.92), followed by SOFA (AUC 0.92), CRP (AUC 0.84) and APACHE II score (AUC 0.83). Conclusion Monitoring of PCT in SIRS-positive patients raises possibility to distinguish between patients with sepsis and those with non-infectious SIRS. A significant correlation between PCT and SOFA, and APACHE II score in non-surviving septic patients indicates that PTC combined with clinical score could be useful for assessing severity of infection.