The ability of Procalcitonin, lactate, white blood cell count and neutrophil-lymphocyte count ratio to predict blood stream infection. Analysis of a large database

The global burden of death due to sepsis is considerable. Early diagnosis is essential to improve the outcome of this deadly syndrome. Yet, the diagnosis of sepsis is fraught with difficulties. Patients with blood stream infection (BSI) are at an increased risk of complications and death. The aim of this study was to determine the diagnostic accuracy of four readily available biomarkers to diagnose BSI in patients with suspected sepsis. In this retrospective, observational, Electronic Medical Record based study we compared the accuracy of procalcitonin (PCT), serum lactate concentration, total white blood cell (WBC) count and the neutrophil-lymphocyte count ratio (NLCR) to diagnose BSI in adult patients presenting to hospital with suspected sepsis. Based on the blood culture results patients were classified into 1 of the following 5 groups: i) negative blood cultures, ii) positive for a bacterial pathogen, iii) positive for a potential pathogen, iv) fungal pathogen and v) potential contaminant. Group 2 was further divided into Gram –ve and Gram +ve pathogens. Receiver operating characteristic (ROC) curves were constructed to compare the diagnostic performance of the biomarkers. There were 1767 discreet patient admissions. The median PCT concentration differed significantly across blood culture groups (p < 0.0001). The highest median PCT concentration was observed in patients with a Gram-negative pathogen (17.1 ng/mL; IQR 3.6–49.7) and the lowest PCT in patients with negative blood cultures (0.6 ng/mL; IQR 0.2–2.8). The AUROC was 0.83 (0.79–0.86) for PCT, 0.68 (0.64–0.72) for the NLCR, 0.55 (0.51–0.60) for lactate concentration and 0.52 (0.48–0.57) for the WBC count. The AUROC for PCT was significantly greater than that of the NLCR (p < 0.0001). A PCT less than 0.5 ng/mL had a negative predictive value of 95% for excluding BSI. The best cut-off value of PCT for predicting BSI was 1.5 ng/ml. Our results suggest that PCT of less than 0.5 ng/mL may be an effective screening tool to exclude BSI as the cause of sepsis, while the diagnosis of BSI should be considered in patients with a PCT above this threshold. The total WBC count and blood lactate concentration may not be reliable biomarkers for the diagnosis of BSI. The NLCR may be a useful screening test for BSI when PCT assays are not available.