Uselfulness of biomarkers to predict bacteremia in patients with infection in the emergency department

Between all patients attended in the Emergency Department (ED), 14.3% have an infectious disease diagnosis. Blood cultures (BC) are obtained in 14.6% of patients and have a profitability of 20%, whereas 1% are considered as contaminated and 1-3% of positive cultures correspond to discharge patients ("hidden bacteraemia"). The highest number of confirmed bacteraemias comes from the samples of patients with urinary tract infections, followed by community-acquired pneumonia. The suspicion and detection of bacteraemia have an important diagnostic and prognostic significance and could modify some important making-decisions (admission, BC request, administration of appropriate and early antimicrobial, etc). Therefore, finding a predictive model of bacteraemia useful and applicable in ED has become the objective of many authors that combine different clinical, epidemiological and analytical variables, including infection and inflammatory response biomarkers (IIRBM), as they significantly increase the predictive power of such models. The aim of this review is to highlight the evidence showed in recent published articles, to clarify existing controversies, and to compare the accuracy of the most important IIRBM to predict bacteremia in patients attended due to infection in the ED. Finally, to generate different recommendations that could help to define the role of IIRBM in improving the indication to obtaining BC, as well as in immediate decision-making in diagnosis and treatment (early and adequate antibiotic treatment, complementary tests, other microbiological samples, hemodynamic support measures, need for admission, etc.)