Diagnostic accuracy of SeptiCyte® RAPID to discriminate sepsis from non-infectious critical illness in patients meeting sepsis criteria according to sepsis-3 definition at ICU admission

OBJECTIVES: The aim of this study was to validate the SeptiCyte® RAPID assay, a molecular test to distinguish sepsis from sterile inflammation, by determining its diagnostic accuracy in critically ill patients who meet criteria for sepsis according to Sepsis-3 definition on ICU admission DESIGN: This is an observational, prospective, and multicenter study. SETTING: Carried out in seven hospitals in Andalusia (Spain). A 2.5 mL whole blood sample was collected and tested in a SeptiCyte RAPID kit on a real time PCR platform (IdyllaTM). A score from 0 to 15 (SeptiScore™) was generated that falls into four bands based on the increasing likelihood of infection-positive systemic inflammation. PATIENTS: Patients aged 18 years or older, admitted to the ICU with a diagnosis of sepsis. MAIN RESULTS: We enrolled 354 patients, of whom 86 (24.3%) did not present sepsis at the researchers´ discretion. SeptiCyte® RAPID showed an AUC of [0.84 (CI95% 0.79-0.87)] for differentiating sepsis from sterile systemic inflammation. SeptiCyte® RAPID was significantly better for sepsis diagnosis than CRP [0.75 (CI95% 0.70-0.80)] (p =0.003) but without significant differences with PCT [0.80 (CI95% 0.75-0.84)]. SeptiScore distribution in patients with sepsis was higher than patients with sterile inflammation, with a PPV of 68.8% and 92.2% (Bands 3 and 4) for sepsis diagnosis and a PPV of 100% for sterile inflammation (Band 1). Independent risk factors for sepsis were estimated probability of sepsis [OR 8.02 (CI 95% 4.50-14.28), p