Abstract
We aimed to develop presepsin as a marker of diagnosis of severe infections of either bacterial and viral origin. The derivation cohort was recruited from 173 hospitalized patients with acute pancreatitis or post-operative fever or infection suspicion aggravated by at least one sign of the quick sequential organ failure assessment (qSOFA). The first validation cohort was recruited from 57 admissions at the emergency department with at least one qSOFA sign and the second validation cohort from 115 patients with COVID-19 pneumonia. Presepsin was measured in plasma by the PATHFAST assay. Concentrations more than 350 pg/ml had sensitivity 80.2% for sepsis diagnosis in the derivation cohort (adjusted odds ratio 4.47; p < 0.0001). In the derivation cohort, sensitivity for 28-day mortality prognosis was 91.5% (adjusted odds ratio 6.82; p: 0.001). Concentrations above 350 pg/ml had sensitivity 93.3% for the diagnosis of sepsis in the first validation cohort; this was 78.3% in the second validation cohort of COVID-19 aiming at the early diagnosis of acute respiratory distress syndrome necessitating mechanical ventilation. The respective sensitivity for 28-day mortality was 85.7% and 92.3%. Presepsin may be a universal biomarker for the diagnosis of severe infections of bacterial origin and prediction of unfavorable outcome.