ABSTRACT
Background: Early detection of the illness trajectory in SARS-CoV-2 infected patients is crucial for patients and healthcare workers. A practical, non-invasive approach, with simple measurement for decision-making, is necessary for a pandemic to discriminate between high and low-risk patients, even though both may appear mild initially. Method(s): To determine COVID-19 disease severity within 10-days, distinguishing severe vs. moderate+mild and severe+moderate vs. mild, urinary L-type fatty acidbinding protein (L-FABP) was assayed less than four days after receiving a diagnosis. Severity was divided into mild (without oxygen), moderate (with oxygen) and severe (IMV/ECMO), and L-FABP within four days of diagnosis were used to predict severity within 10 days after onset. For risk prediction, urinary L-FABP was compared with interleukin-6. Next, the L-FABP point-of-care test (POCT) was further examined for its utility in risk screening in the international multicenter study. Result(s): Of the 224 participants in the study, 173 initially had a mild form. The AUC for a severe outcome was 96.3% in creatinine-adjusted L-FABP and 93.5% in net value. The AUC for a mild outcome was 85.0% and 84.9% net. The predictability of L-FABP for either a severe or mild outcome was superior to that of IL-6. Dynamics further showed better predictability of L-FABP than IL-6 during 10-days after onset. Finally, a multicenter-multicountry study confirmed the L-FABP POCT risk prediction of a severe outcome with an AUC of 88.9% and a mild outcome of 68.5%. Conclusion(s): Urinary L-FABP can predict the patient's risk of illness severity in COVID-19. L-FABP may detect severe cases more effectively and earlier than IL-6. L-FABP POCT is implementable for patient management.