A Lever and a Place to Stand to Predict COVID-19 Progression: Developing a Prognostic Model Based on Day Five from Symptoms Onset

Abstract

Background: A major limitation of predictive prognostic models in COVID-19 patients is the heterogeneity of disease stage and population. This study aims at identifying a panel of clinical and laboratory parameters that at day-5 of symptoms onset could predict disease progression within 11 days in hospitalized COVID-19 patients.Methods: Single-centre, prospective cohort study on hospitalized adult COVID-19 patients. Patient-level epidemiological, clinical, and laboratory data were collected at fixed time-points: day-5, -10, and -15 from symptoms onset and in case of intensive care unit (ICU) admission, discharge, or death. COVID-19 progression was defined as in-hospital death and/or ICU and/or respiratory failure (PaO2/FiO2 ratioResults: Among 235 patients with COVID-19 prospectively admitted in a 3-month period, 88 (37%) experienced COVID-19 progression. A model including male sex, age >65 years, cardiovascular disease, and at least three abnormal laboratory parameters among CRP > 80 U/L, AST > 45 U/L, ALT > 40 U/L, NLR > 4·5, LDH > 250 U/L, and CK > 80 U/L showed an AUC of 0·73 (95%CI: 0·66 - 0·81) for predicting disease progression. Conclusion: An easy-to-use panel of laboratory/clinical parameters computed at day-5 from symptoms onset predicts, with fair discrimination ability, COVID-19 progression. Assessment of these features at day-5 from symptoms onset could facilitate clinicians’ decision making and be used to increase patient population in clinical trials in hospitalized patients.Trial Registration: COVID 19-VR registry (ClinicalTrials.gov NCT04497194).Funding Statement: This study is part of React-COVID-19 project funded by Fondazione CARIVERONA.Declaration of Interests: The authors have no conflict of interests to be declared.Ethics Approval Statement: The study was approved by the hospital Institutional Review Board (IRB 2577CESC).