A Simple Scoring Model Predicting the Outcome of COVID-19 Patients: Tanta COVID score

Abstract

Background & Aim: COVID-19 is a worldwide pandemic with high rates of morbidity and mortality, and an uncertain prognosis leading to an increased risk of infection in health providers and limited hospital care capacities. In this study, we have proposed a predictive, interpretable prognosis scoring system with the use of readily obtained clinical, radiological and laboratory characteristics to accurately predict worsening of the condition and overall survival of patients with COVID -19. Methods: This is a single-center, observational, prospective, cohort study. A total of 347 patients infected with COVID-19 presenting to the Tanta university hospital, Egypt, were enrolled in the study, and clinical, radiological and laboratory data were analyzed. Top-ranked variables were identified and selected to be integrated into a Cox regression model, building the scoring system for accurate prediction of the prognosis of patients with COVID-19. Results: The six variables that were finally selected in the scoring system were lymphopenia, serum CRP, ferritin, D-Dimer, radiological CT lung findings and associated chronic debilitating disease. The scoring system discriminated risk groups with either mild disease or severe illness characterized by respiratory distress (and also those with hypoxia and in need for oxygen therapy or mechanical ventilation) or death. The area under the curve to estimate the discrimination performance of the scoring system was more than 90%. Conclusion: We proposed a simple and clinically useful predictive scoring model for COVID-19 patients. However, additional independent validation will be required before the scoring model can be used commonly.