Prognostic Scoring System for Mortality of Hospitalized COVID-19 Patients in Resource-Limited Settings: A Multicenter Study from COVID-19 Referral Hospitals

Abstract

Background: Many studies identified the risk factors and prognostic factors related to in-hospital COVID-19 mortality using sophisticated laboratory tests. Cost and the availability of supporting blood tests may be problematic in resource-limited settings. This multicenter cohort study was conducted to assess the factors associated with mortality of COVID-19 patients aged 18 years and older, based on history taking, physical examination, and simple blood tests to be used in resource-limited settings. Methods: The study was conducted between July 2020 and January 2021 in five COVID-19 referral hospitals in Indonesia. Among 1048 confirmed cases of COVID-19, 160 (15%) died during hospitalization. Results: Multivariate analysis showed eight predictors of in-hospital mortality, namely increased age, chronic kidney disease, chronic obstructive pulmonary disease, fatigue, dyspnea, altered mental status, neutrophil-lymphocyte ratio (NLR) ≥ 5.8, and severe-critical condition. This scoring system had an Area-under-the-curve (AUC) of 84.7%. With cut-off score of 6, the sensitivity was 76.3% and the specificity was 78.2%. Conclusion: The result of this practical prognostic scoring system may be a guide to decision making of physicians and help in the education of family members related to the possible outcome.

Keywords

• covid-19
• prognostic
• predictive score
• mortality
• resource-limited settings