Usefulness of the COVID-GRAM and CURB-65 scores for predicting severity in patients with COVID-19

Aim: The aim of this study was to determine the usefulness of COVID-GRAM and CURB-65 scores as predictors of the severity of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in Caucasian patients. Methods: This was a retrospective observational study including all adults with SARS-CoV-2 infection admitted to Hospital Universitario Marqués de Valdecilla from February to May 2020. Patients were stratified according to COVID-GRAM and CURB-65 scores as being at low–medium or high risk of critical illness. Univariate analysis, multivariate logistic regression models, receiver operating characteristic curve, and area under the curve (AUC) were calculated. Results: A total of 523 patients were included (51.8% male, 48.2% female; mean age 65.63 years (standard deviation 17.89 years)), of whom 110 (21%) presented a critical illness (intensive care unit admission 10.3%, 30-day mortality 13.8%). According to the COVID-GRAM score, 122 (23.33%) patients were classified as high risk; 197 (37.7%) presented a CURB-65 score ≥2. A significantly greater proportion of patients with critical illness had a high COVID-GRAM score (64.5% vs 30.5%; P < 0.001). The COVID-GRAM score emerged as an independent predictor of critical illness (odds ratio 9.40, 95% confidence interval 5.51–16.04; P < 0.001), with an AUC of 0.779. A high COVID-GRAM score showed an AUC of 0.88 for the prediction of 30-day mortality, while a CURB-65 ≥2 showed an AUC of 0.83. Conclusions: The COVID-GRAM score may be a useful tool for evaluating the risk of critical illness in Caucasian patients with SARS-CoV-2 infection. The CURB-65 score could be considered as an alternative.