Abstract
INTRODUCTION: While the CURB-65 score predicts mortality in community-acquired pneumonia (CAP), its performance in COVID-19 CAP is suboptimal. Hyperglycaemia correlates with an increased mortality in COVID-19. This analysis aims to enhance predictive accuracy for in-hospital mortality among COVID-19 patients by augmenting the CURB-65 score with objective variables, including markers of dysglycaemia. DESIGN: A single-centre retrospective observational analysis assessed the effectiveness of the CURB-65 score in predicting in-hospital mortality among adult patients with moderate to severe COVID-19 from March to September 2020. Using a binary logistic regression model, two extended CURB-65 scores which include markers of dysglycemia are proposed to enhance the predictive capability of the CURB-65 score for in-hospital mortality. RESULTS: Among 517 patients admitted, 117 (22.6%) died. Using the CURB-65 score, 393 patients (76%) were classified as low risk, 91 (17.6%) as medium risk and 33 (6.4%) as high risk. 37 patients were diagnosed with new-onset dysglycaemia, of which 22 (59.5%) died (p<0.001). Of those with dysglycaemia who died, 41% and 23% were classified as low risk and high risk using the CURB-65 score. The CURB-65 score demonstrated a modest area under the receiver operator characteristic curve (AUC) of 0.75 (95% CI 0.70 to 0.81) for in-hospital mortality in COVID-19 CAP. An Extended CURB-65 Score 1, incorporating an admission of fasting plasma glucose (FPG) and neutrophil to lymphocyte ratio, showed improved prognostic performance with an AUC of 0.80 (95% CI 0.76 to 0.85). When lactate and lactate dehydrogenase were added to these parameters (Extended CURB-65 Score 2), the AUC was 0.82 (95% CI 0.78 to 0.86). The integrated discrimination index showed an 11% and 24% higher discrimination slope when using the Extended CURB-65 Scores 1 and 2, respectively. CONCLUSIONS: The addition of common biochemical parameters including an admission FPG enhances the prognostic performance of CURB-65 for in-hospital mortality among patients with COVID-19.