Abstract
The elevated risk of severe COVID-19 outcomes in patients with diabetes underscores the need for effective predictive tools. This study aimed to assess the predictive accuracy of APACHE II, CURB-65, SOFA, and NEWS2 scores at critical time points in diabetic patients diagnosed with COVID-19, aiming to guide early and potentially life-saving interventions. In a prospective cohort study conducted from January 2021 to December 2023, adult patients with type 1 or type 2 diabetes and confirmed SARS-CoV-2 infection were evaluated. Clinical scores were calculated at admission and five days post-symptom onset, with data analyzed using receiver operating characteristic (ROC) curves and logistic regression to determine areas under the curve (AUC) and hazard ratios (HR) for severe outcomes. Among the 141 diabetic patients studied, ROC analysis revealed high AUC values for SOFA (0.771 at admission, 0.873 at day five) and NEWS2 (0.892 at admission, 0.729 at day five), indicating strong predictive accuracy for these scores. The APACHE II score's AUC improved from 0.698 at admission to 0.806 on day five, reflecting worsening patient conditions. Regression analysis showed significant HRs associated with exceeding threshold scores: The SOFA score HR at day five was 3.07 (95% CI: 2.29-4.12, p < 0.001), indicating a threefold risk of severe outcomes. Similarly, the APACHE II score showed an HR of 2.96 (95% CI: 2.21-3.96, p < 0.001) at day five, highlighting its utility in predicting severe disease progression. The SOFA and NEWS2 scores demonstrated excellent early predictive accuracy for severe COVID-19 outcomes in diabetic patients, with significant AUC and HR findings. Continuous score monitoring, especially of APACHE II and SOFA, is crucial for managing and potentially mitigating severe complications in this vulnerable population. These tools can effectively assist in the timely escalation of care, thus potentially reducing morbidity and mortality among diabetic patients during the COVID-19 pandemic.