Abstract
BACKGROUND: Sepsis is one of the leading causes of morbidity and mortality worldwide. Early detection and reliable prediction of sepsis-related mortality are crucial. The prognostic value of National Early Warning Score 2 (NEWS-2) in sepsis patients with cardiovascular and neurovascular diseases is not well elucidated. OBJECTIVES: This study aims to evaluate and compare the qSOFA, SOFA, qPitt, and NEWS-2 scores for predicting mortality in sepsis and septic shock patients. DESIGN: An observational retrospective study was conducted involving patients with sepsis or septic shock. METHODS: Data were collected between 2023 and 2024. Baseline measures included sociodemographic and clinical characteristics. All subjects were identified from a database of a highly complex healthcare institution focused on the diagnosis and treatment of cardiovascular and neurovascular diseases. The scores qSOFA, SOFA, NEWS-2, and qPitt were collected from medical records and analyzed. A sample size was estimated. The area under the receiver operating characteristic curve (AUROC) analysis was established for each score. A multivariate analysis adjusted for confounding factors through logistic regression was performed. RESULTS: A total of 126 participants were included, with a median age of 72 years (IQR: 59-81), and 76 (60.32%) being men. The overall mortality rate in the population was 19.8%, with 84% of deaths occurring in males. Statistically significant differences in variables such as heart failure, mild liver disease, and atrial fibrillation were observed between the groups (p < 0.05). The NEWS-2 score with a cut-off point of ⩾8 showed a significant association with mortality (p = 0.0001). The AUROC for NEWS-2 was 0.80, compared to a ROC of 0.6960 for qPitt, 0.6776 for SOFA, and 0.5868 for qSOFA (χ(2) = 16.60, Prob > χ(2) = 0.0009). CONCLUSION: NEWS 2 exhibits better performance in predicting mortality among elderly adults with cardiovascular and neurovascular diseases suffering from sepsis or septic shock, compared to the qPitt, SOFA, and qSOFA scores.