The predictive performance of NEWS, MEWS, SOFA and SAPS II in outcomes of bacteremic and non-bacteremic sepsis
Manish Anant Talkar, Durga Shankar Meena, Deepak Kumar, Gopal Krishana Bohra, Vibhor Tak, Amit Kumar Rohila, Ankur Sharma, MahendraKumar Garg
Abstract
BACKGROUND: This study evaluates the prognostic accuracy of sepsis scores and inflammatory biomarkers in predicting mortality, bacteremia, and septic shock among sepsis patients, with comparisons across bacteremic and non-bacteremic groups. METHODS: This prospective observational study was conducted in a tertiary care center in western India. A total of 196 sepsis patients were enrolled and categorized into bacteremic (n = 95) and non-bacteremic (n = 101) groups. Key sepsis scores analyzed included Sequential Organ Failure Assessment Score (SOFA), Modified Early Warning Score (MEWS), National Early Warning Score (NEWS), and Simplified Acute Physiology Score (SAPS II). Outcomes measured included 30-day mortality, bacteraemia prediction, and septic shock. Diagnostic performance was assessed using area under curve (AUCs), sensitivity, specificity, and odds ratio (OR) analysis. RESULTS: The overall 30-day mortality rate was 27.0%. Among all patients, NEWS (AUC: 0.861, 95% CI: 0.81-0.92, p < 0.001) and SOFA (AUC: 0.738, 95% CI:0.65-0.82, p < 0.001) demonstrated strong predictive capabilities for 30 days mortality. Subgroup analysis revealed that, in bacteremic sepsis, NEWS (AUC: 0.94, 95% CI: 0.89-0.99, p < 0.001) and ferritin (AUC: 0.804, 95% CI: 0.71-0.90, p < 0.001) showed the highest predictive values, while SOFA (AUC: 0.852, 95% CI: 0.76-0.94, p < 0.001) and HsCRP (AUC: 0.623, 95% CI: 0.51-0.74, p = 0.033) were most predictive for mortality in non-bacteremic sepsis. MEWS (≥ 7) demonstrated the strongest association with bacteremia (OR: 5.1, 95% CI: 2.4-11, p < 0.001, adjusted for age, and comorbidities), followed by SOFA (≥ 6), (OR: 2.4, 95% CI: 1.3-4.2, p = 0.005). For septic shock prediction, SOFA (≥ 6) (OR: 5.7) and procalcitonin (≥ 2.5 ng/mL) (OR: 4.9) were the strongest predictors. CONCLUSIONS: NEWS and SOFA demonstrated strong predictive accuracy for 30-day mortality, with ferritin providing additional value in bacteremic patients. For bacteremia prediction, MEWS (≥ 7) showed the highest association. These tools can provide critical support for clinicians in risk stratification and early intervention in severe sepsis. CLINICAL TRIAL NUMBER: not applicable.