Sepsis remains a leading cause of mortality in intensive care units (ICUs) globally; however, the effectiveness of conventional prognostic scoring systems varies across healthcare settings and patient populations. This study aimed to evaluate the comparative discriminative ability of the Acute Physiology and Chronic Health Evaluation II (APACHE II), Sequential Organ Failure Assessment (SOFA), and Charlson Comorbidity Index (CCI) in predicting 28-day mortality among critically ill septic patients. We conducted a retrospective observational study involving 392 consecutive adult patients diagnosed with sepsis according to the Sepsis-3 criteria, who were admitted to the ICU of a tertiary pulmonary care center in Serbia from January 2017 to December 2020. APACHE II scores were derived from the worst physiological values recorded within the first 24 hours of ICU admission; SOFA scores reflected the highest total score assessed during the same period; and CCI was calculated based on comorbidities present at admission. Discriminative performance was evaluated using receiver operating characteristic curve analysis. The overall 28-day mortality rate was 51.3%, which increased to 68.1% in patients with septic shock, with pneumonia being the source of infection in 97.4% of cases. APACHE II demonstrated the highest area under the curve (AUC) in the overall sepsis cohort (0.692, 95% CI 0.643-0.747), followed by SOFA (0.682, 95% CI 0.629-0.735) and CCI (0.667, 95% CI 0.613-0.720). In patients with septic shock, SOFA (AUC 0.671, 95% CI 0.561-0.782) and APACHE II (AUC 0.646, 95% CI 0.539-0.754) significantly outperformed CCI (AUC 0.423, 95% CI 0.302-0.543; p=0.006 and p=0.013, respectively), with no statistically significant difference between SOFA and APACHE II (p=0.828). Optimal cut-off values were identified as SOFA ≥8, APACHE II ≥21, and CCI ≥4, with corresponding sensitivities of 59.7%, 67.7%, and 56.2%, respectively. Both APACHE II and SOFA exhibited modest and comparable discriminative abilities for mortality prediction in this high-severity population, while CCI demonstrated limited utility in cases of septic shock. These findings underscore the continued relevance of conventional scoring systems and highlight the necessity for population-specific validation in high-acuity settings.
Gavrilovic et al. (Fri,) studied this question.