Risk factors and prediction model for mortality risk in older adults septic patients

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Background The early identification of septic patients at high risk of mortality is essential for improving clinical outcomes. This study aimed to investigate the risk factors associated with 28-day mortality among older adults septic patients admitted to the intensive care unit (ICU) and to develop a predictive model based on these factors. Methods A total of 165 older adults septic patients admitted to the ICU of our hospital between April 2022 and March 2025 were included in this study. Multivariate logistic regression analysis was employed to identify independent risk factors associated with mortality. A prediction model was subsequently constructed and evaluated using receiver operating characteristic (ROC) curve analysis, with the area under the curve (AUC) used to assess its predictive accuracy. Results Among the 165 septic patients, 59 experienced 28-day mortality. Multivariate logistic regression analysis indicated that higher hemoglobin levels were associated with a reduced risk of 28-day mortality (odds ratio [OR]: 0.961; 95% confidence interval [CI]: 0.938–0.985; p = 0.001). A simplified risk estimation approach based on hemoglobin levels demonstrated good discriminatory performance (AUC: 88.3%; 95% CI: 82.7–93.9%). However, given that hemoglobin was the sole independent predictor retained in the multivariable model, the predictive performance should be interpreted as reflecting the prognostic value of this single biomarker rather than a true multivariable risk model. The findings should be considered exploratory and hypothesis-generating, requiring external validation in larger cohorts. Conclusion This exploratory study identified hemoglobin as a potentially useful biomarker for risk stratification in older septic patients. A simplified risk estimation approach based on hemoglobin demonstrated good preliminary discriminatory performance, but these findings should be interpreted with caution and require external validation in larger cohorts before clinical application.