Risk factors for mortality and development of a predictive model in pediatric sepsis.

Abstract

Objective: To investigate the clinical characteristics and risk factors associated with mortality in pediatric sepsis patients, and to establish a predictive model for early identification of high-risk children.

Methods: A retrospective cohort study was conducted including 143 pediatric sepsis cases admitted to the Pediatric Intensive Care Unit of the Second Hospital of Hebei Medical University from January 2020 to December 2024. Clinical data, laboratory indicators, and treatment history were collected. Univariate and multivariate logistic regression analyses were performed to identify risk factors for mortality. A nomogram model was constructed based on significant predictors, and its predictive performance was evaluated by receiver operating characteristic (ROC) curve analysis.

Results: Among the 143 cases, 121 survived and 22 died. Significant differences were observed between the survival and death groups in lymphocyte count, platelet count, albumin, D-dimer, liver function tests (ALT, TBIL), CALLY index, and pre-admission glucocorticoid use (P < 0.05). Multivariate analysis identified platelet count (OR = 0.992, 95% CI: 0.987-0.997), D-dimer (OR = 7.571, 95% CI: 2.642-21.698), and CALLY index (OR = 0.532, 95% CI: 0.323-0.877) as independent risk factors for mortality. The nomogram model incorporating these factors showed good predictive accuracy with an area under the ROC curve of 0.859 (95% CI = 0.742-0.953).

Conclusion: Platelet count, D-dimer level, and CALLY index are valuable indicators for assessing prognosis in pediatric sepsis and can aid in early risk stratification. The established nomogram provides a useful tool for clinical decision-making to improve outcomes in high-risk pediatric sepsis patients. Further multicenter prospective studies are warranted to validate and refine these findings.