Establishment and validation of predictive model for children with pneumonia complicated with sepsis
10.3760/cma.j.cn115455-20230918-00265
- VernacularTitle:肺炎并发脓毒症儿童预测模型的建立与验证
- Author:
Rui ZHANG
1
;
Feng ZHU
1
Author Information
1. 徐州医科大学附属徐州儿童医院急诊科，徐州 221002
- Publication Type:Journal Article
- Keywords:
Pneumonia;
Sepsis;
Child;
Models, statistical
- From:
Chinese Journal of Postgraduates of Medicine
2025;48(2):150-153
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To establish and validate a predictive model for children with pneumonia complicated with sepsis.Methods:The clinical data of 45 children with pneumonia complicated with sepsis(observation group) and 45 children with simple pneumonia (control group) who were treated at the Xuzhou Children′s Hospital, Xuzhou Medical University from January 2019 to November 2022 were retrospectively selected. The clinical characteristics of the two groups were compared. Multivariate Logistic regression analysis was used to screen the risk factors of pneumonia complicated with sepsis. The predictive model was established and verified by R4.0.3 statistical software, and the predictive value of the predictive model for pediatric pneumonia complicated with sepsis was analyzed by the receiver operating characteristic (ROC) curve.Results:The levels of procalcitonin, pediatric critical case score (PCIS) in the observation group were lower than those in the control group: (70.20 ± 0.30) ng/L vs. (51.70 ± 0.26) ng/L, (71.73 ± 14.29) scores vs. (83.42 ± 7.78) scores, and the ratio of pathological type of lobular pneumonia, acidosis, and electrolyte metabolism disorder in the observation group were higher than those in the control group: 42.22%(19/45) vs. 13.33%(6/45), 62.22%(28/45) vs. 26.67%(12/45), 53.33%(24/45) vs. 17.78%(8/45), there were statistical differences ( P<0.05). The results of multivariate Logistic regression analysis showed that PCIS, acidosis, and electrolyte metabolism disorder were independent risk factors of pneumonia complicated with sepsis in children ( P<0.05). The nomogram, clinical decision curve, correction curve, and ROC curve were made, the area under the curve (AUC) of the training set was 0.863 (95% CI 0.756 - 0.971), and the AUC of the validation set was 0.862 (95% CI 0.753 - 0.971). Hosmer-Lemeshow Goodness-of-Fit Test was performed on the model in the validation set, χ2 = 9.50, P = 0.302, indicated that the model had good clinical value and reliability in predicting pneumonia complicated with sepsis in children. Conclusions:The prediction model can identify high-risk children with pneumonia complicated sepsis early, which may have potential significance for the next step of prevention and treatment.
