Construction and validation of a prediction model for sepsis-associated delirium prognosis
10.3760/cma.j.issn.1671-0282.2024.05.010
- VernacularTitle:脓毒症相关谵妄患者预后预测模型构建及验证
- Author:
Xiaoqi BAI
1
;
Qiong GU
;
Jun XU
;
Huijie YU
Author Information
1. 嘉兴市第一医院急诊科，嘉兴 314000
- Keywords:
Sepsis associated delirium;
MIMIC-Ⅲ;
Prognosis;
Prediction model
- From:
Chinese Journal of Emergency Medicine
2024;33(5):651-657
- CountryChina
- Language:Chinese
-
Abstract:
Objective:The study aimed to analyze the risk variables influencing the prognosis of patients with sepsis-associated delirium (SAD) in the Intensive Care Unit (ICU) and build a prediction nomogram.Methods:This was a retrospective cohort study that includes patients with SAD in the Medical Information Mart for Intensive CareⅢ database (MIMIC-Ⅲ) database as training cohort, and patients who were hospitalized in the First Hospital of Jiaxing from January 2021 to September 2022 as validation cohort. Inclusion criteria: (1) age≥18 years old; (2) being admitted to the ICU for the first time; (3) the length of ICU stay≥24 h; (4) consistent with the diagnosis of sepsis; (5) the diagnosis of delirium was identified by CAM-ICU questionnaire. The general information, vital signs, past medical history and laboratory examination results of the patients were collected, and the outcome was 28-day mortality. Multiple logistic regression was used to identify independent influencing factors and the nomogram was constructed. The validity of the prediction model was determined using multiple indicators, including calibration curve, the area under the receiver operating characteristic curve (AUC), decision curve analysis (DCA), and Hosmer-Lemeshow test.Results:A total of 250 patients were included in the training cohort and 154 patients were in the validation cohort. The multiple logistic regression demonstrated that age ( OR=1.057, 95% CI: 1.030-1.084, P<0.001), respiratory frequency ( OR=1.117, 95% CI: 1.037-1.202, P=0.003), lactic acid ( OR=1.137, 95% CI:1.011-1.279, P=0,032), hemoglobin ( OR=0.983, 95% CI: 0.970-0.997, P=0.020), SOFA score ( OR=1.184, 95% CI: 1.070-1.309, P=0.001) were independent risk factors associated with the 28-day mortality of patients with SAD. The AUC of the nomogram created by the five factors above was 0.773 (95% CI: 0.705-0.841), and the Hosmer-Lemeshow test showed that the model was a good fit ( P=0.875). The DCA curve indicates that the model has potential net benefit. The AUC was 0.864 (95% CI: 0.799-0.928) in the validation cohort, and the Hosmer-Lemeshow test showed that the model was a good fit ( P=0.488). The DCA curve indicates that the model of the validation cohort had potential net benefit. Conclusion:The prediction model based on age, respiratory frequency, lactate, hemoglobin, and SOFA scores shows valuable capability of predicting the prognosis of patients with SAD, which could help clinicians identify risk factors at first time and make earlier intervention.
