The predictive effect of advanced improved early warning scoring model on the occurrence of in-hospital cardiac arrest in critically ill patients
10.3760/cma.j.cn431274-20240103-00009
- VernacularTitle:高级改良早期预警评分模型对危重症患者院内心脏骤停发生的预测作用
- Author:
Wenbo ZHANG
1
;
Wei GU
Author Information
1. 河北医科大学急诊医学教研室，石家庄 050017
- Keywords:
Early warning score;
Predictive model;
Critical illness;
In-hospital cardiac arrest
- From:
Journal of Chinese Physician
2024;26(7):1046-1052
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To explore the relevant factors affecting in-hospital cardiac arrest (IHCA) in critically ill patients, establish an advanced modified early warning score (AMEWS) model based on the Modified Early Warning Score (MEWS), and explore the predictive effect of this model on the occurrence of IHCA in critically ill patients.Methods:A retrospective collection was conducted on 1230 critically ill patients admitted to the Chuiyangliu Hospital affiliated to Tsinghua University from January 2019 to March 2023. The patients were divided into two groups based on whether they developed IHCA or not, with 362 patients in the IHCA group and 868 patients in the NO-IHCA group. The clinical data of two groups of patients were compared, and the receiver operating characteristic (ROC) curve was used to obtain the optimal cutoff values for predicting the occurrence of IHCA, including MEWS, warning symptoms, electrocardiogram, lactate (LAC), blood potassium (K), troponin (TNI), brain natriuretic peptide (BNP), left ventricular ejection fraction (LVEF), and C-reactive protein (CRP). Multivariate logistic regression analysis was used to screen for independent risk factors that affected the occurrence of IHCA in critically ill patients, and an AMEWS model was constructed.Results:LAC, K, TNI, BNP, CRP, and MEWS in the IHCA group were higher than those in the NO-IHCA group, while LVEF was lower than that in the NO-IHCA group. The differences in warning symptoms and electrocardiogram were statistically significant (all P<0.05). The ROC curve analysis showed that the optimal cutoff values for MEWS, LAC, K, TNI, BNP, LVEF, and CRP to predict the occurrence of IHCA were 4 points, 2.20 mmol/L, 4.40 mmol/L, 0.10 ng/ml, 361 pg/ml, 50%, and 5 mg/L, respectively. The results of multiple logistic regression analysis showed that MEWS, warning symptoms, electrocardiogram, LAC, TNI, BNP, LVEF, and CRP were independent influencing factors for the occurrence of IHCA in critically ill patients (all P<0.05). Based on the risk factors obtained from multiple logistic regression analysis, an AMEWS model was constructed. With changes in warning symptoms and different types of electrocardiograms, MEWS score, LAC level, TNI level, BNP level, CRP level increased, LVEF score decreased, and the corresponding score of the column chart model increased, corresponding to an increased risk of IHCA in critically ill patients. In the AMEW model, patients with a score of 0-9 were classified as low-risk group (patients only needed to stay under observation and did not need to be hospitalized), those with a score of 10-14 are classified as medium risk group (the risk of disease changes increased and hospitalization was required), and those with a score of ≥15 were classified as high-risk group [the risk of death significantly increased and required intensive care unit (ICU) treatment]. Conclusions:MEWS, warning symptoms, electrocardiogram, LAC, TNI, BNP, LVEF, and CRP are related factors affecting the occurrence of IHCA in critically ill patients. The AMEWS model established based on MEWS has high predictive value for the occurrence of IHCA in critically ill patients.
