Establishment of an early warning score model in children with hand-foot-mouth disease
10.3760/cma.j.cn101070-20190816-00754
- VernacularTitle:儿童手足口病早期预警评分模型的构建
- Author:
Guiling YU
1
;
Lu MEI
;
Xiaoxia XU
;
Yan KONG
Author Information
1. 青岛市妇女儿童医院护理部 266034
- From:
Chinese Journal of Applied Clinical Pediatrics
2020;35(22):1708-1711
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct an early warning score model for the identification of hand-foot-mouth disease (HFMD) in children using the decision tree, and to provide scientific evidence for early recognition of critical HFMD in clinical practice.Methods:The medical records of 2 488 children with severe and critical HFMD hospi-talized in Qingdao Women and Children′s Hospital from January 2014 to December 2016 were retrospectively analyzed.Multivariate Logistic regression analysis was made to screen the independent risk factors of critical HFMD.An early warning score model for identifying children with HFMD was established by using the chi- square automatic interactive detection method in the decision tree, and the effectiveness of the model was evaluated by the receiver operating characteristic curve (ROC). Results:The multivariate Logistic regression analysis identified 5 independent risk factors of critical HFMD, including age<3 years old, blood sugar≥8.0 mmol/L, body temperature≥39.0 ℃, limb shaking and trembling, and hand-foot herpes.Two risk factors, i.e., blood sugar and body temperature, were included in the tree chart through the decision tree, and scoring criteria were established according to the tree chart.Combining the early warning scoring system for children, the warning model were finally set up, which consisted of 5 dimensions and inclu-ded 9 evaluation indicators: consciousness, skin color, heart rate, capillary refill time, respiratory frequency, participation of respiratory muscles, oxygen therapy, blood sugar and body temperature.Each dimension ranged from 0 to 3 points, totaling 0-15 points.The area under the ROC, sensitivity, specificity, maximum Youden index, and cut-off value of the model were 0.911, 0.910, 0.774, 0.684, and 6.5, respectively. Conclusions:The early warning score model established in this paper can accurately identify critical HFMD, with high sensitivity and specificity.
