Risk factors of secondary epileptic seizures in children with febrile seizures and construction of nomogram prediction model
10.3760/cma.j.cn115455-20220519-00463
- VernacularTitle:热性惊厥患儿继发癫痫发作的危险因素及构建列线图预测模型
- Author:
Min TANG
1
;
Jian ZHU
Author Information
1. 恩施州民族医院儿科，恩施 445000
- Keywords:
Seizures, febrile;
Epileptic seizures;
Risk factors;
Nomogram prediction model
- From:
Chinese Journal of Postgraduates of Medicine
2024;47(2):129-134
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To analyze the risk factors of secondary epileptic seizures in children with febrile seizures and to construct a nomogram prediction model.Methods:A total of 235 children with febrile seizures who were admitted to Enshi State Hospital for Nationalities from August 2018 to September 2021 were selected. According to whether the children had secondary epileptic seizures during the 6-month follow-up, the children were divided into the seizure group (62 cases) and no-seizure group (173 cases). The best cut-off value of each factor were obtained by the receiver operating characteristic (ROC). Multivariate Cox regression analysis was used to analyze the independent risk factors of secondary seizures in children with febrile seizures. The R software "rms" package was constructed to predict secondary seizures in children with febrile seizures. High-risk nomogram models, calibration curves was used for internal validation of nomogram models, and decision curves to assess the predictive power of nomogram models.Results:The age of the patients in the seizure group was lower than that in the no-seizure group: (14.45 ± 1.54) months vs. (21.47 ± 2.18) months; and the proportion of family history of epilepsy, the proportion of perinatal (abnormal), the proportion of seizure type (comprehensive), the proportion of electroencephalogram (EEG) (abnormal), the number of seizures, the duration of seizure, the tumor necrosis factor-alpha (TNF-α) level in the seizure group were higher than those in the no-seizure group: 56.45%(35/62) vs. 35.84%(62/173), 59.68% (37/62) vs. 15.61%(27/173), 70.97%(44/62) vs. 36.99% (64/173), 74.19% (46/62) vs. 20.81% (36/173), (5.45 ± 2.32) times vs. (2.04 ± 1.02) times, (18.89 ± 4.29) min vs. (12.62 ± 2.34) min, (25.65 ± 5.32) ng/L vs.(18.21 ± 2.29) ng/L, there were statistical differences ( P<0.05). The area under the curve (ACU) of age, number of convulsions, duration of convulsion, and TNF-α were 0.906, 0.913, 0.899, and 0.890, respectively; the best cut-off values were 3 years, 4 times, 15 min, 21 ng/L; age (≤3 years), family history of epilepsy (yes), type of seizures (generalized), perinatal period (abnormal), number of seizures (≥4 times), duration of seizures (≥15 min) were febrile seizures independent risk factors for secondary epileptic seizures in children ( P<0.05), the C-index of this nomogram prediction model was 0.744 (0.567-0.932); the decision curve showed that when the risk threshold was greater than 0.11, the clinical net benefit provided by this prediction model. The benefits were all higher than individual independent risk factors and provided a significant additional net clinical benefit in predicting a high risk of seizures secondary to febrile seizures in children with febrile seizures. Conclusions:This study constructed a nomogram model of the risk of secondary seizures in children with febrile seizures based on age, family history of epilepsy, type of seizures, perinatal period, number of seizures, and duration of seizures. Important strategic guidance.
