儿童化脓性脑膜脑炎并发癫痫的危险因素分析及列线图预测模型构建与验证

Risk factors for epilepsy in children with purulent meningitis and establishment and validation of a nomogram prediction model

摘要:

目的 分析儿童化脓性脑膜脑炎感染后并发癫痫的危险因素，构建并验证列线图预测模型。方法 回顾性分析2020年1月至2024年6月南阳市中心医院儿童神经内科收治的486例化脓性脑膜脑炎患儿。根据是否发生癫痫分为癫痫组（89例）与非癫痫组（397例）。收集患儿相关资料，比较2组患儿相关资料的差异。采用多因素Logistic回归分析筛选癫痫的影响因素。基于影响因素构建列线图预测模型。结果 486例患儿中，癫痫发生率为18.3%（89例）；其中早发型31.5%（28例），晚发型68.5%（61例）。发作类型以全面性强直－阵挛发作为主。多因素Logistic回归分析显示，年龄、肺炎链球菌感染、脑脊液蛋白水平、神经元特异性烯醇化酶（NSE）、S100β蛋白、脑实质病变和脑电图重度异常或特殊模式是影响癫痫发生的危险因素。基于上述因素构建的列线图模型预测癫痫发生的曲线下面积（AUC）为0.891，最佳截断值为142分。Bootstrap内部验证显示模型具有良好的校准度。根据列线图总分将患儿分为低危组、中危组和高危组，癫痫发生率分别为3.8%、19.4%和63.8%（P<0.001）。结论 基于危险因素构建的列线图预测模型具有良好的区分度和校准度，可为临床早期识别化脓性脑膜脑炎并发癫痫的高危患儿提供可靠工具。

Abstract:

Objective To investigate the risk factors for epilepsy in children with purulent meningitis， and to establish and validate a nomogram prediction model.Methods A retrospective analysis was performed for 486 children with purulent meningitis who were admitted to Nanyang Central Hospital from January 2020 to June 2024. Aaccording to the presence or absence of epilepsy， they were divided into epilepsy group with 89 children and non-epilepsy group with 397 children. Related data were collected and compared between the two groups. A multivariate logistic regression analysis was used to investigate the risk factors for epilepsy， and a nomogram prediction model was established based on these risk factors.Results For the 486 children， the incidence rate of epilepsy was 18.3% （89/486）， among whom 28 （31.5%） had early-onset epilepsy and 61 （68.5%） had late-onset epilepsy. Generalized tonic-clonic seizure was the main type of seizure. The multivariate logistic regression analysis showed that age， pneumococcal infection， cerebrospinal fluid protein levels， neuron specific enolase， S100β protein， brain parenchymal lesions， and epilepsy-like or special patterns on electroencephalography were risk factors for the onset of epilepsy. The nomogram model established based on the above factors had an area under the curve of 0.891 for predicting the onset of epilepsy， with the optimal cut-off value of 142 points. Bootstrap internal validation showed that the model had good calibration. According to the total score of the nomogram， the children were divided into low-risk， medium-risk， and high-risk groups， and the incidence rates of epilepsy were 3.8%， 19.4%， and 63.8%， respectively （P<0.001）， suggesting that the model had good discriminatory ability.Conclusions The nomogram prediction model established based on the risk factors has good discriminatory ability and calibration and can provide a reliable tool for early identification of children with purulent meningitis at a high risk of epilepsy in clinical practice.