| 摘要: |
| 【】目的:基于Lasso回归分析筛选结核性脑膜炎(TBM)患者预后的影响因素,构建列线图预测模型。方法:选取2022年5月~2025年8月期间本院接收205例的TBM患者为研究对象,根据出院时预后评估分为预后良好组143例和预后不良组62例。采用Lasso回归分析和二元logistic回归分析确定TBM患者预后不良的独立风险因素,并使用列线图构建TBM患者预后不良的预测评分系统；绘制ROC曲线、校准曲线和决策曲线验证列线图预测模型的区分度、校准度和临床适用性。结果:预后不良组年龄、营养不良中重度、脑积水、脑梗死、意识改变、抽搐、TBM分期Ⅲ期、外脑室引流、机械通气、脑脊液蛋白高于预后良好组,入院时GCS评分低于预后良好组(P＜0.05)。Lasso和二元logistic回归分析显示,年龄增加、营养不良中重度、有脑积水、入院时GCS评分减少、TBM分期Ⅲ期、脑脊液蛋白增加均为TBM患者预后不良的独立危险因素(P＜0.05)。ROC曲线结果显示,该列线图预测模型的AUC(95%CI)为0.937(0.896~0.978)；校准曲线结果显示,该列线图预测模型的预测概率与实际概率之间高度吻合；决策曲线结果显示,与两种极端情况相比,该列线图预测模型在0.02~0.99的风险阈值范围内提供了更高的净收益,模型预测区分度、校准度、临床适用性较高。结论:本研究构建的列线图可准确预测TBM患者预后不良风险,可能为患者的临床治疗提供参考。 |
| 关键词: |
| DOI： |
| 分类号: |
| 基金项目: |
|
|
|
郭睿
|
|
Second Affiliated Hospital of Air Force Military Medical University
|
| Abstract: |
| Objective: To screen the influencing factors of prognosis in patients with tuberculous meningitis (TBM) through Lasso regression analysis and construct a nomogram prediction model. Methods: Totally 205 patients with TBM who were admitted to our hospital from May 2022 to August 2025 were selected as the subjects. They were assigned into the good prognosis group (143 cases) and the poor prognosis group (62 cases) based on the prognosis assessment at discharge. Lasso regression analysis and binary logistic regression analysis were employed to identify the independent risk factors for poor prognosis in TBM patients, and a prediction scoring system for poor prognosis in TBM patients was constructed using a nomogram. In addition, ROC curves, calibration curves and decision curves were drawn to verify the discrimination, calibration and clinical applicability of the nomogram prediction model. Results: The age, moderate-severe malnutrition, hydrocephalus, cerebral infarction, altered consciousness, convulsions, TBM stage Ⅲ, external ventricular drainage, mechanical ventilation, and cerebrospinal fluid protein in the poor prognosis group were higher than those in the good prognosis group, while the GCS score at admission was lower than that in the good prognosis group (P < 0.05). Lasso and binary logistic regression analyses showed that increasing age, severe malnutrition, hydrocephalus, low GCS score at admission, TBM stage Ⅲ, and high cerebrospinal fluid protein were all independent risk factors for poor prognosis in TBM patients (P < 0.05). ROC curve analysis showed that the AUC (95% CI) of this nomogram prediction model was 0.937 (0.896 - 0.978). The calibration curve results showed that the predicted probability of the nomogram prediction model was highly consistent with the actual probability. The decision curve analysis showed that, compared with the two extreme scenarios, this nomogram prediction model provided higher net benefits within the risk threshold range of 0.02 to 0.99, and the discrimination, calibration, and clinical applicability of the prediction model were higher. Conclusion: The nomogram constructed in this study can accurately predict the risk of poor prognosis for patients with tuberculous meningitis, and may provide a reference for the clinical treatment of these patients. |
| Key words: |
