三叉神经痛球囊压迫术后并发症预测模型构建
该回顾性研究纳入164例接受经皮球囊压迫术的三叉神经痛患者,采用LASSO回归筛选危险因素并建立多因素logistic回归模型预测术后并发症。模型最终纳入神经分支、糖尿病、压迫时长和CD4+水平,显示出良好区分度(AUC=0.793)与校准度,决策曲线分析提示有临床实用价值。
要点
- 神经分支、糖尿病、压迫时长和CD4+水平是术后并发症独立危险因素。
- 预测模型区分度良好,AUC为0.793。
- 模型可用于术前风险评估并指导个体化预防干预。
核心结论
- 研究成功建立三叉神经痛患者经皮球囊压迫术后并发症的预测模型,为术前风险评估和个体化预防提供框架。
研究设计与样本
- 回顾性分析164例接受经皮球囊压迫术的三叉神经痛患者,采用LASSO回归筛选危险因素并构建多因素logistic回归预测模型。
主要发现
- 并发症组与非并发症组在神经分支、高血压、压迫时长和CD4+水平方面存在显著差异。
- 最终模型确定神经分支、糖尿病、压迫时长和CD4+水平为独立危险因素。
- 模型区分度良好(AUC=0.793)且校准度佳,决策曲线分析显示有临床实用价值。
关键数值与推荐
- 模型纳入的独立危险因素包括神经分支、糖尿病、压迫时长和CD4+水平。
- 模型AUC为0.793,提示区分度良好。
临床怎么用
- 可用于术前识别高危患者并制定针对性预防措施,降低术后并发症风险。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: Trigeminal neuralgia (TN) patients undergoing percutaneous balloon compression (PBC) may experience postoperative complications, warranting the investigation of influencing factors and preventive strategies. METHODS: A retrospective analysis of 164 TN patients who underwent PBC was conducted to compare baseline characteristics and identify risk factors using LASSO regression. A multivariate logistic regression model was developed to predict complications. RESULTS: Significant differences were observed between complication and non-complication groups related to nerve branches, hypertension, compression duration, and CD4+ levels. The final model identified nerve branches, diabetes, compression duration, and CD4+ levels as independent risk factors. The model exhibited good discriminatory power (AUC = 0.793) and calibration, with meaningful clinical utility shown through DCA. CONCLUSIONS: The study established a predictive model for postoperative complications in TN patients undergoing PBC, offering a framework for preoperative risk assessment and tailored preventive interventions.
引用
- Du Y, Xi Z, Wang Y, et al. Predictive Model Development for Postoperative Complications in Trigeminal Neuralgia Patients Undergoing Percutaneous Balloon Compression. World Neurosurg. 2026. PMID: 41544974. doi:10.1016/j.wneu.2026.124797
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。