老年经蝶垂体瘤术后水钠代谢紊乱的危险因素分析:一项单中心回顾性研究

Analysis of risk factors for water and sodium metabolism disorders in elderly patients after transsphenoidal pituitary tumor surgery:a single-center retrospective study

中文关键词: 老年  经蝶窦垂体瘤手术  水钠代谢紊乱  危险因素  预测模型

英文关键词: elderly  transsphenoidal pituitary tumor surgery  water and sodium metabolism disorders  risk factor  pre-diction model

中文摘要:

      目的 分析老年患者经蝶垂体瘤术后发生水钠代谢紊乱的危险因素,构建及验证风险预测模型.方法 采用回顾性分析,选取2021年1月—2023年12月于复旦大学附属华山医院神经外科行经蝶垂体瘤手术老年患者429例,通过交叉验证法,按6:4比例分为建模组(257例)和验证组(172例).建模组通过二项Logistics回归分析确定危险因素并建立风险预测模型,验证组通过受试者工作特征曲线(ROC)评估模型预测效果.结果 二项Logistics回归分析显示:合并糖尿(OR=3.810)、手术时长(OR=1.755)、肿瘤直径(OR=3.531)、术后血糖(OR=3.091)、术后高热(OR=5.111),是术后水钠代谢紊乱的危险因素(P＜0.05).验证组的ROC曲线下面积(AUC)为0.847(95％CI:0.788～0.905,P＜0.01),灵敏度为0.692,特异度为0.864,约登指数为0.560.结论 本研究构建的风险预测模型对高危老年患者预测效果好,可为临床医护人员提供一定的参考.

英文摘要:

      Objective To analyze the risk factors of water and sodium metabolism disorder in elderly patients after transsphenoidal pituitary tumor surgery,and establish and validate a risk prediction model.Methods A retrospective analysis was conducted.429 elderly patients who underwent transsphenoidal pituitary tumor surgery in Department of Neurosurgery,Huashan Hospital Affiliated to Fudan University from January 2021 to December 2023 were selected.By cross-validation meth-od,they were divided into modeling group(257 cases)and validation group(172 ones)according to a ratio of 6:4.In the modeling group,the risk factors were identified through binomial logistics regression analysis,and a risk prediction model was established.In the validation group,the predictive effect of the model was evaluated by receiver operating characteristic curve(ROC).Results Binomial logistic regression analysis showed that concurrent diabetes(OR=3.810),surgical duration(OR=1.755),tumor diameter(OR=3.531),postoperative blood glucose(OR=3.091),and postoperative high fever(OR=5.111)were risk factors for postoperative water and sodium metabolism disorders(P＜0.05).In the validation group,the area under ROC curve(AUC)was 0.847(95％CI:0.788-0.905,P＜0.01),the sensitivity was 0.692,the specificity was 0.864,and Yoden index was 0.560.P＜0.01).Conclusion The risk prediction model constructed in this study has a good prediction effect on high-risk elderly patients,and can provide a certain reference for clinical medical staff.