嗜铬细胞瘤和副神经节瘤患者术中血压剧烈波动预测模型的构建

Establishment of Prediction Model of Intraoperative Blood Pressure Severe Fluctuation for Patients with Pheochromocytoma and Paraganglioma

目的:分析嗜铬细胞瘤和副神经节瘤(pheochromocytoma and paraganglioma,PPGL)患者术中血压剧烈波动的危险因素,建立血压波动预测模型。方法:回顾性分析2015年1月-2021年1月笔者所在医院进行嗜铬细胞瘤切除术或包块切除术,且术后经病理确诊为PPGL患者的临床资料。包括人口学特征、肿瘤特征、术前合并症、术前检查、术前准备,麻醉因素、手术因素。计算每位患者术中血压变异度,并将患者分为血压剧烈波动组和非剧烈波动组,比较两组术后并发症发生情况、ICU入住率、ICU入住时间、术后总住院时间、住院期间死亡情况。采用多因素Logistic回归分析患者术中血压剧烈波动的独立危险因素,构建预测模型。结果:研究最终共纳入126例患者,剧烈波动组40例,非剧烈波动组86例。剧烈波动组术后并发症发生率及ICU入住率均显著高于非剧烈波动组,差异有统计学意义(P<0.05);剧烈波动组ICU入住时间及术后住院总时间均长于非剧烈波动组,差异有统计学意义(P<0.05)。两组均无住院期间死亡病例。术前有典型症状及24 h尿VMA水平是血压剧烈波动的独立危险因素(P<0.05)。预测模型ROC曲线下面积为0.799,灵敏度为72.5%,特异度为80.2%。结论:血压剧烈波动与术前有典型症状及24 h尿VMA水平密切相关。临床医师可在术前使用预测模型,做好术前准备,减少血压波动程度及持续时间。

Objective:To analyze the risk factors of intraoperative severe fluctuation of blood pressure in patients with pheochromocytoma and paraganglioma(PPGL)and establish a prediction model of blood pressure fluctuation.Method:The clinical data of patients who underwent pheochromocytoma resection or mass resection and were pathologically confirmed as PPGL in our hospital from January 2015 to January 2021 were retrospectively analyzed.Including demographic characteristics,tumor characteristics,preoperative comorbidities,preoperative examination,preoperative preparation,anesthesia factors,surgical factors.The blood pressure variability of each patient was calculated,and the patients were divided into the sever fluctuation group and the non-sever fluctuation group.The incidence of postoperative complications,ICU occupancy rate,ICU stay time,postoperative total length of hospital stay and death during hospitalization were compared between the two groups.Multivariate Logistic regression was used to analyze the independent risk factors of intraoperative blood pressure severe fluctuation,and a prediction model was established.Result:A total of 126 patients were finally included in the study,with 40 cases in the sever fluctuation group and 86 cases in the non-sever fluctuation group.The incidence of postoperative complications and ICU occupancy rate in the sever fluctuation group were significantly higher than those in the non-sever fluctuation group(P<0.05).The ICU stay time and postoperative total length of hospital stay in the sever fluctuation group were significantly longer than those in the non-sever fluctuation group(P<0.05).There was no death during hospitalization in both groups.Preoperative typical symptoms and 24 h urine VMA level were independent risk factors of intraoperative blood pressure severe fluctuation(P<0.05).The area under ROC curve of the prediction model was 0.799,the sensitivity was 72.5%,and the specificity was 80.2%.Conclusion:Preoperative typical symptoms and 24 h urine VMA level were closely associated with intraoperative blood pressure fluctuation.Clinicians can use the prediction model before operation to make preoperative preparations and reduce the degree and duration of blood pressure fluctuation.