背景:尽管最近取得了进展,但开颅手术治疗幕上原发性恶性脑肿瘤的并发症发生率仍然很高。
目的:本研究的目的是表征与这些并发症相关的因素。
方法:数据来自2016年至2019年的国家外科质量改进计划数据库。包括接受开颅手术切除幕上原发性恶性脑肿瘤的患者。协变量包括人口统计学/合并症,术前实验室值,美国麻醉医师协会(ASA)分类,手术时间,术后并发症。采用后向和前向选择的多变量逻辑回归来评估死亡的独立预测因子,住院时间延长,术后卒中伴神经功能缺损(CVA),和计划外的重新接纳。使用接受者工作曲线下面积(AUC)评估模型的预测拟合。
结果:纳入8965例,CVA术后30天的风险为1.9%,计划外再入院的10.1%,1.2%用于长期住院,死亡为2.4%。年龄,ASA类别,播散性癌症,术前功能依赖,术后呼吸系统并发症是30天死亡率的预测因素(AUC,0.83;P<0.001)。手术时间延长对CVA的预测效果最好(P<0.001),年龄,ASA类别,和最近的体重减轻(AUC,0.63;P=0.009)。非选择性手术状态可预测住院时间延长,从入院到手术的时间,再插管,和术后脓毒症(AUC,0.78;P<0.001)。长期使用类固醇预测了计划外的再入院,术后血栓并发症,器官/空间手术部位感染,深静脉血栓形成,术后全身性脓毒症,和脓毒性休克(AUC,0.68;P<0.001)。
结论:我们的研究确定了这部分脑肿瘤患者开颅手术后30天主要并发症的预测因素。
The rate of complications remains significant after craniotomy for supratentorial primary malignant brain tumors despite recent advances.
The goal of this study is to characterize factors associated with these complications.
Data were extracted from the National Surgical Quality Improvement Program database from 2016 to 2019. Patients who underwent a craniotomy for resection of supratentorial primary malignant brain tumors were included. Covariates included demographics/comorbidities, preoperative laboratory values, American Society of Anesthesiologists (ASA) classification, operative time, and postoperative complications. Multivariable logistic regression with backward and forward selection was used to evaluate independent predictors of death, prolonged hospitalization, postoperative stroke with neurologic deficit (CVA), and unplanned readmission. Predictive fit of the model was evaluated using the area under the receiver operating curve (AUC).
Of 8965 included cases, the 30-day postoperative risks were 1.9% for CVA, 10.1% for unplanned readmission, 1.2% for prolonged hospitalization, and 2.4% for death. Age, ASA category, disseminated cancer, preoperative functional dependence, and postoperative respiratory complications were predictors of 30-day mortality (AUC, 0.83; P < 0.001). CVA was best predicted by increased operation time (P < 0.001), age, ASA category, and recent weight loss (AUC, 0.63; P = 0.009). Prolonged hospitalization was predicted by nonelective surgery status, time from admission to surgery, reintubation, and postoperative sepsis (AUC, 0.78; P < 0.001). Unplanned readmission was predicted by chronic steroid use, postoperative thrombotic complications after surgery, organ/space surgical site infection, deep vein thrombosis, postoperative systemic sepsis, and septic shock (AUC, 0.68; P < 0.001).
Our study identifies predictors of major 30-day complications after craniotomy for this subset of patients with brain tumor.