建立预测老年患者非神经外科非心血管手术围术期死亡风险的危险指数评分的临床研究

目的 建立预测老年患者非神经外科非心血管手术围术期死亡风险的危险指数评分.方法 选取2012年12月至2016年3月于北京大学人民医院接受非神经外科非心血管手术治疗的年龄大于或等于65岁的住院患者11 144例,分为死亡组及存活组.对比两组患者的一般资料、合并疾病、术前化验室检查、手术麻醉情况.行多因素Logistic回归分析该人群围术期死亡的危险因素,采用Bootstrapping法行内部校验,通过校正方程的偏回归系数确立危险指数评分的参数权重.结果 围术期病死率为1.0％(111例).通过回归分析获得8个独立预测因素,并定义危险指数评分:美国医师协会分级(工级0分,Ⅱ级3分,Ⅲ或Ⅳ级4分)、BMI(＜24 kg/m2 0分,≥24 kg/m2-1分)、肾功能不全(1分)、慢性阻塞性肺病(3分)、需胰岛素治疗的糖尿病(2分)、低清蛋白血症(1分)、低钠血症(1分)、实施全身麻醉(1分).危险指数评分:＜6分的患者为低危,6～7分的患者为中危,＞7分的患者为高危,高危患者围术期死亡的实际预测风险大于10％.围术期死亡危险指数评分展现出了较好的诊断识别能力(c-statistic=0.878).结论 老年患者非神经外科非心血管手术的围术期死亡风险可以通过危险指数评分进行预测,这一评分可以帮助筛选围术期死亡的高危人群,以便针对性的给予更精细化的围术期管理.

Study on establishing risk index scoring for predicting perioperative mortality risk in aged patients undergoing non-neurologic and non-cardiovascular surgery

Objective To develop a risk index scoring for predicting perioperative mortality risk in aged patients undergoing non-neurologic and non-cardiovascular surgery.Methods A total of 11 144 inpatients aged ＞65 years old undergoing non-neurologic and non-cardiovascular surgery in the People's Hospital of Peking University from December 2012 to March 2016 were selected and divided into the death group and survival group.The following variables were compared between the 2 groups:general data,comorbidities,preoperative laboratory tests and operation anesthesia.A multivariate Logistic regression analysis was performed on the risk factors for perioperative death in this group.The Bootstrapping method was performed for conducting internal validation.The parameters weighing of risk index scoring was established by correcting the partial regression coefficient of equation.Results The perioperative mortality was 1.0％ (111 cases).Eight independent predicting factors were obtained by the regression analysis.Then the risk index scoring was defined:classification of the Association of American Physicians (grade Ⅰ:0 point;grade 1:3 points;grade Ⅲ or 1:4 points),BMI(＜24 kg/m2:0 point;＞24 kg/m2:-1 point),renal insufficiency(1 point),chronic obstructive pulmonary disease(3 points),diabetes needing insulin treatment(2 points),preoperative hypoalbuminemia (1 point),preoperative hyponatremia (1 point) and general anesthesia (1 point).The patients with risk scores＜6 points were classified as low risk,patients with risk score 6-7 points were classified as intermediate risk and those with risk score＞ 7 were classified as high risk.The actual predicting risk of perioperative death in high risk patients ＞10％.The perioperative mortality risk index exhibited better diagnostic recognition ability (c-statistic=0.878).Conclusion The perioperative mortality risk of aged patients undergoing non-neurologic and non-cardiovascular surgery can be predicted by the risk index scoring,which can help to screen the high-risk individuals of perioperative death in order to give more carefully perioperative management.