Logistic回归预测肺叶切除术后并发症  

Forecasting Model of Pneumonectomy Complication with Logistic Regression

作　　者：李君荣[1] 孙丽萍[1] 蔡映云[2] 董鹤嘉[2] 徐正浪[2] 舒宝刚[2] 

机构地区：[1]镇江医学院,镇江市212007 [2]上海医科大学中山医院

出　　处：《中国医院统计》1998年第4期203-205,共3页Chinese Journal of Hospital Statistics

摘　　要：本文对473例肺叶切除病人进行研究,其中105例术后发生心、肺并发症。用Logistic回归进行多因素分析提示有6个因素是肺切除术后并发症的主要危险因素。它们是年龄、慢支史、FEV1/FEV1预计值、FVC/FVC预计值、肺切除的段数、麻醉时间等。并在此基础上建立了回归模型。使用该模型预测术后并发症的敏感性为82％、特异性为65％,超过临床经验预测。This study is to investigate 473 cases pneumonectomy among which 105 cases had developed postoperative complication. The research showed 6 risk factors which could forecast postoperative complication of pneumonectomy. They are age, history of bronchitis, preoperative FEV1/FEV1 predicted value, FVC/FVC predicted value, anaesthesia time and the numbers of segments resected. According to Logistic regression theory, forecasting model of pneumonectomy complication can be made. The model has many advantages compared with clinical estimation, and one advantage is its greater sensitivity.

关 键 词：术后并发症 LOGISTIC回归 肺叶切除术 肺切除术后 VVC 多因素分析 病人 提示 基础 时间 

分 类 号：R655[医药卫生—外科学] R734.2[医药卫生—临床医学]