重型肝病患者的预后生存模型的研究  

Research on the Prognostic Model for Patients with Severe Liver Disease

作　　者：武文芳[1] 张晶[2] 段钟平[2] 马斌荣[1] 

机构地区：[1]首都医科大学生物医学工程学院,北京100069 [2]首都医科大学附属北京佑安医院,北京100069

出　　处：《中国医疗设备》2008年第8期6-9,共4页China Medical Devices

基　　金：首都医学科研发展基金(2002-2039)

摘　　要：目的研究影响重型肝炎患者预后的各种因素,建立适合我国肝病特点的预后判断模型。方法选择2001年3月至2006年6月确诊为重型肝炎的527例患者的临床指标和随访资料,用COX比例风险模型进行单因素和多因素分析,建立生存模型。用400例病例进行验证,评价模型判断能力。结果COX回归分析结果显示消化道出血、年龄、腹水程度、丙氨酸转氨酶、谷草转氨酶,丙氨酸转氨酶、总胆红素、胆固醇、血清肌酐、凝血酶原活动度是影响预后的重要因素;验证结果显示模型ROC曲线下面积好于CTP、MELD评分。结论本预后判断模型适合我国以病毒性肝炎为主的特点,对评价我国重型肝炎患者的预后及对临床选择治疗方法有较高的指导意义。Objective To analyze prognostic factors of patients with severe liver disease and develop a prognostic model for them, which is fitful to the characteristics of our country. Methods 527 patients with severe liver disease from March 2001 to June 2006 were evaluated, the follow-up information were collected, including the survival state （death/survival）. The survival model was established by using COX proportional hazard model. New 400 cases were used to test and evaluate the decision capability of the model. Results Nine prognostic factors were extracted by using COX, the prognostic index （PI） was calculated. By statistics analyze to compare the ROC area （AUC） of this system model with CTP and MELD, this system meld is better than CTP and MELD. Conclusion The developed model is fitful to the characteristics of our country where there are a lot of virus hepatitis patients, so the model is valuable in the prognostic evaluation of severe liver patients and very useful to guide clinicians in selecting treatment methods in China.

关 键 词：重型肝炎 回归方程 预后分析 COX比例风险模型 

分 类 号：TB114[理学—概率论与数理统计] R575.1[理学—数学]