摘要
本文运用Cox模型,对117例肝癌根治术后的14项预后因子进行多因素分析。筛选出HBsAg(X_3)、肿瘤大小(X_s)肝硬变(X_9)、瘤栓(X_(11))及AFP转归(X_(12))等五项指标为影响预后的重要因子。由此建立的术后预测模型为h(t,z)=ho(t)exp(-0.9723Z_3+0.5599Z_s+1.0913Z_9+1.0158Z_(11)+1.6676Z_(12)),并用预后指数判断预后,敏感度为86.67％,可靠性达80.34％。本组病例的术后2年、5年及10年生存率分别为59.83％、33.33％及19.66％,总体预后甚好。Cox模型拟合采用南京铁道医学院卫生统计教研室编制的软件,并在IBM PC/XT微机上实现。
Cox model was applied to the multivariate analysis of 117 patients with liver cancer under radical operation. Among 14 factors, HBsAg (X_3), tumor size (X_5), cirrhosisX_9), thrombus in vessel (X_(11)), and AFP outcome (X_(12)) were most important factors which influence prognosis of postoperation. Cox model established is: h(t; z)=h_o(t)exp(-0.9723Z_3+0.5599Z_5+1.0913Z_9+1.0158Z_(11)+1.6676Z_(12)) The predicting value (PV) can be used as evaluation of prognosis. Its sensitivity and reliability were 86.67％ and 80.34％ respectively. The 2-, 5-, and 10- year survival rates after operation were 59.83％, 33.33％, and 19.66％, respectively. Statistical analysis was performed on an IBM PC/XT computer using Cox model software made by Nanjing Railway Medical College.
关键词
COX模型
预后因素
肝癌
Cox model
Prognosis of liver cancer
Survival analysis
