影响膀胱肿瘤患者长期存活因素回归性分析探究.doc

影响膀胱肿瘤患者长期存活因素回归性分析探究 　【摘要】 目的 对膀胱肿瘤预后多因素进行综合评价，探讨影响膀胱肿瘤患者长期存活的相关指标。方法 通过对284例膀胱肿瘤病案回顾分析，对患者和(或)家属进行问卷调查、电话咨询等方式获取临床资料并采用目前常用的生存分析的统计学方法，计算生存率，生存率比较用χ2检验，应用计算机软件进行回归分析，建立Cox比例分析模型，检验病理分期分级及各种治疗方式与患者长期存活的关系。结果 284例患者中，最终存活189例，死于膀胱肿瘤71例，死于其他疾病24例。总体5年和10年生存率分别为72.6%和61.5%；影响膀胱肿瘤患者长期存活的主要因素有临床分期(X5)、分级(X6)、手术加膀胱灌注(X11)、手术加化疗(X13)、手术加放化疗(X14)，并据此建立生存分析回归模型：h(t，X)=ho(t)exp(11202X5+0.9124X6-0.9461X11-0.8745X13-1.0247X14)，并对生存5年和5年以上及生存不足5年而死亡的71例膀胱肿瘤患者的预后进行预测，并将预测结果与随访的实际结果进行比较，其灵敏度、特异度及符合率分别为75.0%、82.98%、8082%，两者基本相符。结论 临床分期分级是影响膀胱肿瘤患者的危险因素，能缩短生存时间；手术加膀胱灌注(X11)、手术加化疗(X13)、手术加放化疗(X14)是保护因素，说明综合治疗因素具有重要的临床意义，是影响膀胱肿瘤患者长期存活的主要因素。 【关键词】 膀胱肿瘤 癌 预后 生存率 Cox模型 Regressive analysis of factors influencing the long-term 【Abstract】 Objective To study the factors influencing the long-term surviving of patients with bladder carcinoma.Methods 284 cases of bladder carcinoma，treated from 1980 to 2000 and followed up for 6 months to 18 years，were studied retrospectively and analyzed by Cox regression models in order to testify the relationship between stages，grades，kinds of treatment methods and the long-term surviving of patients.Results 189 cases of the patients still lived at the end of follow up，71 cases died of the bladder carcinoma and 24 cases died of other diseases.The overall 5 and 10 years survival rates were 72.6% and 61.5% respectively.Five factors showed significant relation to the long-term surviving of patients：clinical stage，grade，surgical operation and intravesical therapy，surgical operation and CT，surgical operation and RCT.A predicting equation was established，and the predicting values were calculated according to the clinical data.The predicting and actual values were compared，and the sensitivity，specificity and overall concordance were 75.0%，82.98% and 80.28% respectively.Conclusion Clinical stages and grades are risk factors and surgical operation plus intravesical therapy，surgical operation plus CT，surgical operation plus RCT are protective factors of bladder