Ginés P, Quintero E, Arroyo V, Terés J, Bruguera M, Rimola A, Caballería J, Rodés J, Rozman C
Hepatology. 1987 Jan-Feb;7(1):122-8. doi: 10.1002/hep.1840070124.
To investigate the natural history of compensated cirrhosis, 293 consecutive patients without previous major complications (ascites, jaundice, encephalopathy or gastrointestinal hemorrhage) were studied in terms of morbidity (probability of developing decompensated cirrhosis during follow-up) and survival. Patients were diagnosed by liver histology between 1968 and 1980. Median follow-up was 63 months. Decompensation of cirrhosis was considered when a patient first developed one of the major complications of the disease. Ten years after diagnosis, the probability of developing decompensated cirrhosis and the survival probability rate were 58 and 47%, respectively. A multivariate survival analysis (Cox's regression model) using clinical, biochemical and histological data obtained at diagnosis disclosed seven factors that predicted prognosis: serum bilirubin; serum gamma-globulin concentration; hepatic stigmata; prothrombin time; sex; age, and alkaline phosphatase. According to the contribution of each one of these factors to the final model, a prognostic index was constructed that allows calculation of the estimated survival probability. The predicting value of this index was validated by a split sample testing technique.
为了研究代偿期肝硬化的自然病史，我们对293例无既往重大并发症（腹水、黄疸、肝性脑病或胃肠道出血）的连续患者进行了发病率（随访期间发展为失代偿期肝硬化的概率）和生存率方面的研究。患者于1968年至1980年间经肝脏组织学诊断。中位随访时间为63个月。当患者首次出现该疾病的一项重大并发症时，即视为肝硬化失代偿。诊断后十年，发展为失代偿期肝硬化的概率和生存率分别为58%和47%。使用诊断时获得的临床、生化和组织学数据进行的多变量生存分析（Cox回归模型）揭示了七个预测预后的因素：血清胆红素；血清γ-球蛋白浓度；肝脏体征；凝血酶原时间；性别；年龄和碱性磷酸酶。根据这些因素中每一个对最终模型的贡献，构建了一个预后指数，用于计算估计的生存概率。该指数的预测价值通过拆分样本测试技术进行了验证。
