Prindiville T, Miller M, Trudeau W
Am J Gastroenterol. 1987 Jul;82(7):655-9.
The purpose of the study was to evaluate prognostic indicators in patients receiving endoscopic injection sclerosis for bleeding esophageal varices. The results below were obtained from a prospective nonrandomized study in which the patients were subdivided into the following groups (subsets): elective sclerotherapy--10, active variceal bleeding at endoscopy--47, active variceal bleeding with spontaneous resolution at the time of endoscopy--21, and gastric variceal bleeding--nine. The data were analyzed as follows: 14 variables possibly affecting 6-wk survival were evaluated by stepwise logistic regression analysis. Variables examined were subset, age, sex, Child's classification, ascites, comorbid conditions, SGOT, bilirubin, protime, albumin, volume of initial bleed, volume of rebleed, rebleed rate, and Blakemore tube use. Survival curves were also compared using the Kaplan-Meier methodology. The following results were obtained. First, the subsets above had significantly different survival curve patterns with the elective endoscopic injection sclerosis group having the best prognosis and the gastric variceal bleeders a significantly poorer survival. Second, stepwise logistic regression analysis revealed the following four significant variable prognostic indicators: comorbid factors, subset, Child's classification, and serum albumin. Although the latter two variables, Child's class (p = 0.03) and albumin (p = 0.1) had an impact on six week survival, this was no longer seen when subset (p = 0.0124) and comorbid factors (p = 0.003) were taken into account. We conclude that comorbid factors and subsets of variceal bleeding in themselves were the only two prognostic indicators showing a statistically significant association with 6-wk and long-term survival. In contrast, the more usual clinical prognostic parameters, i.e., the physical findings and biochemical test of liver function in patients with cirrhosis and acute variceal bleeding, were of lesser prognostic magnitude.
本研究的目的是评估接受内镜下注射硬化剂治疗食管静脉曲张出血患者的预后指标。以下结果来自一项前瞻性非随机研究，患者被分为以下几组（亚组）：择期硬化治疗组——10例，内镜检查时活动性静脉曲张出血组——47例，内镜检查时活动性静脉曲张出血且已自发止血组——21例，胃静脉曲张出血组——9例。数据分析如下：通过逐步逻辑回归分析评估了14个可能影响6周生存率的变量。所检查的变量包括亚组、年龄、性别、Child分级、腹水、合并症、谷草转氨酶、胆红素、凝血酶原时间、白蛋白、初次出血量、再出血量、再出血率以及使用Blakemore管情况。还采用Kaplan-Meier方法比较了生存曲线。得到了以下结果。首先，上述亚组的生存曲线模式有显著差异，择期内镜下注射硬化剂治疗组预后最佳，胃静脉曲张出血患者的生存率明显较差。其次，逐步逻辑回归分析揭示了以下四个显著的变量预后指标：合并症因素、亚组、Child分级和血清白蛋白。尽管后两个变量，即Child分级（p = 0.03）和白蛋白（p = 0.1）对6周生存率有影响，但在考虑亚组（p = 0.0124）和合并症因素（p = 0.003）时，这种影响不再明显。我们得出结论，合并症因素和静脉曲张出血的亚组本身是仅有的两个与6周和长期生存率有统计学显著关联的预后指标。相比之下，更常见的临床预后参数，即肝硬化和急性静脉曲张出血患者的体格检查结果及肝功能生化检查，其预后意义较小。
