Burroughs A K, Mezzanotte G, Phillips A, McCormick P A, McIntyre N
Academic Department of Medicine, Royal Free Hospital, London, United Kingdom.
Hepatology. 1989 Jun;9(6):801-7. doi: 10.1002/hep.1840090602.
The importance of variable time of entry for analysis of survival following variceal bleeding has recently been disputed. In a study of 194 cirrhotic patients with bleeding esophageal varices in whom 2-day mortality was 3%, statistically significant differences in both survival and rebleeding rates were obtained by shifting the starting point for analysis of survival by 2 weeks following admission to hospital or by 5 days for the analysis of rebleeding. In addition, the curve of hazard function for death or for failure to control bleeding following admission clearly showed that any change in entry time in a study of variceal bleeding would introduce bias and alter survival or rebleeding rates. Thus, the starting point for analysis following variceal hemorrhage is an important confounding variable when calculating both survival and rebleeding. It should always be taken into account, particularly in clinical trials, which are often performed in centers where patients are referred from other hospitals at different times following bleeding.
对于分析静脉曲张出血后的生存情况而言，可变的进入时间的重要性最近受到了质疑。在一项针对194例肝硬化合并食管静脉曲张出血患者的研究中，2天死亡率为3%，通过将入院后生存分析的起始点推迟2周或再出血分析的起始点推迟5天，在生存率和再出血率方面均获得了具有统计学意义的差异。此外，入院后死亡或出血控制失败的风险函数曲线清楚地表明，在静脉曲张出血研究中，进入时间的任何变化都会引入偏差并改变生存率或再出血率。因此，在计算生存率和再出血率时，静脉曲张出血后的分析起始点是一个重要的混杂变量。应始终予以考虑，尤其是在临床试验中，因为临床试验通常在不同中心进行，而患者在出血后的不同时间从其他医院转诊而来。
