生存资料的二次研究系列之五：三级随机效应Meta分析模型在预后相关Meta分析中的应用

摘要

医学研究中的结局指标往往不止一个。在涉及长期随访的研究中，时间-事件（TTE）结局往往有多个，如无进展生存与总生存。上述结局指标之间存在关联，分开合并的结果会增加I类错误概率。以三级随机效应Meta分析模型针对多个相互关联的TTE结局进行Meta分析的方法，可以在一定程度上避免上述问题。%Multiple outcomes are often involved in medical research. In studies concerning long term follow-up, generally multiple time-to-event (TTE) outcomes are investigated, for example, progression-free survival and overall survival. These outcomes are correlated with each other, and split-and-merge results will increase the probability of type I error. The 3-level random-effects Meta-analysis model is used in the Meta-analysis on multiple correlated TTE outcomes may prevent above mentioned problems.