Kuss O, Gromann C
Institut für Medizinische Epidemiologie, Biometrie und Informatik, Martin-Luther-Universität Halle-Wittenberg, Magdeburger Str. 27, 06097 Halle (Saale), Germany.
Methods Inf Med. 2007;46(6):662-8. doi: 10.3414/me0422.
We reintroduce an exact Mantel-Haenszel (MH) procedure for meta-analysis with binary endpoints which is expected to work especially well in sparse data, e.g., in meta-analyses of safety or adverse events.
The performance of the exact MH procedure in terms of empirical size and power is compared to the asymptotic MH and to the two standard procedures (fixed effects and random effects model) in a simulation study. We illustrate the methods with a meta-analysis of postoperative stroke occurrence after off-pump or on-pump surgery in coronary artery bypass grafting.
We find that in almost all situations the asymptotic MH procedure outperforms its competitors; especially the standard methods yield poor results in terms of power and size.
There is no need to use the reintroduced exact MH procedure; the asymptotic MH procedure will be sufficient in most practical situations. The standard methods (fixed effects and random effects model) should not be used in the sparse data situation.
我们重新引入一种用于二元终点荟萃分析的精确曼特尔 - 亨塞尔（MH）方法，预计该方法在稀疏数据中表现出色，例如在安全性或不良事件的荟萃分析中。
在一项模拟研究中，将精确MH方法在经验大小和功效方面的表现与渐近MH方法以及两种标准方法（固定效应和随机效应模型）进行比较。我们通过对冠状动脉搭桥手术中不停跳或停跳手术后发生术后中风的荟萃分析来说明这些方法。
我们发现，在几乎所有情况下，渐近MH方法都优于其竞争对手；特别是标准方法在功效和大小方面产生较差的结果。
无需使用重新引入的精确MH方法；在大多数实际情况下，渐近MH方法就足够了。在稀疏数据情况下不应使用标准方法（固定效应和随机效应模型）。
