本文针对二分类变量结局指标相对（而非绝对）治疗效果的不一致性。证据本身不会因不同研究结果具有一致性而升级，但可能因不一致而降低质量级别。衡量一致性的标准包括点估计值的相似性、可信区间的重叠程度以及统计学判定标准包括异质性检验和I2。系统评价作者应提出并检验少数几个与患者、干预措施、结局指标以及方法学相关的先验假设以探寻异质性来源。当不一致性很大且无法解释时，因不一致性而降低质量级别是恰当的，特别当某些研究显示有显著益处而其他显示无益甚至有害时（而非仅是疗效大与疗效小的比较）。明显的亚组效应可能不可靠。如果亚组效应满足以下条件，其可信度将会增加：基于少数几个有具体方向的先验假设、亚组比较来自研究内而非研究间、交互检验的P值小、结果有生物学意义。
The paper presents two statistical methods to compare summary estimates of different subgroups in meta-analysis. It also shows how to use Z test and meta-regression model with dichotomous data and continuous data in R software to explain the similarities and differences between the two statistical methods by examples.
To perform a meta-analysis of single nucleotide polymorphism needs to calculate gene frequency. This paper employs allele model as an example to introduce how to calculate gene frequency and display the process of a meta-analysis of single nucleotide polymorphism data using Review Manager 5.3 software.
Subpopulation treatment effect pattern plot (STEPP) method is a method for examining the relationship between treatment effects and continuous covariates and is characterized by dividing the study population into multiple overlapping subpopulations to be analyzed based on continuous covariate values. STEPP method has a different purpose than traditional subgroup analyses, and STEPP has a clear advantage in exploring the relationship between treatment effects and continuous covariates. In this study, the concepts, advantages, and subpopulation delineation methods of the STEPP method are introduced, and the specific operation process and result interpretation methods of STEPP method analysis using the STEPP package in R language are presented with examples.