A systematic review provides a summary of the data from the results of a number of individual studies. If the results of the individual studies are similar, a statistical method (called meta-analysis) is used to combine the results from the individual studies and an overall summary estimate is calculated. The meta-analysis gives weighted values to each of the individual studies according to their size. The individual results of the studies need to be expressed in a standard way, such as relative risk, odds ratio or mean difference between the groups. Results are traditionally displayed in a figure, like the one below, called a forest plot. 系统性回顾提供一定数量单个研究的数据汇总。如果这些研究的结果相似,那么可以使用荟萃分析(一种统计方法)综合所有单个研究的结果并给出一个总体的概括性评估。荟萃分析会根据单个研究的样本数量给与不同的权重。单个研究的结果应该用标准化的方式进行描述,例如在各组之间使用相对风险,比值比或者平均标准差等专业标准。结果通常用图表的形式展现,例如下面这张森林图。  The forest plot depicted above represents a meta-analysis of 5 trials that assessed the effects of a hypothetical treatment on mortality. Individual studies are represented by a black square and a horizontal line, which corresponds to the point estimate and 95% confidence interval of the odds ratio. The size of the black square reflects the weight of the study in the meta-analysis. The solid vertical line corresponds to ‘no effect’ of treatment - an odds ratio of 1.0. When the confidence interval includes 1 it indicates that the result is not significant at conventional levels (P>0.05). 这张森林图描述了5个评估某假设性治疗对死亡率影响的荟萃分析结果。单个研究用黑色方块表示。平行线(横线)代表估值点以及比值比的95%可信区间。黑色方块的大小反映在单个研究在该荟萃分析中的权重。实心的垂直线代表治疗效果为零,也就是比值比为1。当可信区间包括1时,这表明跟传统方法相比,结果并不具有显著性(p值大于0.05)。因此,黑色方块越偏向左侧,则越有利于该假设性治疗。黑色方块越偏向右侧,则越不利于该假设性治疗。 The diamond at the bottom represents the combined or pooled odds ratio of all 5 trials with its 95% confidence interval. In this case, it shows that the treatment reduces mortality by 34% (OR 0.66 95% CI 0.56 to 0.78). Notice that the diamond does not overlap the ‘no effect’ line (the confidence interval doesn’t include 1) so we can be assured that the pooled OR is statistically significant. The test for overall effect also indicates statistical significance (p<> 图表下方的钻石形代表所有5个研究95%可信区间的比值比汇总结果。在这个例子中,表明该治疗方法能够降低34%的死亡率(比值比为 0.66, 95%的可信区间在0.56和0.78之间)。注意,钻石形并不和“效果为零”的垂直线相重叠(其可信区间并不包含1)。所以我们可以确定总体的比值比(OR)达到统计学上的差异。对整体治疗效果的检验也达到统计学差异(p值小于0.0001) Exploring heterogeneity对异质性的说明 Heterogeneity can be assessed using the “eyeball” test or more formally with statistical tests, such as the Cochran Q test. With the “eyeball” test one looks for overlap of the confidence intervals of the trials with the summary estimate. In the example above note that the dotted line running vertically through the combined odds ratio crosses the horizontal lines of all the individual studies indicating that the studies are homogenous. Heterogeneity can also be assessed using the Cochran chi-square (Cochran Q). If Cochran Q is statistically significant there is definite heterogeneity. If Cochran Q is not statistically significant but the ratio of Cochran Q and the degrees of freedom (Q/df) is > 1 there is possible heterogeneity. If Cochran Q is not statistically significant and Q/df is < 1="" then="" heterogeneity="" is="" very="" ="" unlikely.="" in="" the="" example="" above="" q/df="" is=""><1 (0.92/4="0.23)" and="" the="" p-value="" ="" is="" not="" significant="" (0.92)="" indicating="" no=""> 异质性可以用“eyeball”检验以及其他正式的统计学检验,例如Cochran Q检验来评估。用“eyeball”检验时,我们试图寻找已具概括性评估的各研究的可信区间的重叠情况。在上述的例子中,垂直的虚线穿过总体比值比并跨过各个研究的平行线,这表明这些研究之间是具有同质性。异质性也可以通过Cochran Q检验来评估。如果Cochran Q检验的结果具有统计学差异,那么各研究之间具有异质性。如果Cochran Q检验没有统计学差异但是Cochran Q和自由度之比大于1,表明之间可能存在异质性。如果Cochran Q检验没有统计学差异但是Cochran Q和自由度之比小于1,表明之间存在异质性的可能性很小。在上面的这个例子中Q/df 的比值 <1 (0.92/4=""> Note: The level of significance for Cochran Q is often set at 0.1 due to the low power of the test to detect heterogeneity. 注意:因为Cochran Q检验对于检查异质性的能力偏弱,故该检验的差异水平经常设为0.1。 |