mantelhean.test r语言_Meta分析常用教程：R语言

> funnel(metaor)

方法2

> library(meta) > metaor1=metabin(a,n1,c,n2,data=sample1,sm="OR",comb.fixed = TRUE,comb.random = FALSE,studlab = study) > metaor1 OR 95%-CI %W(fixed)wang 2001 0.3096 [0.1326; 0.7227] 23.5li 2009 0.1964 [0.0378; 1.0198] 8.9zhan 2013 0.4056 [0.1646; 0.9993] 17.7he 2006 0.4118 [0.1972; 0.8599] 26.3jiang 2014 0.3434 [0.1522; 0.7751] 23.6Number of studies combined: k = 5 OR 95%-CI z p-valueFixed effect model 0.3514 [0.2366; 0.5218] -5.18 < 0.0001Quantifying heterogeneity: tau^2 = 0; H = 1.00 [1.00; 1.01]; I^2 = 0.0% [0.0%; 1.3%]Test of heterogeneity: Q d.f. p-value 0.84 4 0.9326Details on meta-analytical method:- Mantel-Haenszel method

> forest(metaor1)

方法3

> library(rmeta)载入需要的程辑包：grid > metaor2=meta.MH(n1,n2,a,c,data=sample1,names=study,statistic = "OR")> summary(metaor2)Fixed effects ( Mantel-Haenszel ) meta-analysisCall: meta.MH(ntrt = n1, nctrl = n2, ptrt = a, pctrl = c, names = study, data = sample1, statistic = "OR")------------------------------------ OR (lower 95% upper)wang 2001 0.31 0.13 0.72li 2009 0.20 0.04 1.02zhan 2013 0.41 0.16 1.00he 2006 0.41 0.20 0.86jiang 2014 0.34 0.15 0.78------------------------------------Mantel-Haenszel OR =0.35 95% CI ( 0.24,0.52 )Test for heterogeneity: X^2( 4 ) = 0.84 ( p-value 0.9326 )

> plot(metaor2)