Table of Contents
Henmi & Copas (2010)
The Methods and Data
The data come from studies examining the effectiveness of wrist acupuncture point P6 stimulation for preventing postoperative nausea (Lee & Done, 2004).
library(metafor) dat <- dat.lee2004 dat <- escalc(measure="OR", ai=ai, n1i=n1i, ci=ci, n2i=n2i, data=dat)
The contents of dat
are then:
id study year ai n1i ci n2i yi vi 1 1 Agarwal 2000 18 100 20 100 -0.1301 0.1303 2 2 Agarwal 2002 5 50 18 50 -1.6219 0.3090 3 3 Alkaissi 1999 9 20 7 20 0.4184 0.4218 4 4 Alkaissi 2002 32 135 31 139 0.0792 0.0825 5 5 Allen 1994 9 23 10 23 -0.1795 0.3595 6 6 Andrzejowski 1996 11 18 12 18 -0.2412 0.4838 7 7 Duggal 1998 69 122 80 122 -0.3805 0.0697 8 8 Dundee 1986 3 25 12 25 -1.9124 0.5390 9 9 Ferrera-Love 1996 1 30 1 30 0.0000 2.0690 10 10 Gieron 1993 11 30 19 30 -1.0931 0.2871 11 11 Harmon 1999 7 44 16 39 -1.3021 0.2759 12 12 Harmon 2000 4 47 6 47 -0.4531 0.4643 13 13 Ho 1996 1 30 13 30 -3.0990 1.1702 14 14 Rusy 2002 24 40 71 80 -1.6600 0.2294 15 15 Wang 2002 16 50 53 88 -1.1687 0.1394 16 16 Zarate 2001 28 110 25 111 0.1610 0.0995
Random-Effects Model
We can fit a random-effects model (using the DerSimonian-Laird estimator) to these data with:
res <- rma(yi, vi, data=dat, method="DL") print(res, digits=2)
Random-Effects Model (k = 16; tau^2 estimator: DL) tau^2 (estimated amount of total heterogeneity): 0.33 (SE = 0.22) tau (square root of estimated tau^2 value): 0.58 I^2 (total heterogeneity / total variability): 60.96% H^2 (total variability / sampling variability): 2.56 Test for Heterogeneity: Q(df = 15) = 38.42, p-val < .01 Model Results: estimate se zval pval ci.lb ci.ub -0.68 0.20 -3.43 <.01 -1.07 -0.29 *** --- Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
These results match what is reported on page 2978.
Henmi & Copas Method
hc(res, digits=2)
method tau2 estimate se ci.lb ci.ub rma DL 0.33 -0.68 0.20 -1.07 -0.29 hc DL 0.33 -0.51 0.22 -1.00 -0.03
Again, these results match what is reported on page 2978.
References
Henmi, M., & Copas, J. B. (2010). Confidence intervals for random effects meta-analysis and robustness to publication bias. Statistics in Medicine, 29(29), 2969–2983.
Lee, A., & Done, M. L. (2004). Stimulation of the wrist acupuncture point P6 for preventing postoperative nausea and vomiting. Cochrane Database of Systematic Reviews(3), CD003281.