Solved – Publication bias in meta-analysis of diagnostic accuracy in R

meta-analysispublication-biasr

Are there any implemented methods in R that could be used to assess publication bias in a meta-analysis of diagnostic accuracy (Egger's test, funnel plots)?
One way would be to change sensitivites and specifities to log diagnostic odds ratios (logDOR), run meta-analysis on the logDOR using common rfx-models and then e.g. use funnel() and regtest() from the metafor package. However as I understand the bivariate approach (Reitsma et al., 2005) is to be prefered. But if a run a bivariate linear model, how can I assess publication bias?

library(mada)
data(Dementia)
(fit <- reitsma(Dementia))
summary(fit)

Best Answer

There is an older paper of Deeks et al justifying the use of the diagnostic odds ratio for this purpose:

on this method:

http://www.ncbi.nlm.nih.gov/pubmed/16085191

The work of Deeks et al is not related to the bivariate model. From my limited experience I can nevertheless recommend this approach.

Related Solutions

Solved – How to go about doing Egger’s test for publication bias in a meta-analysis

An option is to use the free statistical software R (link). I generally use the metafor package, but there are also other packages out there that can be used for this purpose (see this page for more information about R packages related to meta-analysis).

An example using the metafor package:

install.packages("metafor") # Install metafor package

library(metafor) # Load metafor package

dat <- dat.bangertdrowns2004 # Use these data as example

res <- rma(yi = yi, vi = vi, data = dat) # Random-effects meta-analysis

regtest(res) # Conduct Egger's test

Regression Test for Funnel Plot Asymmetry

model:     mixed-effects meta-regression model
predictor: standard error

test for funnel plot asymmetry: z = 2.4904, p = 0.0128

?regtest # Description of the function in the metafor package

It might be important to realize that the rma() function by default uses the restricted maximum likelihood estimator (REML) for estimating the between-study variance in a meta-analysis. If Review Manager is using a different estimator for the between-study variance, you may also get slightly different results. You can, however, specify the estimator for the between-study variance in the rma() function. See ?rma for more information.

R Correlation – How to Conduct Three-Level Meta-Analyses to Compare Correlations and Correct for Publication Bias

To compare the A-B and A-C correlations, you need to account for the dependency therein, not just by including random effects, but also for the dependency in the sampling errors. The rcalc() function (see here) helps with with computing those and constructing the proper V matrix. Note that you will need to include the B-C correlations in your dataset (since the covariance between $r_{AB}$ and $r_{AC}$ depends on $r_{BC}$). Also, instead of random = ~ 1 | study.id/es.id, you might want to use random = ~ type | study.id, struct="UN".
Including the standard error as a predictor in your model (in addition to type) would be the straightforward generalization of the 'regression test for funnel plot asymmetry'. See:

Fernández-Castilla, B., Declercq, L., Jamshidi, L., Beretvas, S. N., Onghena, P. & Van den Noortgate, W. (2021). Detecting selection bias in meta-analyses with multiple outcomes: A simulation study. The Journal of Experimental Education, 89(1), 125-144. https://doi.org/10.1080/00220973.2019.1582470

Rodgers, M. A. & Pustejovsky, J. E. (2021). Evaluating meta-analytic methods to detect selective reporting in the presence of dependent effect sizes. Psychological Methods, 26(2), 141-160. https://doi.org/10.1037/met0000300

Best Answer

Related Solutions

Solved – How to go about doing Egger’s test for publication bias in a meta-analysis

R Correlation – How to Conduct Three-Level Meta-Analyses to Compare Correlations and Correct for Publication Bias

Related Question