0
$\begingroup$

I am performing a meta-analysis. Most of my data does not have an odds ratio with a Confidence Interval. The majority of the studies I have included in my meta-analysis only have a p-value for a particular question with a sample size. Can I create a plot with only the sample size and p-values for each study? Is it possible to have a meta-analysis without a Forest Plot based on Odds Ratios?

$\endgroup$

1 Answer 1

1
$\begingroup$

Just to clear up a possible mis-conception here first there is no requirement to have odds ratios to draw a forest plot you just need an effect size with its confidence interval.

I think what you are searching for is the albatross plot first suggested by Harrison and colleagues in an article entitled "The albatross plot: A novel graphical tool for presenting the results of diversely reported studies in a systematic review" in Research Synthesis Methods 2017 (8) pages 281 to 289. Fortunately it seems to be open access here. Basically the plot consists of plotting the transformed $p$-values on the $x$-axis and the transformed sample sizes on the $y$-axis. It is then possible to superimpose contours of constant effect size.

I believe the authors made available code for Stata although I do not use Stata so cannot confirm that. It is also implemented in R in the metap package. Disclaimer: I am the author of that R package.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.