Working with a meta analysis, I have to combine odds ratios from two different strata into one stratum, I know that M.H formula of calculating the weighted average of the odds ratios, however the problem lies in the fact that this gives me a weighted average of crude odds ratio. The authors in the paper have already calculated adjusted odds ratios for the two strata I want to combine, so my question is, is there a way by which I can calculate the weighted average of the adjusted odds?
I am not sure I understand your question, but in compositional data analysis we use geometric means. The justification is given by the fact that the sample space of compositional data, i.e. the simplex, has a Euclidean space structure, known as Aitchison Geometry. Weighting is in this framework powering.
Try to look at the metafor package in R.
See also:
Viechtbauer, W. (2007). Con dence intervals for the amount of heterogeneity in meta-analysis. Statistics in Medicine, 26, 37-52.
Wolfgang Viechtbauer: Conducting Meta-Analyses in R with the metafor Package. Journal of Statistical Software, August 2010, Volume 36, Issue 3.
http://www.jstatsoft.org/v36/i03/paper
Just use MetaXL (www.epigear.com) as it allows you to input data as OR and confidence intervals and thus you can meta-analyze adjusted odds ratios
I use the geometric mean, after converting the ratio to a proportion.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics