Just enter your email
Have you considered using meta-analysis software?
I2 is automatically calculated in Revman (free program) and makes forest plots, quality assessment tables & figures: http://tech.cochrane.org/revman
What was I2 when combined?
You can convert p-values etc. pretty much anything to standardized mean differences using a free online effects calculator (courtesy of David B. Wilson):
I’d probably stick with RR and not OR in this instance, assuming you looked at RCTs: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1114127/