I have been remiss, given that I never noticed your post here Surlethe. Sorry!Surlethe wrote:The post describing the pictures cites this paper: http://www.ncbi.nlm.nih.gov/pubmed/19593153 . I do not see a pdf download. I have been inferring that "OR" is a measure of risk, probably a regression coefficient. The blog post describing the pictures seems trustworthy, so I'm trusting it that the graphs are meaningful.
Honestly, I can't imagine what OR could mean except odd ratio. It's a diagnostic binary outcome. I can't imagine what they'd be reporting except for the odds ratio at all, let alone another thing that's abbreviated OR. You could easily make it into a regression coefficient just by transforming this all into a logit, but honestly, why bother? It's pretty much industry standard as a reporting measure.