[richrichie]

I am suspicious of studies that tout “% change in odds” in cases where the odds are pretty low (low enough to be modelled as Poisson) to start with.