http://www.avaresearch.com/ava-main-website/files/2010040106... puts it on #3, after heart disease and cancer. So, it may be an exaggeration, but not a huge one and certainly not an inexcusable one. It is fairly common to remember that something is way more common (or way rarer) than one would think, and then overestimate the number when trying to produce it later.

http://en.m.wikipedia.org/wiki/Iatrogenesis#Incidence_and_im... has more links, which I didn't read.