Thanks to Doubleday/ Penguin Random House for the ARC. I honestly don't know how to rate this book because I like my popular science books very conversational with loads of anecdotes (think "The Body" by Bill Bryson or "The Man Who Mistook His Wife for a Hat" by Oliver Sacks) and long dry descriptions of research methodology bore me to tears, However, this is not actually a popular science book, and it feels unfair to compare it to one. I found myself skimming through how they did each study to get to the results already. Of course Stephen Leavitt would enjoy this, I thought.

I did learn a lot: about "confounders," about types of ambulances (who knew?), about risk-benefit analyses, about anchoring and left-digit bias. These are all interesting experiments, and cool findings and natural experiments in general are pretty fascinating. I think an e-book is a bad format for this particular title, since by the time you get to the footnotes you've forgotten what they are about, and switching back and forth is too much bother. Random Acts of Medicine was both too long and too short. I wanted at least ten more natural experiments but not all of the details. I'd say they could have put the research methodology in the footnotes but that would have been impossible, given that sometimes they changed the methodology to drill down further or control for different confounders, and in dARC format the footnotes, as I've said, were impossible to use anyway. I'm always happy when anyone points out the problems of medical overtreatment.

I'll say three stars but I'd probably rate a print copy a four in which I could use the footnotes properly. Or not, if I still found the writing pretty dry. ( )