A friend of mine is going through a hard time: on the verge of divorce, hates her job, suffering through depression. And she’s a really clever woman; a physics teacher, in fact. So, when she told me she had started researching her condition―depression―and found a book by a PhD graduate called The Emperor’s New Drugs by Irving Kirsch that said antidepressants aren’t more effective than a placebo, I was shocked. I mean, I was in a room with a former psychiatrist who prescribed antidepressants and found they do work, but that’s just anecdotal, surely?
So I had this strange feeling: a received wisdom about the efficacy of antidepressants which had been attested to by a psychiatrist I knew and a friends’ mother who actually used them and recovered was being challenged by someone I know to be clever citing a book by a PhD whose research was peer reviewed in this area: he’s a lecturer at Harvard Medical School. Now what?
Well, the book isn’t peer reviewed, only the research in the bibliography was researched. I didn’t immediately believe that a Harvard Medical School lecturer and PhD would intentionally write a book saying the contrary to what their peer reviewed science said, because Kirsch wouldn’t be able to claim misinterpretation of his own published work.
But, Kirsch’s peer reviewed research does show that antidepressants are more effective than a placebo. The abstract of a 2008 paper Kirsch was an author of says “Drug–placebo differences increased as a function of initial severity”, and that where it matters―the severely depressed―the authors are in no doubt the antidepressants are statistically and clinically significant (this is where it matters because this is where antidepressants are used as a part of treatment; before this, NICE guidelines recommend therapy).
Not only that, but the Wikipedia page about the book talks about the failures of their statistical analysis. (See the Wikipedia page here.)
So, what’s my point? My point is that a book by a reputable sounding person in a scientific field is still only as good as the peer reviewed literature. It would still be an appeal to an authority to accept Kirsch’s book as an authority on the matter, especially as it doesn’t reflect his research, research on the field as a whole or the consensus of doctors in the field working with the best data available today.