As a Ph.D. student in philosophy, I was required to take a grueling year-long upper-level logic course. The textbook for the course, written by Benson Mates, is unbelievably difficult to read. It is a difficult read for myriad reasons, but the most insurmountable characteristic of the book is that you have to read and understand every sentence perfectly. No words are wasted, and if you are looking forward to a fluffy paragraph to give your brain a break and let your eyes glaze over, you're not going to get it. If your eyes ever glaze over when reading Mates' Elementary Logic, you're going to be helplessly lost once you regain consciousness.
The memories I've suppressed of Benson Mates' Elementary Logic came rushing back to me midway through reading chapter one of Irving Kirsch's The Emperor's New Drugs. This chapter, in which Kirsch describes the scientific process he and his research partner Guy Sapirstein used to compare the efficacy of antidepressant drugs over placebo, gives his readers a crash-course in quantitative meta-analysis. If you don't completely understand the connection between two studies that Kirsch presents or if, for instance, you forget the difference between the technical definitions of "clinical significance" and "statistical significance," you'll have to go back a few pages to refresh your memory. And, to be honest, it doesn't get any easier after the first chapter. I found myself taking copious notes and writing down vocabulary words, not merely because I knew I would eventually be writing a review of his book, but simply to understand the intricate details - all of which are important - that Kirsch presents. That being said, if you have a background in statistics or are willing to make the effort to connect all the dots in The Emperor's New Drugs, you'll be in for the shock of your life.
As mentioned, the premise of the book is to explain Kirsch's meta-analysis (he did three meta-analyses, actually) on the efficacy of antidepressants over placebo in treating depression. The results of Kirsch's meta-analyses show that antidepressants do not, in fact, outperform placebo in relieving depression. In his analyses, Kirsch included data that were published by drug companies and, later, data from clinical trials that were suppressed by drug companies which he requested from the FDA through the Freedom of Information Act. Kirsch's skills as a data analyst clearly lie in being able to statistically dissect collected data and in his not taking trial results at face value: All of the studies he included in his analysis had been used to validate claims made by drug companies that their respective drugs are effective, and were additionally used as a basis for antidepressant drug approvals by the FDA. Kirsch, when comparing the placebo effect to the drug effect, overall, in his meta-analysis, considered a number of other studies that cast doubt on the purity of the drug effects reported by various antidepressant manufacturers.
One germane element of clinical trials that Kirsch highlights, which is not accounted for in clinical trial data results, is the following: When patients are enrolled in a clinical trial, procedures of informed consent require that patients be informed of the possible side effects of the active drug(s) being studied, regardless of whether the patient is given a drug or placebo. All of the drug trials that Kirsch examined were double-blind, meaning that neither the patients nor the medical professionals working directly with them during the trials knew if they were enrolled in the placebo or the drug arm of the study. However, since patients were made aware of the side effects that the active study drug could cause, they had a fairly good idea that they were taking the active drug if they happened to develop side effects after beginning the trial. In fact, it was found that in antidepressant-trial surveys, 89% of patients enrolled in a drug-arm correctly guessed that they were taking the active drug instead of placebo. This, Kirsch argues, must contribute in some way to the perceived drug effect in the trials he analyzed. If patients deduce that they are being given antidepressants and they are told that antidepressants relieve depression, why wouldn't they feel better?
Kirsch presents every imaginable explanation for why antidepressants outperform placebo in trials, and why, in the end, all of these explanations statistically reduce drug effects to insignificance. And that's the point of it all: Through his meta-analysis, which was then replicated by other groups of researchers and was validated by various parties including the drug regulatory agencies of the European Union, France, the Netherlands, and Sweden, Kirsch and his research partner found that antidepressants do not outperform placebo in treating depression. In order to fully understand how he arrives at his conclusion, it is essential to read his book. Nothing short of what he has written will explain his conclusion, objections to his conclusion, and his responses to those objections – remember, every sentence he writes is of utmost importance to his message, although his method of presentation is demanding, to say the least, for a general interest book-reader. For an added bonus, he dedicates an entire chapter to explain why the results of his study discredit the myth that depression is a result of chemical imbalances in the brain, and two other chapters focus on power of the placebo effect. If the results of Kirsch's analyses are sound, antidepressants are merely very powerful placebos.
If you've ever opened up a magazine that includes ads for antidepressants, have spoken to a doctor about depression, or have read any documentation on standard depression treatment, you will find his results to be contradictory to everything you know and have been told about antidepressants. And Kirsch knows it. As he states in his epilogue, "I enjoy rocking boats, especially when they are in need of sinking." And this is a gigantic, multi-billion-dollar boat that he's rocking.
© 2010 Maran Wolston
Maran Wolston is a graduate student in philosophy at the University of Minnesota who teaches college-level philosophy courses in the Minneapolis area.