As ornithology is to birds

Today, in another of her cantankerous and directionless "interviews," Deborah Solomon of the New York Times at least got something right. This time, the subject of her bullying is Louann Brizendine, a professor of neuropsychiatry at the University of California, San Francisco (UCSF). Brizendine, who in her picture insists on the heeled ankle bootie, just came out with a book about why all women are catty bitches. (Finally!)

Near the end of the interview, Solomon notes that while Brizendine draws on other scientists' research in writing her book, she hasn't done any research herself. In response, Brizendine delivers this gem:

I don't like doing clinical research because of placebos. In a "double-blind placebo-controlled study," as they are called, neither the doctor nor the patient knows what the patient is taking. I don't want to give patients a placebo. It's cruel.

I am not a neuropsychiatrist, and UCSF isn't exactly tugging at my sleeve to offer me a job. However, I cannot understand how a person with any exposure to medical science can comprehend this little about its advancement.

When a scientist constructs a study testing a new therapy against a placebo, the driving assumption is that it's unknown whether the therapy is better or worse than the placebo. Unless this ambiguity can be proven to an institutional review board, the study won't be approved. If we already know what's best, why risk study subjects' health--and spend money--to study it?

We don't always have to wait until studies officially end to find out whether placebos are better or worse than therapies under investigation: In many double-blinded placebo-controlled studies, investigators "break" the blinding at intervals in order to analyze existing data. They do this to identify cases in which either group is actually sustaining harm by being involved in the study. Sometimes, this results in the study ending early to prevent treatment group patients from getting a therapy that's actually worse than the placebo. (Remember the Women's Health Initiative trial of estrogen and progesterone? It was stopped early when it turned out that the therapy increased the risk of invasive breast cancer.) Sometimes, studies with a clear benefit from the therapy are stopped early so that the placebo patients can have access to the better therapy as soon as possible. (This happens all the time in studies of blood pressure medicines.)

In placebo-controlled trials involving ill patients, patients in the placebo group do not just get sugar pills instead of anti-cancer drugs or CT scans or whatever's being studied. They get the standard of care for whatever illness they've got, while patients in the treatment group get the standard of care plus whatever therapy is being studied. Generally speaking, a study that provides less than the standard of care to its placebo group will not get past an institutional ethics committee.

My point is, placebos are not cruel. Often enough, they're better than the new therapies we try out on people, and there are mechanisms to stop studies early when they're worse. Moreover, without them, we'd have no way of measuring the effects of new therapies--we wouldn't know what really works.

My dad likes to quote Richard Feynman's saying that "the philosophy of science is about as useful to scientists as ornithology is to birds." I'm not saying this woman is a philosopher. But what she's saying here isn't useful to science, and it verges on being harmful to patients.