Jonah Lehrer wrote an intriguing article in The New Yorker magazine on “the decline effect.” The article discussed a phenomenon of scientific inquiry in which early promising results tend not to hold up over time. Although this occurs in an array of scientific disciplines, the article began with a discussion of neuroleptic drugs. The author described a 2007 meeting during which clinical researchers and representatives of the pharmaceutical companies discussed the fact that more recent studies were reporting less robust efficacy of these drugs than had been found in earlier studies.
“A recent study showed an effect that was less than half of that documented in the first trials, in the early nineteen-nineties. Many researchers began to argue that the expensive pharmaceuticals weren’t any better than first-generation antipsychotics, which have been in use since the fifties. ‘In fact, sometimes they now look even worse,’ John Davis, a professor of psychiatry at the University of Illinois at Chicago, told me. “
Researchers at the FDA published a report on this in The Journal of Clinical Psychiatry and a link to an article about this study was posted on this website a few weeks ago. The researchers were trying to understand these declining results by comparing data from studies done from 1991 to 1998 with those done between 1999 and 2008. They also compared results from studies done in North America to those done elsewhere.
Indeed, the more recent studies show a smaller difference between active treatment and placebo groups than the earlier ones. But, as I wrote in an earlier blog on a new antipsychotic lurasidone, what was most striking to me, is just how small the treatment effect was – even in the earlier studies.
The researchers focused on the results of the PANSS (Positive and Negative Symptom Scale). This is a 30 item questionnaire that measures symptoms such as voices and delusions (so called positive symptoms) as well as apathy (so called negative symptoms). Each item is rated on a scale of 1-7 thus the score can be anywhere from 30 to 210 points. At entry into the study, the average score was about 90 points.
For all of the studies, the average decline in the PANSS in the active treatment group was 15 points and in the placebo group 6.4. The effect size is the difference between these numbers and that is about 8 points.
This is not a big difference.
The prevailing belief among psychiatrists is that these drugs are extremely effective.
This is a belief that needs to be reexamined.