A study in Journal of the American Medical Association about the effectiveness of anti-depressants has been widely reported. The conclusions, that drug therapies are most beneficial for people with the most serious symptoms, but are no better than a placebo for those with mild symptoms, does not square with the clinical observations of doctors. New York Times columnist Richard Friedman called the results "provocative" and "confusing if not alarming."
Friedman explained what to look for:
"The study is a so-called meta-analysis -- not a fresh clinical trial, but a combined analysis of previous studies. A common reason for doing this kind of analysis is to discover potential drug effects that might have been missed in smaller studies. By aggregating the data from many studies, researchers gain the statistical power to detect broad patterns that may not have been evident before."He cautions about the methodological problems of analysis of 718 people selected from six of a possible 23 studies comparing a drug to a placebo. The six studies looked at comparisons to two anti-depressants (Paxil was one), and may give rise to premature conclusions. Another caution about jumping to conclusions from the study comes from another psychiatrist, Daniel Carlat, who observes that placebo-controlled trials "are rarely generalizable to the real world." He explains how candidates for these studies are selected from clinical experience.
More studies
Another noteworthy depression study appeared in this month's Archives of General Psychiatry with controversial information about disparities based on racial and ethnic characteristics in access to and therapy with psychotherapy and psychopharmacologic tools. The NIMH sponsored study found that "African Americans and Mexican Americans were least likely to receive any care or care consistent with [APA} practice guidelines. And an article in the Annals of Psychotherapy about the risk of suicide for teens concludes that anti-depressants reduce hospital readmissions.
