Dishonest Research Sells Meds

By: Garry Cooper, LCSW

Reprinted with permission from Psychotherapy Networker magazine. Copyright, Psychotherapy Networker, Inc.

 

Seemingly every week, medication studies announce they’ve discovered the latest and best psychotropic, found a new use for an existing medication, or “proven” that one medication is better than its competition for a specific condition. According to Johns Hopkins University psychiatrist Daniel Safer, however, even therapists who depend on peer-reviewed journals for information may still get hoodwinked by the plethora of misleading research.

Safer’s not a foe of medications, but of purposely misleading research that leads to poor prescribing, a trend he says is accelerating. In the September 2002  Journal of Mental and Nervous Disease, Safer exposes more than a dozen ways in which the nearly invisible hand of pharmaceutical companies skews research to favor their medications. That article follows an eye-opening study in the February 2000  Journal of the American Medical Association, in which Safer and others found a large increase in antidepressant and stimulant prescriptions for preschoolers.

Safer’s 2002 article, a virtual primer on how to critically read medication studies, suggests that therapists who merely read journal abstracts may as well get their medication information from the popular media. Abstracts don’t reveal spurious study designs and self-serving data analysis. Here are some common research tricks Safer warns about:

Manipulating side effects to achieve a desired result. Safer says that at least three pharmaceutical companies, seeking to tout their atypical antipsychotics, compared their medications to 20 mg daily of the older haloperidol, even though the clinically effective level of haliperidol is 4 to 10 mg daily. That virtually guaranteed worse side effects for haloperidol. Sometimes the side effect manipulations are subtler. In a head-to-head comparison, the manufacturers of the SSRI fluvoxamine (Luvox) used a dose at the low end of its clinical effectiveness and a dose of fluoxetine (Prozac) at the high end. Another way to circumvent side effects is to avoid studying them: an SSRI study of 3,000 subjects administered a 23-item side effects questionnaire that never mentioned sexual side effects, a known problem with many SSRIs.

Manipulating measurements and scales. One study of an antipsychotic touted the subjects’ 20-percent improvement on the Brief Psychiatric Rating Scale, not mentioning that this still left them in the “profoundly disabled” range. Another company-influenced study, seeking to establish its medication as an effective alternative to Ritalin, emphasized that both medications resulted in a 25-percent decrease in AD/HD. However, a third of the placebo subjects also improved 25 percent. Researchers also use the element of time to slant their results. If study designers know that their medication acts more quickly than its rival, they’ll run the trial for a shorter period.

Blitzing the journals, or, “This medication must be good because I keep hearing about it.” To inflate the impression of a buzz for their medication, pharmaceutical companies often publish the same study, slightly refocused, in different journals. With their virtually unlimited financial resources, they also run multiple studies of their medication, each insignificantly altered.

In addition, companies also withhold publication of studies that don’t work out to their advantage. A 2002 metanalysis of antidepressants by Irving Kirsch, using pharmaceutical companies’ raw data submitted to the Food and Drug Administration (FDA) rather than their published studies, found that they aren’t much more effective than placebos.

Withholding unfavorable results may occasionally lead to dangerous results. An article in the August 7 New York Times says that GlaxoSmithKline, manufacturers of Paxil, published just one of its nine studies of Paxil’s effects on children and adolescents. But a study by British researchers of all the raw data submitted to the FDA finds that Paxil is ineffective on adolescent depression and poses a substantial risk of suicidal ideation. Although several journals require researchers to state their affiliations and funding, that may still not be enough. The only safe way to read pharmaceutical research is to approach it like any other advertising–read it critically, skeptically, examine the fine print, and if the fine print isn’t there, ask yourself why not.