Drug Efficacy Results Skewed in Clinical Trials
Two Danish scientists — Helle Krogh Johansen MD, DMSc, and Peter C Gotzsche MD, DMSc — were poring over old findings. What they discovered both surprised and disturbed them. In investigating 15 trials comparing an antifungal agent, fluconazole, with a rival agent, amphotericin B, they found that in three large trials comprising 43 percent of the patients, the amphotericin B group were included with patients taking nystatin, which is ineffective in the circumstances.
Further skewing the results in favor of fluconazole was their discovery that in many of the studies, the amphotericin B was administered orally when it should have been taken intravenously.
These so-called scientific trials, in other words, were as rigged as an old Chicago election, as phony as a WWF match, and as crooked as the people who promote boxing. What makes the matter of grave concern to all of us is that the findings implicate the drug’s manufacturer, Pfizer, which provided support for at least 10 of the studies representing 92 percent of the subjects. Pfizer also happens to produce Zoloft, the second-largest-selling antidepressant accounting for $1.8 billion in annual sales.
These revelations are reported in the November 10 Journal of the American Medical Association, and in an editorial of the same issue Drummond Rennie, MD indicates that there may be a lot of this going on, namely:
- In a 1989, article co-author Gotzsche, in performing a meta-analysis of 244 trials of nonsteroidal anti-inflammatory drugs, found 44 multiple publications of 31 of the clinical trials, often with different authors and key discrepancies.
- In 1996, two researchers attempted to perform a meta-analysis on the effects of the antipsychotic agent, risperidone. After research which they described as “vexing” and “intolerably time-consuming,” they whittled down to nine what had been presented as 20 articles and several unpublished reports, many under different names.
- Another meta-analysis looked at 84 trials involving presumably 11,980 patients investigating the effects of an anti-vomiting agent on postoperative patients, which, on closer inspection, turned out to be only 8645 patients in 70 trials. The inclusion of the duplicate data led to a 23 percent overestimation of the effectiveness of the drug.
- In their article, Johansen and Gotsche cite Eli Lilly producing research (showing Prozac in a favorable light) that was only available to them.
The editorial suggests as one solution an open registry of research, but acknowledges that a major stumbling block is the reluctance of the drug companies to come to the party. Two companies – Schering Health Care and Glaxo Wellcome – now register all their clinical trials, but no one is exactly holding their breaths waiting for the others to follow suit.
And so we have what amounts to a conspiracy to deceive us, which should come as no surprise to people in our positions. As street-wise consumers we will read new studies with a more jaundiced eye, noting that in head-to-head matches with a rival drug the scoring is likely to be as fair as the judging of the first Lennox Lewis-Evander Holyfield fight. The same applies to placebos – that is, until the drug companies actually figure out a way to market the things. And should that day ever come, then expect to see the long-maligned placebo finally “winning” a few decisions.
For the Johansen and Gotzsche article, please see:
For the Rennie editorial, please see:
This article was reprinted here with permission. Please visit Suite101.com‘s Depression section, which is overseen by the author and consider becoming a member. The author also publishes an informative weekly newsletter, McMan’s Depression and Bipolar Weekly.
Psych Central. (2016). Drug Efficacy Results Skewed in Clinical Trials. Psych Central. Retrieved on October 26, 2016, from http://psychcentral.com/lib/drug-efficacy-results-skewed-in-clinical-trials/