Studies Drug Companies Don’t Want You to See

*The following is an excerpt from an article “The Medicating of America,” published in Living Safer Magazine and The Legal Examiner.

Before a pharmaceutical drug can be sold in the U.S., it must first be approved by the Federal Drug Administration (FDA). Part of this process is requiring both laboratory and animal testing to understand the efficacy, or effectiveness, of the drug as well as reveal any potential risks or unwanted side effects. One might think that the FDA commissions such tests themselves. This is rarely the case, although test results must be submitted to them during the application process. No matter. Surely these tests are carried out by unbiased sources that report all significant findings, right? Wrong again, in many cases.

In 1997, in an effort to provide healthcare professionals and the public easy access to information on publicly and privately sponsored clinical studies, Congress passed new legislation that allowed for the creation of a website called clinicaltrials.gov. Overseen by the National Institutes of Health, it acted as a registry for the results of many studies. In 2007, Congress expanded the submission requirements, mandating that medical researchers release results of all trials within a year of a study’s completion. Many editors of major medical journals decided to only publish study results that appeared on the website.

At about the same time, an FDA medical officer by the name of Erick Turner led a research review that was looking to determine which studies of antidepressants such as Prozac, Paxil and Zoloft got published and which didn’t. The findings were published in The New England Journal of Medicine in January 2008, and revealed some interesting anomalies. For starters, Turner found that of the clinical trials which were published, positive findings popped up 94 percent of the time—the indication being that the drugs were effective in treating depression.

Yet, when Turner and his team looked at all studies that had been completed, published or not, they saw a different result. This time, it was a draw. Half the studies showed positive results and half the studies showed negative results; meaning that there was only a 50/50 chance that the drugs would have the desired effect.

So almost a decade ago, Turner’s group put forth the idea of “publication bias” when it comes to pharmaceutical testing—the belief that most people are only aware of the studies that are published, and those are often the reports that only display the desired results.

Understanding that this study occurred almost a decade ago, one might believe that the exposure led to a fundamental change in the selection and publication of clinical trials. Yet in 2015, a follow-up report was published in JAMA Psychiatry that determined 43 of 45 articles (96 percent) published offered positive conclusions. Additionally, the study showed that a “positive” study was five times more likely to be published when compared to trials that did not offer a positive outcome. Ultimately, the study concluded that various reporting biases were present in testing the efficacy of FDA-approved, second-generation antidepressants, which led to an overly positive representation of significant findings in the scientific literature.