Study: Medical Journals Don't Give Doctors Full Picture on Psychiatric Drugs
Publication bias can blur distinctions between effective and ineffective drugs
Helen Silvis Of The Skanner News
March 20, 2012Doctors learn which treatments work best by reading about the latest drug research in medical journals. But the journals don’t publish every drug study, and they prefer to report studies that show positive results from drugs.
That’s a problem, says researcher Erick Turner, M.D., an assistant professor at the OHSU School of Medicine, because doctors are not told which drugs work better than others.
Turner’s research team today released a study of antipsychotic drug research in the peer-reviewed journal PLoS Medicine, titled “Publication Bias in Antipsychotic Trials: An Analysis of Efficacy Comparing the Published Literature to the US Food and Drug Administration Database.”
Antipsychotic drugs are used to combat symptoms of a range of mental illnesses, including schizophrenia and bipolar disorder. The study looked at the newer antipsychotics. They include Abilify and Seroquel, which have been marketed directly to consumers as cures for severe depression, as well as the drugs: Risperdal, Fanapt, Zyprexa, Invega, Consta and Geodon.
The research team found journal articles on antipsychotics showed only a slight bias toward positive studies. In fact, most of the studies, 20 of 24, did have positive results. But the research also revealed that one drug worked less well than the others, yet journal articles did not offer doctors that important detail.
“As far as the FDA (Food and Drug Administration) is concerned, so long as a drug beats the placebo, then it is effective,” Turner says. “But I think doctors would like to know that Fanapt was actually statistically inferior to the others.”
In 2008, Turner led a research team that found antidepressant drugs were getting a positive spin. His research, published in the New England Journal of Medicine, showed that medical journals were biased toward publishing studies that report positive effects.
That matters, because about half of all antidepressant drug studies found no statistically significant benefits for patients, compared to placebos (sugar pills).
“According to the published literature, the results of nearly all of the trials of antidepressants were positive,” that study reported. “In contrast, FDA analysis of the trial data showed that roughly half of the trials had positive results.”
The bias toward positive studies was less in this year’s antipsychotic study than in the 2008 study on antidepressants. Turner says it’s probably because antipsychotics are generally more effective than antidepressants.
“When you compare between drug classes and use FDA data, it's clear that, overall, antipsychotics are more effective than antidepressants. But when you rely on the data in medical journals, the difference between these two drug classes is obscured,” Turner said.
Drug companies must report all their studies to the federal Food and Drug Administration, and their results are published online. But Turner says both the medical journals and the FDA could do a better job.
The FDA posts all the drug studies on its website, but finding that information takes many page clicks even if you know where to look. The search function won’t help doctors much either.
“The search doesn’t work, and it takes many steps to find information,” says Turner, who used to work for the FDA and knows his way around the site. “I think it’s just a stone-age way of doing things. They are just not thinking about the user on the other end.”
Hint: Information about research studies is listed in the drug approval history.
And if you want to know about a drug that was approved before 1996, you need to submit a Freedom of Information request and wait till it’s returned. Doctors and patients would benefit if the studies were easily available online, Turner says.
As for the medical journals, Turner says they should take a step back and look at study protocols – the plans researchers make before they start. When researchers change those protocols, it’s often because the results look better with a different protocol.
“They could look at the protocols and ask ‘Is this good science?’ and if it is, they should commit to publishing regardless of the outcome.”
Search for drug trials by drug name here.