A recent study in the Annals of Oncology (which you can't read online unless you pay big bucks) found this. I found this describing it:
"Reports on Breast-Cancer Drugs Are Often Biased, Research Finds
Researchers often exaggerate the benefits and play down the potentially toxic side-effects of breast-cancer drugs in clinical trials, according to a new study published in the Annals of Oncology. One way they do this is by emphasizing some secondary benefit of the drug, even if it failed to achieve the primary objective (such as, say, lengthening the life of the patient). High toxicity findings frequently go unmentioned in abstracts and conclusions—perhaps, the authors suggest, “to make the results more attractive.”
The researchers looked at 164 randomized clinical trials from 1995 to 2011 and found evidence of bias in reporting toxicity in 67 percent of those trials. While they note that the pharmaceutical industry is “increasingly influential” in the sponsorship of such trials, they didn’t find any association between industry involvement and bias.
So what do the authors propose be done? Journal editors, regulators, and reviewers should apply “a critical eye to trial reports” to ferret out instances of bias, they say—implying that they have been failing in that responsibility up to now."
Well is anyone else shocked here? First we must insist researchers take lie detector tests. Next markers are not allowed to put their spin on any results. Finally, editors, regulators and reviewers should smarten up.
Doh! Really. This does not surprise me at all. I don't put blame on editors, regulators and reviewers. I am sure they are working with what is presented to them. Is it their job to decipher research? Unless they are scientists I am not sure they are qualified. No the onus goes back to the researchers who run the studies to make sure the data is presented correctly.