Researchers have shown that there is significant bias in the reporting of breast cancer clinical trials. The group investigated randomised, controlled Phase III clinical trials with over 200 participants between 1995 and 2011. They looked at how the toxicity (side effects) of the drug and the primary end points (main purpose of the drug) were reported in journal publications. The researchers were particularly interested in abstracts of published reports since the workload of doctors often limits the time they have to read the entire trial report.
Of the 164 trials the researcher found, only 30 (18%) were registered on ClinicalTrials.gov. Of these, 7 reported a different end point than what was stated at the time of registration. The primary end point was more likely to be reported in the abstract if the trial results were positive. Over half (56%) of the trials had negative results yet 33% were given a positive spin based on secondary end points – patient outcomes that were not the principal aim of the drug.
Toxicity of cancer drugs is an important consideration for patients and doctors since side effects can have considerable adverse effects on the patient’s general health and quality of life. This study found 67% of trials underplayed the toxicity of the drug, particularly if the trial was successful in achieving the primary end point.
Thanks to Matthew Lam from Breakthrough Breast Cancer for alerting us to this paper. Matthew has written more on the subject in detail in his blog, and there’s a summary of the paper in NICE’s bulletin Eyes on Evidence.