Cancer trial data less likely to be published

The research, in the BMJ, found that only 1 in 5 cancer trials are published in journals.

This compares with around 2 in 5 studies on autoimmune, infectious, respiratory and cardiovascular diseases, which are most likely to be published.

The authors said this may be because cancer studies tend to have fewer participants, and are less likely to be randomised. In addition, negative trial results are more common in cancer studies.

Trial data is important, because it affects drug approval and clinical guidelines and decision-making.

Researchers analysed in detail which features of clinical trials are linked to the likelihood of results publication, looking at disease area, trial size and design.

They looked at studies registered since 2005 on clinicaltrials.gov, a database of trials taking place worldwide. They looked at the same database, along with PubMed, to look for publication of these trials.

Studies which are larger, longer in duration and randomised were more likely to be published in journals.

The pattern of smaller, less well-designed studies being published less often was particularly evident for high-impact journals.

Overall, more trial results were reported on clinicaltrials.gov than in medical journals, but researchers said a lot more data remains to be published.

Lead researcher Magdalena Zwierzyna, of the Institute of Cardiovascular Science at UCL, said:

'The struggle for transparency must continue. There are a lot more clinical data out there that should be made available. We hope that the findings from our research will improve awareness of this problem, and ultimately lead to better access to data from all clinical trials.'

Researchers found that half of all clinical trials required to disclose their results fail to do so. Of 27,835 registered trials required to disclose results, only 15,084 (54.2%) had disclosed their data.

This finding is in agreement with previous studies.

Researchers also found industry studies were more likely to publish trial data compared to non-commercial trials.

The research in the BMJ can be read in full on the BMJ website. Lead researcher Magdalena Zwierzyna has also written a related blog post.