Radiotherapy Clinical Trials Not Being Published?

[Choppy]

I just noticed this article based on a talk presented at the European Society for Radiotherapy and Oncology conference.

Failure to publish trial results exposes patients to risks without providing benefits

An analysis of 802 trials with a primary completion date of before 1 January 2013 showed that 655, or 81.7%, did not publish even a summary result.

It's rather concerning on a number of levels: primarily what's the point of conducting a trial if you're not going to publish the results?

 

[jim mcnamara]

If failure - inconclusive results also - is a factor in not publishing as your link indicates
This study:
https://nccih.nih.gov/research/results/stjohnswort
Found that St John's Wort extract was not effective for treating depression. This was useful to anyone who read it.

There must be something else going on here. And it is a broad problem apparently, called publication bias.
Anyway, here is a link on publication bias and how it affects meta-analysis:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3733739/

 

[Vanadium 50]

Many major journals, e.g. Nature, refuse to publish negative results. That's right - they would have refused to publish Michelson-Morely.

It's gotten so bad that in some fields, there are journals that publish only negative results: e.g. New Negatives in Plant Science. Go botany!

 

[Choppy]

I think I was generally aware that in some cases results aren't published, but the fact that the vast majority of them did not even publish a summary caught me by surprise. The article talks about how in the US since 2007, publication of the results is actually required by law.

 

[Laroxe]

There are lots of things going on at the moment in an effort to restore credibility to a lot of science publication, the so called "crisis of confidence" there is increasing evidence of a whole range of issues ranging from downright corruption, misrepresenting finding, failure in the peer review system, hiding results, changing success criteria part way through trials and so it goes on. The problem is that a great deal of money rests on publication and impact factors, Poppers ideas around research would suggest that negative results are in fact more important than positive results but many people see failure to achieve significant findings as a failure and don't even submit paper, and for them that do, journals want papers with findings, it attracts custom and so there is an estimated 3 to 1 bias in favour of papers showing significant results. Its a bit frightening that medical research is in fact considered to be among the least reliable. A quick read of some of the stuff on Retraction Watches website should increase your anxiety a little more.