Monday, 25 February 2013

ASA 2013: Fiona Godlee on Who is an author and why does it matter?

In the first session at the Association of Subscription Agents and Intermediaries conference in London, Fiona Godlee, Editor-in-Chief of the British Medical Journal discussed the challenges of authorship credit and the move towards contributorship. Who is an author and why does it matter?

Life and research is far more complicated than it was. We now have multi-centre/author/function research going on. There are questions around who did text, the statistics and the lab work. Research has always been, but is even more so today, competitive with researchers fighting for first author place. Then there is the commercial interest, use of journals for marketing by medical/pharmacy industry

The International Committee of Medical Journal Editors has guidelines on authorship credit and have added to these to try and tackle some of the emergent issues. The idea of contributorship revolves around credit and responsibility, being explicit about roles and contributions and revisiting responsibility for complex studies with guarantor role.

The things that bother editors are:
  • bias
  • data manipulation or suppression
  • dupliate publication or salami publication
  • fabriation of data
  • falsification of data
  • gift authorship
  • plagiarism
  • self delusion
  • undeclared conflict of interest
  • wrong observations, analysis or references.

Source: Stephen Lock, BMJ
You can map these across to what constitutes errors in good faith, 'trimming and cooking' or fraud. Stephen Lock at the BMJ mapped these issues (see Fiona's slide, right).

The reality is that the lack of open data is creating major problems with verifying research. Questions around Tamiflu are a case in point, with contradictory statements confusing the picture, which was compounded by the lack of data available for checking.

Concealment of data is a very serious offence. In the US you have to register your trial and publish data within a year of the end of the trial by law. Yet a large proportion do not. The BMJ are trying to do their bit by insisting that clinical trial data must be available to be published alongside the article. 

With the BMJ open data campaign, Tamiflu is the test case. However, industry funded trials are not the only transgressors. Non-commerical trials are just as guilty. But there are signs of compliance. The UK HTA programme is an exemplar - they badger authors, withhold funding and publish on an open access website. This approach has resulted in near 100% compliance.

No comments:

Post a Comment