Half of all EU-funded health research fails to yield papers

More than 50 per cent of European Union-funded health research projects do not result in any publications, researchers have found.

A multinational team led by Mike Galsworthy, a senior research associate in University College London’s research and development office, looked at the outcomes of EU health research spending during its fifth and sixth framework programmes, which ran from 1998 to 2006.

The team searched databases for papers associated with nearly 3,700 health-related grants assigned unique alphanumeric codes. They were able to find papers associated with only 50 per cent of such grants in Google Scholar, and with just 44 per cent in the PubMed life sciences database.

Using the Google Scholar figure, the total value of the grants that did not lead to papers was €570 million (£460 million).

However, this accounted for only 15 per cent of total EU spending on health research, owing to the greater tendency of larger projects to result in papers.

The figures were published in The Lancet in September, and will be discussed in a forthcoming article in the Journal of Health Services Research & Policy.

Dr Galsworthy admitted that his method would not have discovered articles that had not correctly cited the grant code or that were written in languages other than English, but he was confident they were not large in number.

However, he also noted that a large previous study of conference abstracts suggested that it was typical for half not to result in a paper. For this reason, his main conclusion was not that “EU-funded [scientists are] lazier than anyone else” but that the EU should follow the lead of the US National Institutes of Health and establish a system to automatically harvest papers associated with its grants.

He said the European Commission’s only existing knowledge of its research outputs derived from a 2010 survey of more than 10,000 principal investigators, 22 per cent of whose responses were analysed. He described this as “a pretty barmy way to approach literature informatics in the 21st century”, which would not be tolerated as pressure to account for EU spending increased.

He admitted that some grants - such as those for training or commercial partnerships - were not intended to lead to publications, but said a database could flag up those that were.

Dr Galsworthy said researchers should also be required to make datasets publicly available to ensure that EU-funded research “does not fall prey to waste, data misrepresentation or publication bias”.

He said the UK was “much more diligent in publishing than many places in the EU”, and suggested that increased scrutiny of output might promote more of a “publish or perish culture” across the Continent.

“The EU has traditionally focused more on bureaucracy than output and I personally know of projects that have been lazy with outputs as a result,” he said.

paul.jump@tsleducation.com

http://tinyurl.com/ak53b7r