Why do collaborative research?BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39297.500544.94 (Published 09 August 2007) Cite this as: BMJ 2007;335:304
- Anisur Rahman, reader in rheumatology
Much of the most valuable medical research is done by large teams of people, often collaborating across several centres. Examples include randomised controlled trials that prove the efficacy of new forms of treatment and genetic studies that use clinical data from many hospitals to establish the linkage of genes to specific diseases.
Such collaborative efforts are undervalued by academic institutions, to the extent that it could be argued that clinical academics who wish to thrive should avoid taking part in such collaborations—unless they are a lead author. Without colleagues who are prepared to collaborate without gaining the kudos of leading, however, none of these studies could be done.
Are the disincentives to collaboration worse than before? One disincentive is the increasing complexity of documentation that researchers need to carry out any form of study involving patients. A researcher who signs up to recruit patients into any multicentre project—even if only blood samples are needed, or a review of case notes—will have to submit several forms to the local ethics committee and may have to deal with the research and development departments of several trusts. All of this may require hours of work spread over several months before the research can even begin.
Once the project starts much time may be invested in recruiting participants, collecting samples, and collating information from case notes. In studies that are not funded by drug companies, often no funding is available for staff outside the lead centre. Collaborators at peripheral centres can either do this work themselves, which is laborious, or try to delegate it to junior staff. This could be considered exploitation if the juniors in question are unlikely to be listed as authors in the final paper.
What reward will ensue from this investment of time and effort? When the research is finally published, there will be a long list of authors. Most of the collaborators will be somewhere in the middle of this list. To get some idea of how this outcome will be viewed by academic departments we can consider the “publication score” recently proposed by Imperial College London as a part of the mechanism for assessing academic performance. The contribution of a particular paper to a person's score is calculated by multiplying the impact factor of the journal that published it by an author position weight and then dividing by the number of authors. The author position weight is 5 for the first and last authors and 1 for any other position from fourth onwards. It is immediately obvious that the amount of credit to be obtained from being part of a large collaboration is very small. Although this publication score has been criticised (www.dcscience.net/goodscience/?p=4) and has not been adopted by other institutions, many may feel that it simply formalises a mode of thinking that already exists. This model implies that clinical academics should concentrate on papers with few authors in which they can lay claim to first or last authorship. Anything else is a mark of failure.
Is this too cynical? Why not just accept that collaboration is a virtue in itself without expecting any other reward? If you have helped to bring a new form of treatment to patients, is that not reward enough? Isn't that why we chose to be clinical academics in the first place? Maybe all this is true, but why should involvement in multicentre collaborative research be a cross to bear? Why shouldn't we actually assess whether individuals or departments are successful collaborators and include that assessment when judging their contribution to research? It shouldn't be difficult. Does a person do any collaborative research? What do the lead authors of the collaboration think of that person's contribution? What was the impact of the resulting research?
Whether as researchers or patients we all benefit from the fact that people in different institutions work together. Let us make it more attractive for individuals to choose that option. Otherwise, one possibility is that the only large scale collaborations will be those funded by the drug industry. The industry can pay people to deal with the bureaucracy and data entry needed and can reward institutions financially for taking part in the research. Do we want this to be the only incentive that persuades people to collaborate?
Why not just accept that collaboration is a virtue in itself without expecting any other reward?
Competing interests: AR is a clinical academic who has been involved in six multicentre collaborative projects since 2000.