Setting better standards in reporting – and doing – animal research
Last week I gave a talk at the British Association for the Advancement of Science (BA) BA Festival of Science as part of a session organised by CAMARADES, a collaboration that provides a supporting framework for groups involved in the systematic review and meta-analysis of data from animal studies in experimental stroke. The basic premise of the session was that animal experiments, in stroke research at least, are not being done to the level of rigour that we have come to expect of human studies: for example sample size calculations, true randomization to experimental groups and blinding of those who assess outcomes to the experimental group are all rare. The problem is that studies that don’t adhere to such best practice are more likely to come up with positive findings. Other speakers discussed best practice in human studies, the specific problems in investigating stroke, and other perspectives on the problem. The issue I was speaking on is what journals can do here and I summarised this in two points:
1. Make sure that what’s reported is the truth, the whole truth, and nothing but the truth
2. Make all the studies available for all to read and reuse.
For the first point we need to have a record of what research is being done, such as the registration that now happens for all clinical trials; good reporting guidelines – such as CONSORT and its offspring for clinical trials; and – this is where journals especially have a crucial role – we need to ensure that there is a place for all papers whether the findings are new and exciting or simply need to be recorded. PLoS ONE is one place for these latter studies. The second point comes down to the need to ensure that all these studies can be read and assimilated either by humans or machines into the totality of evidence; open access to the published literature then becomes an absolute necessity.
So the answers seem fairly straightforward but are those that do animal experiments and the journals that report them ready to accept that better standards are needed?
This seems like a valuable concept. Although prospective IACUC approval is intended to mitigate against inappropriate use of experimental animals, perhaps a comprehensive registry (daunting even to consider!) could add another layer of accountability to the process. If purchasers or breeders of animals were responsible for registering them and reporting ultimate disposition / fate, we might have a database of experimental data (whether or not the results were deemed publishable), as well as some means to monitor the ethical use of these animals.
Would you consider each and every animal purchased for experimental use to be “enrolled” in a trial? How would you monitor animals bred in-house? If such a comprehensive enrollment isn’t used, how can you ensure that all experiments are included in the database?
I couldn´t agree more in the need of better standarisation of studies with animals an the necesity of making everything public. I hope researchers cope with this for the outcomes of these studies (good or bad results) help science grow.
Submited by :