Letters

Animal research

Journal editors could help raise profile of three Rs of animal research

Ethics committees have influenced animal experiments in Sweden

Three Rs should be registration, randomisation, and reviews (systematic)

Would middle ground approach give "added value"?

More funding must go towards finding alternative non-animal methods

Various fallacies persist in vivisection debate

Journal editors could help raise profile of three Rs of animal research

EDITORThe RSPCA (Royal Society for the Protection of Cruelty to Animals) believes that the use of laboratory animals presents an ethical dilemma that affects everyone. Animals are used for a broad range of purposes, and thus it is not possible or constructive to make sweeping statements about scientific validity, justification, or suffering. Issues subject to such polarised debate are inevitably difficult to resolve. In this case, sentient animals capable of suffering are involved.

The human race has a moral imperative to reduce the pain, suffering, and distress that it inflicts on other animals to an absolute minimum. In the case of animal experiments, the principle of the three Rsreplacement, reduction, refinement¹is an excellent practical starting point, and the RSPCA is committed to promoting and supporting it. Greater consideration is now given to implementing this concept, but it is not universally applied (or understood), and much is still to be achieved. One likely reason for this is that the three Rs are often seen as a separate issue and not part of mainstream life sciences.

One way to raise the profile of the three Rs, and push for them to pervade the whole of the life sciences, would be for journal editors to insist that published papers include comprehensive information about the lifetime experiences of the animals involved. This could include their source, transport, husbandry and care, group sizes and structure, enrichment, what was actually done to the animals, protocol refinements, welfare problems (and what was done about them), and the animals' eventual fate (and why).

Journal space is at a premium, but it is possible to convey this information in surprisingly few words. ^{2 3} Such detail would provide guidance for other researchers wishing to improve animal welfare and reduce suffering, as well as giving other interested parties an insight into what laboratory animals actually experience and why. This would help to depolarise the debate and enable people with an interest to contribute to it in an informed and constructive way.

Although the BMJ rarely publishes animal research, perhaps it would consider this kind of comprehensive editorial policy to help steer the debate towards the middle ground so that it can go further, faster.

Penny Hawkins, senior scientific officer. 
Research Animals Department, RSPCA, Horsham, West Sussex HH13 7WN Research_Animals{at}rspca.org.uk

Ethics committees have influenced animal experiments in Sweden

EDITORThe debate in the United Kingdom on animal research¹ and the impending Home Office review of the system² prompted us to scrutinise the influence that ethics committees have had on animal experimentation in Sweden.

In Sweden the review of animal experimentation by animal ethics committees was made compulsory in July 1979. Until 1998 the decisions of the seven regional committees were advisory; after that date they became regulatory. Scientists may, if their application is rejected, appeal to a higher court.

We analysed whether and how the committees have approved the applications subject to minor changes. We explored the main reasons for these conditions and how their frequency varied over time by analysing protocols from the meetings of three ethics committees (Jan 1989-Sept 2000). Altogether 10 432 applications were processed: 7895 were approved, 1907 were approved with modifications, 427 were postponed, 105 were rejected, and the applicants withdrew 98. Some applications required more than one modification before approval. The table shows how the modifications to project applications that were required by the ethics committees before approval may be grouped.

The frequency of modifications required by the ethics committees varied over the years from 10% to 30% (overall average 18%). The three year moving average showed an increasing trend from 16% to 29% during 1999 and 2000. This may reflect a media debate during this period on the ethics committees' role, which was initiated by a government investigation.³

The areas that the ethics committees focused on has varied with time. In recent years pain relief and ensuring the presence of a licensed supervisor seem to have received increasing attention, whereas the interest in introducing earlier humane end points seems to be decreasing.

About three quarters of the modifications may be categorised as refinement with respect to animal welfare, and the modifications requested have had a positive influence on project designs. The systems in the United Kingdom and Sweden cannot, however, be directly compared. The Home Office Inspectorate has played an active and professional part in the United Kingdom in improving project applications through dialogue with individual applicants. The introduction of ethics committees in the United Kingdom might simply result in this important activity being moved from the inspectorate to being a responsibility of an ethics committee.

Jann Hau, professor. 
Jann.Hau{at}bmc.uu.se

Hans-Erik Carlsson, associate professor. 
Joakim Hagelin, research fellow. 
Department of Physiology, Division of Comparative Medicine, University of Uppsala, BMC Box 572, SE-75123 Uppsala, Sweden

Three Rs should be registration, randomisation, and reviews (systematic)

EDITORHaving found no reliable evidence from human clinical research about how much fluid should be given in the resuscitation of bleeding trauma patients, I am working with others on a systematic review of controlled trials of fluid replacement in animal models.¹ To date we have identified about 70 controlled trials.

Some trials are properly randomised but many are not. The potential for selective publication of trials showing more promising treatment effects is considerable, and few if any of the trials set their results in the context of a systematic review of all previous trials. Reducing bias is as important in animal research as in clinical research, and it would seem appropriate to apply the strategies used to improve the quality of clinical research to improve animal research. I therefore propose the following three Rs of animal research:

• Registration: prospective registration of all trials in animals to reduce the potential for publication bias
• Randomisation: proper randomisation to reduce the potential for selection bias
• Reviews: systematic reviews to reduce bias and increase precision.