Preface content

3Rs at Pirbright

Applying the 3Rs at Pirbright

Two calvesThe principles of the 3Rs, Reduction (in numbers), Refinement (of procedures) and Replacement (with laboratory procedures) were developed over 50 years ago as a framework for humane animal research.

Improvements in veterinary medicines and diagnostics are the result of years of research and much of the time is spent doing research in laboratories (in-vitro). However, the sheer complexity of viral diseases and the hosts’ immune responses to them means that research with animals (in-vivo) is essential to better understand the disease and to develop new methods of control.

At The Pirbright Institute this usually entails working with the species for whose improved health the research is directed (cattle, poultry, sheep and pigs). We also use small numbers of mice and rabbits in our research.

Whilst the number of animals used at the Institute is tiny compared to the millions of animals that benefit from the Institute’s research, we strive to apply the principles of the 3Rs for the benefit of animals used in research and for the quality of the data that they yield.


Pirbright scientists researching African swine fever virus (ASFV) are committed to the 3Rs principles; they are constantly looking for new ways to reduce, refine and replace the use of animals in their research. 

Before beginning any research involving animals, experimental design is carried out in collaboration with statisticians to ensure statistically relevant results are obtained. Vaccine candidates are extensively tested in cell cultures before advancing to the stage of animal experiments to reduce the number of animals needed.

The use of standardised clinical and pathology scoring facilitates comparisons between experiments and thus reduces the numbers of groups which have to be used in each experiment. Continued re-evaluation of results from experiments mean the data collected can be refined and the outcomes of experiments can be predicted based on cell culture and analysis of host responses that correlate with protection.

Cells taken from immune animals are also used to identify those ASFV proteins that are recognised by the host immune system which may be capable of inducing protection. This greatly reduces the requirement for animal experiments to obtain this information.

Foot-and mouth disease virus (FMDV) is an extremely variable virus. Successful control involves identifying the right vaccine to use and demonstrating the efficacy of batches of the selected vaccine. This used to involve vaccination of cattle (a group of cattle for each vaccine to be assessed) which would then be challenged (inoculated) with virus. The Institute now uses serologically based methods that have greatly reduced both the number of cattle that are used and those that are challenged.

Together with mathematical modellers at Glasgow University, researchers are working towards a robust and challenge-free computational based model for the assessment of FMD vaccine efficacy.

Colonies of midges and mosquitoes bred at the Institute to study diseases which can be spread by biting insects were traditionally fed using mice. The Institute has developed a method by which the insects can be fed successfully using artificial membrane based feeding units so that mice are no longer required for this purpose.

Formal statistical review of animal experiments is part of the ethical review process at The Pirbright Institute to ensure that all studies are designed appropriately and have sufficient statistical power to detect effects that are biologically meaningful.

Pirbright's in-house statistican, Dr Simon Gubbins assists researchers with their grant submissions to optimise animal numbers and experimental design.

Experiments are designed based on the magnitude of effect that needs to be detected for a study to be informative (e.g. difference in levels of viraemia; proportion of animals protected by a vaccine), the expected level of variation amongst animals and the power (typically 80%) and confidence (typically 95%) required. In most cases, the required number of animals can be determined by standard methods. For more complex designs, however, group sizes may need to be determined by simulating the statistical model underlying the experiments in computer packages.

Back to Animals in research

Trim content

® The Pirbright Institute 2018 | A company limited by guarantee, registered in England no. 559784. The Institute is also a registered charity.