This website uses cookies to improve your experience. Learn more about cookies and how to manage them.

Animal research: reflections on a busy December

Back to Homepage

Published: 13 Mar 2019
By Elliot Lilley

The use of animals in research raises significant ethical issues and as pharmacologists we need to recognise and deal with these issues. UK legislation such as the Animals (Scientific Procedures) Act 1986 (ASPA) is based around utilitarian principles of harm versus benefit. Research should be conducted in such a way that the benefits (new knowledge, new treatments for human and non-human disease) outweigh any harm to animals (with regard to ASPA, this refers to pain, suffering, distress and lasting harm). For this system to work, benefits need to be realistic and deliverable. However, the widely reported ‘reproducibility crisis’ makes this challenging. How can animal research produce benefit if data derived from research cannot be trusted?

The Society and the British Journal of Pharmacology (BJP) have not shied away from the issue of reproducibility in biomedical research. As I have written previously in Pharmacology Matters, the BJP has been proactive in trying to improve the quality of the research it publishes by setting out clear expectations for authors with respect to reporting of experimental design and animal research. The Society has also championed this issue and in December I participated in a symposium at Pharmacology 2018 organised in collaboration with the UK Laboratory Animal Science Association (LASA).

Pharmacology 2018

The symposium was organised by Dave Lewis and Angela Kerton, and featured some excellent speakers (including myself!). The session 'Tackling the reproducibility crisis in preclinical drug discovery' started with a short, scene setting introduction from Dave who highlighted the recent rise in publications challenging the quality of reported research and the consequent difficulty to reproduce experimental findings. This was followed by an interactive session facilitated by Angela Kerton and Jean-Phillippe Mocho. Delegates were given pens and sticky notes and asked to provide their thoughts on what, in their opinion, was the major cause of poor reproducibility in animal research. These thoughts were collected and analysed for themes which would be reported back to the session at the end. While these comments were being processed, the next speaker – Sara Wells spoke on genetically altered animal technology. Sara gave an excellent overview of how this technology, including CRISPR, can be a powerful research tool if used properly. She highlighted examples of how lack of rigour can lead to problems with interpretation of animal data and inter- and intra-lab variability. Manuel Berdoy gave an entertaining exploration of the impact of poor experimental design on reproducibility. This has been an area of extensive activity in recent years with numerous publications, and dedicated symposia and meetings. Manuel reiterated the importance of randomisation and blinding, a message that all of us should be familiar with. He emphasised how important it is, from an ethical and animal welfare perspective, to ensure that experiments are performed and reported in such a way to maximise the benefit accrued from this research. The next speaker was Jordi Tremoleda who spoke about the challenges of modelling complex multi-systemic trauma injury and gave some recommendations as to how to improve translation. Jordi suggested that mechanistic modelling, focusing on specific mechanism-based questions, can reduce the impact on animals and improve predictive validity. I was the final speaker in the session and was tasked to present the argument that the reporting of animal husbandry, housing and care (including welfare-related provision) is critically important to scientific quality and reproducibility. There are several factors that can impact on how animals respond in an experimental situation, these include: temperature; the time of day; handling methods and, in some cases, the sex of the experimenter. If these potential confounds are known, surely it makes sense that these elements should be reported in the methods section of a paper? I’ll come back to this a little later in this article. The final part of the session had Angela and Jean-Phillippe feeding back regarding the earlier sticky note session. Several key themes emerged and are summarised in table one. The issues highlighted are nothing new, many of the points raised were integral to the drive to produce the ARRIVE guidelines back in 2010. It is perhaps a little disappointing that they still are recognised as problems eight years after the publication and widespread support of ARRIVE.

Table one

Main topic evoked Sub topics detailed on sticky notes
Experimental design insufficient group sizes
lack of power calculation
lack of measures to reduce bias, (for example blinding, randomisation)
Bayesian approach
take gradual approach to accumulated knowledge
Education and Training detection of adverse effects
good dosing practice
human error
Genetics inbred versus outbred strains
genetic drift
Animal research reporting not exhaustive material and methods section for example chronobiology
lack of reporting of adverse effects
lack of reporting of negative data
Standardised protocols lack of standardised protocols
too common standardisation of routine protocols
Standardised environment husbandry, care and handling of animals
safe lab environmental versus real life for behavioural studies
Choice of animal model translatability from animals to humans
difficulty to get human samples
Public perception limitation due to ethical legislation
how to restore trust
explore alternatives

Joint Home Office and Royal Society of Biology Meeting

This brings me on to two other events I attended in late 2018. Firstly, I was invited to co-lead a workshop at a joint Home Office and Royal Society of Biology Meeting in London – Fraser Darling and I were asked to lead a discussion on reporting guidelines and training opportunities to improve reproducibility. Workshop participants agreed that guidelines such as ARRIVE and PREPARE (more on that later…) were very useful but that greater efforts were needed to improve compliance. Clearly this responsibility falls on many shoulders, with authors, establishments, journal reviewers, editors, funding bodies and regulators all involved. One suggestion was that post-graduate education was a good place to start with thesis submissions being examined against ARRIVE (or similar guidelines) to encourage experiments to be designed and reported to the required standards. If this practice is embedded early enough, maybe practice can be changed?

BJP editorial board meeting

The other event I attended was the BJP editorial board meeting in December (I am the animal ethics editor for the journal) where Amrita Ahluwalia (editor-in-chief) had asked me to give an update on ARRIVE compliance in the journal. I’ve been in my role at the journal since 2013 and have seen some significant improvements in the standards of reporting in the journal. Sadly however, some elements of reporting still need work and I illustrated this at the meeting by showing data from a review of published papers I had performed on adherence to the current editorial reporting standards (based on ARRIVE). Whilst information on the animals themselves and their housing and husbandry have improved, reporting of post-surgical analgesia, for example, is sadly lacking in many manuscripts.

The editorial board were encouraged to review the editorial guidelines and to press authors to improve adherence. Amrita and I will also look at the policies and see if we can improve instructions to authors to make our expectations clearer. As many of you may know, ARRIVE is being reviewed and revised at the moment, we await the output from this and will amend the policy to reflect any changes.

Preparing to ARRIVE

Earlier in this article, I mentioned the PREPARE guidelines. PREPARE (Planning Research and Experimental Procedures on Animals: Recommendations for Excellence) was conceived by Adrian Smith who recognised that hoping to improve the quality and reproducibility of animal research at the reporting stage was a little like arguing about the quality of the lock on the door of the stable from which the horse has already bolted. PREPARE is designed to complement ARRIVE and focuses on factors that need to be considered at the planning stage of an experiment. It consists of a checklist and a comprehensive web resource which is regularly updated. Adrian uses the analogy of an airline pilot who, despite having flown many times, will always undertake rigorous pre-flight checks before take-off. Maybe a pre-study checklist or resource like PREPARE could help to improve the quality of animal research?

In conclusion

I’m optimistic that progress is being made on this issue and that the ‘reproducibility crisis’ will be dealt with. However, we cannot be complacent; the culture of scientific research needs to change, and the Society and its members have a role to play, by improving the quality of the papers we write, and the rigour of peer review we perform.


Blog post currently doesn't have any comments.

If you are a British Pharmacological Society member, please sign in to post comments.

Back to Homepage

Published: 13 Mar 2019
By Elliot Lilley

About the author

Elliot Lilley

Elliot is a Senior Scientific Officer within the RSPCA Research Animals Department. He received his PhD in Pharmacology from King’s College and then began a 15 year career in the pharmaceutical industry. In 2012, he joined the RSPCA where his main areas of work relate to promoting refinement, experimental design, and translational validity. Elliot has been a member of the British Pharmacological Society since 1998, a former member of the meetings committee and a current member of the animal welfare and in vivo pharmacology committee. He is animal ethics editor of the British Journal of Pharmacology.

Related Pages