Scientists using animals in research have a responsibility to ensure that their studies are appropriately designed, conducted, analysed and reported so that they impartially and robustly answer the question they are intended to and truly add to the knowledge base. Failure to fully embrace this responsibility wastes animals and other resources and ultimately undermines the ‘contract’ with the public that permits the use of animals in research.
Unfortunately there is a large body of evidence, including from the NC3Rs (the National Centre for Replacement, Refinement & Reduction of Animals in Research) to show that many animal studies are poorly designed, analysed and reported and that this has significant implications in terms of reproducibility and the translation of findings into potential clinical benefits. For example, failure to translate pre-clinical findings to the clinic in multiple sclerosis has been attributed to shortcomings in the design of animal experiments.1 Here at the NC3Rs, we have developed a new exciting online tool which is designed to tackle this problem – the Experimental Design Assistant (EDA).
In 2010, we published the ARRIVE guidelines, to improve the reporting of in vivo research. To complement this we have developed a resource to help researchers improve the design and analysis of their experiments using animals, avoiding all of the previously identified pitfalls such as failure to avoid subjective bias, using the wrong number of animals and issues with randomisation. The EDA is an online web application, with an associated website which contains a wealth of information and advice on experimental design. It was developed in collaboration with an expert working group of in vivo
scientists and statisticians from academia and industry, and Certus Technology, a team of software designers specialised in innovative software for the life sciences. In the EDA, experiments are represented as diagrams, with nodes and links, which break the experiment into its component parts. This is a novel approach and it is unusual to think about an experiment in this way. For this reason, it might take a little time to get used to the EDA.
The benefits of EDA
Using the EDA has many benefits:
- Building a diagram from basic components leads to a greater understanding of experimental design. The visual representation enables the user to clarify their thoughts and allows quick identification of elements which might impact on the robustness of the findings.
- The system provides feedback on the diagram which helps improve the design. In the ‘back end’ of the system we have built a dataset of rules, which trigger prompts in specific situations. The prompts provide information to help the user optimise their experimental plan and, for example, highlight the implications of some of the design choices made or point out issues with the internal consistency such as where two variables are confounded. We will continue adding rules into the dataset so that the system can recognise more subtle issues and increase the relevance of the feedback for more complex experimental designs.
- The EDA includes dedicated functionalities to provide support for randomisation, blinding and sample size calculation.
- The diagram forms an explicit description of the experimental plan, which can easily be shared within the lab or with collaborators. This transparency helps communication and allows critical review by others.
We are really excited to have launched the EDA and look forward to hearing what you think about it. We have done extensive testing with scientists from a whole range of disciplines, institutions and career stages and the feedback so far has been fantastic. As with any new software, the first release of a system may have some teething issues. If you do encounter any problem, please do let us know at eda@ nc3rs.org.uk. Your feedback will help us ensure that the system improves and evolves according to your needs.
Nathalie Percie du Sert NC3Rs
NC3Rs is a UK-based scientific organisation dedicated to replacing, refining and reducing the use of animals in research and testing (www.nc3rs.org.uk). This article is adapted from a post that originally appeared on the NC3Rs’ blog.
1. Vesterinen et al. 2010 Multiple Sclerosis 16 1044–1055