Reduction is about minimising the number of animals used per experiment, consistent with the scientific objectives. Well-designed and correctly analysed experiments can lead to a reduction in animal use whilst increasing the robustness and reproducibility of the scientific results. Applicants are given opportunities throughout the project licence application to reassure the AWERB and Home Office that the number of animals used is the minimum number that is consistent with the aims of the project.
The project licence application focuses on the following fundamental aspects of good experimental design: the study design (justification for how experimental and control groups were chosen), avoidance of bias (use of randomisation and blinding) and determining sample size (including justification for the chosen effect size). Here we outline a number of resources in this area and how they can be used in project licence applications to address these fundamental concepts.
Resources for study design
Experimental Design Assistant
The Experimental Design Assistant (EDA) is a free-to-use, online resource provided by the NC3Rs to help scientists design robust experiments more likely to yield reliable and reproducible results. The EDA software is accompanied by experimental design information pages that describe the basic components of the design process, all relevant to the project licence application. These include group and sample size calculation, allocation of experimental units to groups (including randomisation) and blinding.
The EDA software provides tailored feedback on the design of individual experiments, with dedicated support for power calculations, randomisation and blinding. Using an appropriate number of animals and employing a rigorous experimental design ensures that the results are useful and prevents waste. Demonstrating your use of the EDA sends a strong signal to the AWERB that your experimental design has been optimised to yield the best possible data from the lowest number of animals. The automatically generated EDA report can be submitted to the AWERB as part of the review process, and EDA diagrams can be supplied as images within the application itself.
Sample size calculation and justification
Justification for the number of animals required is a critical component of a project licence application. This should be demonstrated by providing the values used in your sample size calculation, and the justification for your chosen effect size (i.e. why this size of effect would be of scientific or clinical interest). The MRC has produced a guidance document with worked examples for justifying experimental design and animal numbers, aimed at grant applicants but equally useful for project licence applicants.
For some experiments, it may not always be possible or appropriate to perform a power calculation. In such cases, we recommend seeking advice from the local statistician. To answer some of the most common questions related to sample size calculations, statistician Dr Simon Bate from GSK has written a helpful blog post on how to decide your sample size when the power calculation is not straightforward.
Reporting and reproducibility
The ARRIVE guidelines were developed by the NC3Rs and experts in the field to improve the quality of reporting of in vivo experiments and the reproducibility of animal research. They have been adopted by over a thousand journals and funders, and most research-intensive universities in the UK – hence many researchers include in their project licence application a commitment to using the ARRIVE guidelines when writing up the results of their research for publication.
The ARRIVE guidelines can also be used as a framework for planning research studies. An Explanation and Elaboration document that accompanies the guidelines provides extensive advice on the design of animal experiments and details the rationale for the fundamental experimental design concepts required in project licence applications.
General reduction resources
- The choice of sex of animals is an important experimental consideration with potential scientific and reduction impacts. We give advice on how the choice of sex can impact upon your experimental design in the animal characteristics page of the EDA website.
- Pilot studies can be used to gather information to improve the quality and efficiency of subsequent experiments. We have a resource page on conducting a pilot study that provides guidance on how to carefully plan a pilot study, including logistical considerations and what to do with the information obtained.
- Optimising the production, breeding and use of genetically altered animals is an important means of reduction often overlooked in project licence applications. We have a guidance page on minimising the use of GA mice that links to mouse databases to avoid strain duplication, and includes services for archiving GA lines such as the free MRC Frozen Embryo and Sperm Archive (FESA). Organisations such as MRC Harwell and Jackson Laboratories offer resources and courses on efficient breeding strategies for genetically altered mice.
- Experimental design training opportunities include the FRAME training schools; local experimental design workshops organised by NC3Rs staff; and use of the EDA software and supporting website.
- To support reduction strategies in the design of toxicological or regulatory studies, we have a resource hub which includes information on efficient study designs in this field, identified by NC3Rs and industry experts, along with a bibliography.
Page last updated: 6 July 2020