Refinement: refining animal experiments to minimise pain, suffering and distress


Experiments on animals can cause them pain, suffering or distress. Many people, including researchers, are concerned about animal welfare. The use of animals in research is regulated by specific legislation in many parts of the world, including the UK. Keeping any pain, distress or suffering to a minimum – referred to as refinement – is core to these regulations.

There are not only ethical and regulatory reasons to consider animal welfare in scientific studies. Pain and suffering can affect the reliability of the data obtained in experiments because they can influence the way an animal behaves and its physiology, for example, how the immune system functions. Refinement is therefore essential for scientific reasons too.

How scientific studies affect animal welfare

Depending on the experiment, an animal may undergo a range of procedures that can cause pain, suffering or distress. These include having blood taken, undergoing surgery, being infected with bacteria or viruses, or ingesting a chemical to test how toxic it is. As well as physical harm, in some studies animals may experience psychological harm; for example, they may be placed in environments that cause them to be anxious or depressed.

Back to top

How researchers can refine their experiments

Image shows two mice in a cage – one is in a cardboard tube and another is under a clear red plastic mouse shelter. The cage also has nesting material and wood chips on the floor.
Animals should be housed in a way that maximises their welfare and gives them the opportunity to perform natural behaviours such as nesting, sheltering and climbing.

There are many ways in which researchers can refine experiments to minimise pain, suffering and distress, such as giving pain relief and providing appropriate housing that meets the specific needs of the animals (e.g. nesting material for mice). They can also set limits on the level of suffering caused by defining what are called “humane endpoints”, which are signs that a procedure should be stopped, pain relief should be given, or the animal should be euthanised. Animals are monitored during experiments by researchers and animal care staff, with veterinary advice sometimes sought.

Back to top

How the NC3Rs helps refine animal experiments

In recent years there has been considerable progress made in refining the use of laboratory animals. The NC3Rs has played an important part in this through the research we have funded and the activities that have been led by our own scientific staff. Our work encompasses all species that are used in research, from mice to zebrafish to monkeys, with a focus on procedures that are commonly used or cause particular concern because of the level of pain and suffering involved.

In some cases, our work applies to animals used in specific types of experiments but in others we focus on refinements that will benefit animals regardless of the study. An example of this is our work on changing how mice are picked up from their cage or during the experiment.

Simple changes benefitting millions of animals

Mice are the most commonly used species in research, with many millions being used worldwide every year. Traditionally mice are picked up by the tail, but research funded by the NC3Rs and others has shown that this is stressful for the mice, causing anxiety that can make scientific findings less reliable and reproducible. Amazingly, the simple change of instead using a plastic or cardboard tunnel or cupped hands to pick up mice can make a big difference. Mice quickly become familiar with the tunnel or hand and find them less stressful than being picked up by the tail, with the animals voluntarily approaching the hand or lining up to get in the tunnel. We have worked hard to promote these refined methods and many institutions worldwide are now changing their practices to stop picking up by the tail.

In the following video, an animal technician demonstrates how mice react to being transferred between cages via a cardboard tunnel after only eight days of habituation. As you can see, the mice are very enthusiastic about getting into the tunnel and are not scared of the technician.

Identifying signs of pain

Identifying pain in animals such as rats and mice can be difficult because as prey animals they instinctively conceal signs of suffering to avoid predation. While this has advantages for survival in the wild, in the laboratory it can be difficult to identify animals that are in pain. This is a problem because it means that it is not easy for researchers and animal care staff to know which animals need pain relief.

Image shows the Rat Grimace Scale poster, which includes five types of rat facial expression (orbital tightening, nose bulge, cheek bulge, ear position and whisper change) and images to show this feature not present, moderately present and definitely present.
Our grimace scale posters include images of facial expressions that indicate an animal is in pain or distress so animal care staff can intervene.

We have funded research showing that mice and rats in pain show distinctive behaviours (for example, back-arching) and facial expressions such as squinting their eyes and bulging their cheeks. These facial expressions can be used to create what are known as “grimace scales”, which can be used to better identify animals in pain. There are now grimace scales available for a whole range of animals, from rodents and rabbits through to sheep and monkeys. As these changes in facial expressions can be subtle, we have produced free training materials to help researchers and animal care staff use the grimace scales, including posters to put on the walls in the animal facilities. We are also investing in a new app that will use artificial intelligence and machine learning to automatically detect changes in the facial expressions of mice, helping researchers and animal care staff identify animals where action is needed, for example, increased pain relief.

Other new technologies are also making a big difference for animal welfare. One example of this is a project we have funded called Rodent Little Brother.

Technological innovations for animal welfare

Mice are social animals that live in groups. Many experiments, however, involve separating the mice and taking them out of their “home cage”. This is often done in behavioural experiments where mice are put in different environments or equipment to investigate the effects of interventions, such as genetic modifications or potential drug treatments on their activity levels, movement and memory. These types of experiments can cause suffering because of the separation of animals from their cage mates and the unfamiliar environments. They are also complicated by the fact the studies are carried out during the working day, but as mice are nocturnal the tests are performed when the animals would normally be sleeping.

Working with our collaborators at MRC Harwell (a major UK facility for mouse genetics research) we have funded a project called Rodent Little Brother to develop a cutting-edge system that allows various mouse behaviours to be monitored 24/7 while the animals remain in their home environment with their cage mates (its sister project, Rodent Big Brother, does the same for rats). In the system, the mice are fitted with a small microchip, so that individual animals can be identified, and a camera records their behaviour without any disruption. The findings have been fascinating, identifying new behaviours and social interactions that had not been seen before, particularly during the night when mice are most active. Importantly, the system allows animals to be closely monitored and signs of pain and suffering (such as seizures) to be picked up much earlier than they would have otherwise been if the animals had been checked only during the day.

The full potential of this clever technology depends on the system being able to automatically identify behaviours and alert researchers and animal care staff to any changes. The extensive video footage recorded by Rodent Little Brother is being used to develop algorithms for key behaviours such as climbing, grooming, sleeping, eating and drinking, however, human input is needed to train these algorithms by annotating the footage. We have worked with our colleagues at MRC Harwell to create a citizen science project called Secret Lives of Mice to help with this huge task. We have enlisted over 6,000 people to analyse our first dataset of over 40,000 video clips, helping us take a big step forward for mouse welfare. Read more about volunteer Emma Robinson's experience of supporting the project on our blog.

Image shows three mice in a cage, a still from a black-and-white video feed. Each mouse is marked with a dot (red, green or blue). The red mouse is climbing while the green and blue mice are together at the back of the cage facing away from the camera.

A still frame from a Secret Lives of Mice video clip. Volunteers were asked to describe the behaviour of each mouse (indicated by a different coloured dot) to help train a computer algorithm to detect these behaviours automatically.

Caring about all laboratory animals

We have focused this overview of the NC3Rs’ refinement activities on mice as they are the most commonly used laboratory species, but we have a comprehensive programme to promote the welfare of other species, including:

  • Minimising the pain and suffering monkeys (also referred to as non-human primates) can experience when used in vaccine research studies, including for COVID-19.
  • Encouraging the housing of dogs in groups rather than singly when used in heart monitoring studies during pharmaceutical development, to avoid the animals being stressed.
  • Championing better pain relief for zebrafish used in research by supporting alternatives to fin clipping for identifying individual animals.

Back to top