International 3Rs Prize now open for applications. £30k prize (£2k personal award) for outstanding science with demonstrable 3Rs impacts.

NC3Rs | 20 Years: Pioneering Better Science
Guidance

Using masking/blinding in in vivo experiments

Practical guidance and examples of strategies that can be used to mask group allocation in animal studies.

On this page

Why use masking in an animal experiment?

Masking (also known as blinding) is a methodological process where an animal’s allocation to a specific experimental group is concealed from the people running the study, caring for the animals or analysing the data. Researchers often expect a particular outcome for their experiment which can lead to unintentional subtle differences in the way animals from different experimental groups are handled, how treatments are administered, measurements collected and data interpreted. This has been demonstrated in different types of experiments using a range of species including rats [1], pigs [2], cattle [2], chicken [2], ants [3] and flatworms [4]. Masking researchers to which experimental group animals are part of, or the treatment they receive, reduces the risk that researchers subconsciously influence research results. Unmasked studies tend to overestimate the efficacy of treatments and interventions tested [5-7].

Best practice guidelines recommend the use of masking to increase the rigour and reproducibility of in vivo research but implementing it in practice is often not straight forward. In collaboration with researchers from AstraZeneca, we have investigated current practices in various labs in the UK. Our study identified a number of barriers that prevented researchers from using masking in their in vivo studies, from staff resources and operational constraints through to animal welfare concerns or knowledge gaps [8].

Publication: A qualitative study of the barriers to using blinding in in vivo experiments and suggestions for improvement

How to make it work in practice?

Masking requires careful planning and it would not be realistic to expect every animal experiment to be fully masked throughout. Masking is also not all or nothing, most studies can relatively easily use some form of masking at some stage during the experiment. Increasing the rigour and reliability of animal research can be achieved with small incremental changes and when full masking is not feasible, partial masking may be possible.

To reduce the risk of error it is crucial to plan carefully who will be responsible for the different tasks and where the randomisation sequence (which specifies the intervention each animal received) will be kept securely. Working with the animal facility will address many of the barriers our study identified such as staff resources, operational constraints or welfare concerns. You can read about this in more detail in the paper [8].

Examples of masking strategies for animal experiments

There are many strategies that can be used to mask the group allocation or the intervention itself. Implementing masking at this stage of the experiment can allow it to be maintained throughout the experiment. The table below has been reproduced and adapted from [8].

Type of intervention Strategy to mask the intervention or the group allocation
Drug injection – drug and vehicle have no discernible differences An assistant/colleague/student loads and codes the syringes with the unique animal ID based on the allocation sequence, and the researcher therefore injects the animals without knowing the content of the syringes.
Drug injection – drug and vehicle have visible differences As above but opaque syringes are used.
Drug injection – drug and vehicle have different viscosity An assistant/colleague/student/technician with no vested interest in the study loads the syringes and injects the animals based on the allocation sequence.
Surgery – the surgeon needs to know what procedure to perform The surgery assistant prepares sealed envelopes containing one of the types of surgery. The envelope is opened during the surgery, at the step that differs between surgical groups. See case study in Box 2 [8].
Tissue graft of different genotypes Donor tissue is stored and genotyped at the end of the study.
Inoculation with different cell lines A different lab member not involved in the in vivo experiment grows the cell cultures and codes the cell vials before the inoculation.
Intervention has a safety concern and staff administering it need to wear protective equipment Similar protective equipment is used to handle all animals in the study, including those receiving an inoffensive vehicle. The intervention is prepared by an assistant/colleagues/student who codes the intervention with the unique ID of the subject based on the allocation sequence.

Diet

An assistant/colleague/student with no vested interest in the study aliquots the food into bags coded for each individual cage, and the researcher feeds the animals without knowing what the diet is nor which experimental group the cage belongs to.

Different strategies can also be used to reduce the risk of subjective bias during data cleaning and analysis. These are summarised in table five of the publication, along with the strengths and weaknesses and an assessment of their efficacy [8].

References

  1. Rosenthal R and Fode KL (1963). The effect of experimenter bias on the performance of the albino rat. Behavioral Science.Behavioral Science. doi: 10.1002/bs.3830080302 
  2. Tuyttens FAM et al. (2014). Observer bias in animal behaviour research: can we believe what we score, if we score what we believe? Animal Behaviour. doi: 10.1016/j.anbehav.2014.02.007
  3. van Wilgenburg E and Elgar MA (2013). Confirmation bias in studies of nestmate recognition: a cautionary note for research into the behaviour of animals. PLoS One. doi: 10.1371/journal.pone.0053548
  4. Cordaro L and Ison JR (1963). Psychology of the Scientist: X. Observer Bias in Classical Conditioning of the Planarian. Psychological Reports. doi: 10.2466/pr0.1963.13.3.787
  5. Bebarta V et al. (2003). Emergency medicine animal research: does use of randomization and blinding affect the results? Acad Emerg Med. doi: 10.1111/j.1553-2712.2003.tb00056.x
  6. Macleod MR et al. (2008). Evidence for the efficacy of NXY-059 in experimental focal cerebral ischaemia is confounded by study quality. Stroke. doi: 10.1161/STROKEAHA.108.515957
  7. Vesterinen HM et al. (2010). Improving the translational hit of experimental treatments in multiple sclerosis. Mult Scler. doi: 10.1177/1352458510379612
  8. Karp NA et al. (2022). A qualitative study of the barriers to using blinding in in vivo experiments and suggestions for improvement. PLoS Biol. doi: 10.1371/journal.pbio.3001873
Support to use masking in experiments

The Experimental Design Assistant provides guidance and support for researchers to use masking in their experiments, including a dedicated feature that generates the randomisation sequence and emails it to a third party.

The logo of the NC3Rs Experimental Design Assistant (EDA)