Blinding
The practice of not informing research personnel or data analysts to which experimental group animals have been assigned.
Click here to download a fact sheet on Blinding at the bench.
What is blinding?
Blinding, or masking, is achieved when study personnel are unaware of what experimental group the animals have been assigned to.
Who can be blinded?
All research team members can be blinded. This includes, but is not limited to, animal care staff, personnel administering treatment and assessing outcomes, and data analysts.
Why blind study personnel?
Knowing which group the animals have been allocated to can introduce conscious and unconscious systematic errors. This may include how animals are handled, how outcomes are assessed and how data is analyzed.
When should study personnel be blinded and when can study personnel be unblinded?
Study personnel should be blinded from the beginning of the experiment (study subject allocation) up until the latest possible time point in the study. Ideally, all team members would be unblinded after the data is analyzed.
Study personnel may be unblinded when it is necessary to identify an animal. For example, an injury occurred during when the intervention was administered and the investigators are concerned for the health and safety of the animal. Thus, it is necessary that the investigators become unblinded in order to identify appropriate treatment to provide the animal. In this case, unblinding conditions should be outlined ahead of time and should be reported and explained in the manuscript.
Optimal Practices
You are conducting an experiment analyzing the therapeutic effects of a new treatment on stroke.
To do so, you have allocated rats to receive either a surgery to induce stroke or a sham surgery. Similarly, rats will receive either the placebo or the treatment.
You conduct the experiment according to the following:
- Treatment allocation is outlined on a master list containing rat ID codes and the assigned treatment group.
- Investigators and animal caregivers responsible for treating rats are blind to which groups the rats have been allocated to and do not have access to the master list. One unblinded study personnel does have access to the list and will not disclose the contents to any other lab member.
- When inducing the disease, the study personnel conducting the surgeries would be told by the unblinded team member at the furthest point in the surgery whether or not to induce stroke (after exposing the middle cerebral artery) making reference to the master list.
- Rats are kept in identical cages coded with the rat ID code.
- The intervention and control solutions are indistinguishable (same visible volume, same color). This can be achieved by an unblinded personnel covering syringes with tinfoil so the contents are hidden before handing the treatment to the member responsible for administration.
- Outcomes are assessed by someone who is blinded to which group the animals are allocated to. Data is recorded relative to the rat ID code in a spreadsheet.
- The spreadsheet is shared with a data analyst who only knows the rat IDs and respective collected data.
In this example, all study personnel involved in conducting the experiments and handling the animals are blinded to what treatment the animal has been allocated to receive. This ensures that all animals are consciously and unconsciously treated the same by research personnel.
Suboptimal Practices
You are conducting an experiment analyzing the therapeutic effects of a new treatment on stroke.
To do so, you have allocated rats to receive either a surgery to induce stroke or a sham surgery. Similarly, rats will receive either the control or the treatment.
You conduct the experiment according to the following:
- Rats receiving the control are kept in cages with blue ID tags and rats receiving the treatment are kept in cages with red ID tags.
- A team member administers two treatments that are visibly different and becomes unblinded and aware of which animals are in the same treatment group.
- Outcome assessors are aware of which rats had a stroke and received the treatment. They spend more time exercising these rats instead of the rats that had a stroke and received the control. Because of this, rats that received the treatment score better on the neurobehavioral tests.
Efforts should be made to ensure that the same study protocol is followed for every animal in the study except for group-specific treatments being provided. This is because preconceived notions about the drug may introduce subtle differences in the way research personnel assess, maintain, or exclude animals.
Resources and Tools
Publications outlining importance of blinding in preclinical research:
- Bebarta, V., Luyten, D. and Heard, K. (2003), Emergency Medicine Animal Research: Does Use of Randomization and Blinding Affect the Results?. Academic Emergency Medicine, 10: 684-687.
- Bello, S., Krogsbøll, L. T., Gruber, J., Zhao, Z. J., Fischer, D., & Hróbjartsson, A. (2014). Lack of blinding of outcome assessors in animal model experiments implies risk of observer bias. Journal of Clinical Epidemiology, 67(9), 973–983.
- Karanicolas, Paul J et al. “Practical tips for surgical research: blinding: who, what, when, why, how?” Canadian Journal of Surgery. 53,5 (2010): 345-8.