How to assess for masking/blinding

To assess this domain, you’ll need to go through each outcome and decide:

1. Was blinding necessary to prevent bias?

If death is your outcome, blinding may not be necessary; but if pain is your outcome, the pain rating may be open to interpretation, so blinding would matter.

2. Who needed to be blinded to prevent bias?

For example, if participants assessed their own pain level, they should have been blinded. But if doctors described their patient’s pain level, they should have been blinded too.

3. Was that blinding likely to be broken?

Maybe participants were not told what they received, but it’s easy to figure out who had the treatment and who had the placebo.

Based on these decisions, you should label each outcome as having high or low risk of bias. If you’re not sure, you should label an outcome as unclear.

For reference, here’s a chart about blinding scenarios that indicate a “low risk of bias” and those that indicate a “high risk of bias.”

Low Risk of Bias High Risk of Bias
Blinding ensured and unlikely to be broken Blinding attempted but likely to be broken and outcome likely to be affected
Blinding not done, but unlikely to affect the outcome or the management of the patient Blinding not done and outcome likely to be affected

But, beware! It’s not enough for study authors to say they used blinding or conducted a “double blinded” study. These terms are often defined in different ways.

We’ll talk about that on the next page.