Blinding
- Concept
Hiding who receives which treatment, so that the expectations of patient, doctor or assessor do not change the results.
What it is
Blinding (masking) hides from one or more people in a study which intervention each participant receives. It stops expectations from changing patients' behaviour, the care clinicians give or the way the outcome is measured. Single-blind means the patient does not know; double-blind means neither the patient nor those who treat or assess know; triple-blind adds the person who analyses the data. A placebo that looks identical is the classic tool for achieving it.
Worked example
- Trial of a cream for knee pain: 200 patients, 100 on active cream and 100 on placebo in identical tubes coded 1 and 2.
- In an earlier open pilot (everyone knew their group), 40 of 50 on active cream (80%) and 20 of 50 on nothing (40%) reported improvement.
- Difference in the open pilot: 80% − 40% = 40 percentage points.
- In the double-blind trial, 55 of 100 on active cream (55%) and 45 of 100 on placebo (45%) report improvement.
- Blinded difference: 55% − 45% = 10 percentage points.
- Reading: knowing the group inflated the apparent effect from 10 to 40 points on a subjective outcome such as pain.
Who is blinded and what it prevents
| Who | Bias it prevents |
|---|---|
| Patient | Unequal placebo effect, drop-outs and symptom reports shaped by expectations |
| Treating clinician | Different care or co-interventions by group |
| Outcome assessor | Observer bias when measuring or classifying |
| Data analyst | Analysis choices swayed by the result |
Markers
- Single-blind
- Double-blind
- Placebo
In practice
- In surgery or physiotherapy trials the therapist cannot be blinded, but the assessor who measures the result can.
- Reports should state exactly who was blinded instead of only saying ‘double-blind’.
Common pitfalls
- Confusing blinding with allocation concealment.
- A placebo with a different taste, colour or side effects can reveal the group and break blinding.
- Assuming an open-label study is invalid: with objective outcomes and a blinded assessor the risk of bias can be low.
Test yourself
Think of the answer, then tap the question to see it.
What does double-blind mean?
That neither the patients nor those who treat or assess them know which intervention each one receives.
Which kind of outcome benefits most from blinding?
Subjective ones, such as pain or perceived improvement; objective ones such as death are affected less.
In a surgery trial where the surgeon cannot be blinded, who can be?
The outcome assessor, the data analyst and sometimes the patient.
Connections
Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.
Sources: Chapter 8: Assessing risk of bias in a randomized trial (Cochrane, 2024); Observer bias | Catalog of Bias (Catalogue of Bias Collaboration, 2017).
Spotted a mistake?We are always improving for the community. Tell us what is wrong and our team will review it.Report a mistake
Representative model, for educational purposes only.