Randomisation and blinding

Randomisation assigns participants to groups by chance, which distributes both known and unknown differences between them and is what allows a trial to support a claim about cause. Blinding conceals who received what, so that expectation does not shape how treatment is delivered or how outcomes are judged. Allocation concealment — keeping the upcoming assignment unpredictable during enrolment — is a third safeguard, and a distinct one.

These are the mechanics that separate a trial from a well-intentioned comparison. Our coverage looks at whether they were actually in place, because a trial described as randomised can still have been unblinded where blinding mattered most.