Randomised Controlled Trials

A randomised controlled trial compares outcomes in participants assigned by chance to receive an investigational drug or a control intervention. Random assignment creates comparable groups, so the difference in their outcomes can estimate the drug's treatment effect rather than merely reflect pre-existing differences.

Creating a valid counterfactual

The control group represents what would probably have happened to the treated group without the investigational drug. It may receive a placebo or an established treatment, depending on scientific and ethical requirements.

• Random allocation tends to balance known and unknown prognostic factors between groups on average.
• Allocation concealment prevents recruiters from predicting assignments and consciously or unconsciously influencing enrolment.

Keeping treatment as the main difference

Apart from the assigned intervention, groups should be managed and assessed alike. Blinding of participants, care providers or outcome assessors, where feasible, reduces behavioural differences and biased outcome measurement.

• Pre-specified eligibility criteria, outcomes and follow-up procedures improve comparability and limit selective analysis.
• An intention-to-treat analysis compares participants according to their original assignment, helping preserve the benefit of randomisation.

Estimating and interpreting the effect

The difference in outcome rates or average outcomes between groups estimates the drug's effect, while confidence intervals indicate statistical uncertainty. Randomisation controls confounding on average, but it does not eliminate chance variation, attrition, non-adherence or measurement error.

• Rare-disease trials often have small eligible populations, making adequate sample size, multicentre recruitment and precise effect estimation difficult.
• A statistically detectable effect must still be assessed for clinical importance, safety and applicability to patients beyond the trial sample.