Evidence guide
What Is a Randomized Controlled Trial?
Random assignment makes comparison groups more alike at the start, helping researchers isolate the effect of an intervention.
Why randomize?
Without randomization, people who choose a treatment may differ from those who do not. Random allocation balances both measured and, on average, unmeasured characteristics.
What can still go wrong?
Small samples, missing outcomes, broken blinding, crossover, selective reporting and short follow-up can weaken an otherwise randomized study.
What it can answer
A well-run trial estimates the effect of offering or receiving the tested intervention in the studied population and setting. It does not automatically establish long-term safety or universal benefit.
A practical reading checklist
- Was allocation truly random and concealed?
- Were groups analyzed as assigned?
- How many participants were lost to follow-up?
- Was the primary outcome declared in advance?