Posters · Glossary
What is an RCT (randomised controlled trial)?
An RCT, or randomised controlled trial, is a study in which participants are allocated at random to receive an intervention or a comparison, such as usual care or a placebo, and then followed to compare outcomes. Randomisation makes the groups alike on average in both measured and unmeasured characteristics, so a difference in outcomes can be attributed to the intervention.
Randomised trials sit near the top of most evidence rankings for questions about whether something works, but only when they are run and reported well. A poster about one has to show readers how allocation, loss to follow up and analysis were handled, because that is where trials go wrong.
Nuwan Madhusanka · Co-founder
6 min read · Published
| Randomised controlled trial | Observational study | |
|---|---|---|
| Who decides exposure | Chance, through randomisation | Participants, clinicians or circumstances |
| Confounding | Balanced on average, including unmeasured factors | Can only be adjusted for factors that were measured |
| Typical cost and time | High; recruitment and follow up are expensive | Lower; often uses existing data |
| Best for | Whether an intervention works, under defined conditions | Rare harms, long term outcomes, exposures that cannot be assigned |
| Main risks | Poor allocation concealment, loss to follow up, selective reporting | Confounding, selection bias, measurement bias |
| Reporting guideline | CONSORT | STROBE |
Why randomisation matters
In an observational study, people who choose a treatment differ from those who do not in ways that also affect outcomes, such as health, income or motivation. Adjustment can account for differences that were measured, but not for those that were not. Random allocation removes that problem on average, because chance, not anyone's judgement, decides who gets what. Two conditions keep it working. Allocation must be concealed, so recruiters cannot predict or influence the next assignment, and analysis should keep people in the groups they were randomised to, known as intention to treat, even if some did not follow the allocation. A table of baseline characteristics by arm lets readers see that randomisation produced similar groups; on a poster, three or four key characteristics are usually enough.
Types of trial
A parallel group trial randomises individuals to two or more arms at the same time and is the most common design. A cluster randomised trial randomises groups, such as clinics, schools or wards, when the intervention works at that level or contamination between individuals is likely; it needs more participants because people within a cluster resemble each other. A crossover trial gives each participant both treatments in random order, which suits stable chronic conditions. Superiority trials ask whether one option is better, while non inferiority trials ask whether a cheaper or simpler option is not worse by more than a set margin.
Registration and reporting
Trials should be registered before the first participant enrols, recording the primary outcome and planned analysis, so later readers can check that the reported outcome was the one intended. The World Health Organization's International Clinical Trials Registry Platform links national registries into a single searchable portal. Reporting follows the CONSORT statement, which provides a checklist and the participant flow diagram. Assessment of bias commonly uses the Cochrane risk of bias tool, known as RoB 2, whose domains cover the randomisation process, deviations from intended interventions, missing outcome data, measurement of the outcome and selection of the reported result.
What a trial poster must show
The essentials fit on one board: the design and setting, the eligibility criteria, the intervention and comparator described well enough to repeat, the primary outcome and when it was measured, the sample size calculation, how randomisation and concealment were done, whether anyone was blinded, and the participant flow from assessed to analysed. Results should give each group's outcome, the difference with its confidence interval, and harms. The registration number belongs in the abstract or methods. A separate page on this site covers how to build the participant flow diagram so its counts reconcile. Harms deserve their own line even when none occurred, because a board that reports only benefits leaves readers unable to weigh the trade off.
Common mistakes
Describing a study as randomised when allocation was by alternate days or record numbers is misleading; those methods are predictable. Other frequent problems are reporting per protocol results as the main analysis, switching the primary outcome after seeing the data, omitting the flow of participants, and reporting only relative effects. Cluster trials analysed as if individuals had been randomised overstate precision. Trials stopped early for benefit tend to exaggerate the effect, which should be stated if it happened. Finally, a poster that reports many secondary outcomes in the same type size as the primary outcome makes it hard to see what the trial was designed to answer.
Where it shows up in the poster builder
A trial poster draws on the blocks built for it. The participant flow block, on ten of the forty layouts, holds six nodes on the main path and one aside for exclusions, which fits a simplified CONSORT path. The protocol block, on 38 layouts, takes up to five labelled entries such as design, participants, intervention, comparator and outcome. Key number blocks, on 27 layouts, carry each arm's result and the difference with its interval. The clinical trial example reports a cluster randomised trial of 20 clinics with a 21.2 point difference and its interval of 15.9 to 26.5.
Questions people ask
What is the difference between randomisation and allocation concealment?
Randomisation generates an unpredictable sequence of assignments. Allocation concealment keeps that sequence hidden from the people enrolling participants until each person is irreversibly enrolled. Without concealment, a recruiter who can see the next assignment might steer sicker or healthier patients into a particular arm, undoing the benefit of randomisation even though the sequence itself was random.
Does every RCT need a placebo?
No. A placebo suits drug trials where a dummy treatment is possible and withholding active treatment is ethical. Many trials compare a new intervention with usual care or an existing treatment, which answers the practical question of whether it is better than what people already receive. Behavioural and service interventions often cannot use a true placebo at all.
What does blinding mean in a trial?
Blinding keeps people from knowing which group a participant is in. Participants, clinicians, outcome assessors and analysts can each be blinded. Even when participants cannot be blinded, such as in an exercise trial, blinding the outcome assessor reduces measurement bias. Report who was blinded rather than using terms like double blind, which mean different things to different readers.
Can an RCT be small?
It can, but a small trial has low power and imprecise estimates, and a chance imbalance between groups is more likely. Small trials are appropriate as pilot or feasibility studies that test recruitment and procedures. They should be labelled as such and not interpreted as definitive tests of effectiveness, even if a result reaches significance.
What is intention to treat analysis?
It analyses participants in the groups to which they were randomised, regardless of whether they received or completed the intervention. It preserves the balance created by randomisation and reflects what happens when a treatment is offered in practice. A per protocol analysis, restricted to those who followed the allocation, can be reported alongside it but is more open to bias.
Where do I find a trial's registration number?
In the registry where the trial was recorded, such as a national registry linked to the WHO International Clinical Trials Registry Platform. Registry numbers carry a prefix identifying the registry. Put the number in the abstract or methods of the poster, and a QR code to the registry entry lets readers check the planned outcomes against what is reported.
Make one with posters
The button opens the generator with this use case already described. Change the wording to match your own.
Create a poster with OneCraftRelated questions
- What is a CONSORT flow diagram?A CONSORT flow diagram tracks every participant from screening to analysis. What each box holds, why the arithmetic must reconcile, and how it fits a poster.
- Cohort vs case-control studyCohort vs case control study: a cohort follows people from exposure to outcome, a case control study starts at the outcome and looks back. What each estimates.
- What is statistical power?What is statistical power: the chance a study detects a true effect of a given size. Why 80 percent is the convention, and how power changes with sample size.
- What is the protocol block on a research poster?What is a protocol on a research poster: methods set out as short labelled entries instead of paragraphs. What each entry holds, and when prose works better.
Step by step in the builder: Choose the template, theme and size.
Written and checked by the OneCraft team. Last checked .