CLIO Learning Modules
Study DesignSelectionSample SizeData Collection and AnalysisHuman Subjects

'Me too studies'
Target Population
Attributable Risk
Relative Risk
Data Sources
Study Time
Case Control
Nested Case-Control
Prospective Cohort
Retrospective Cohort
Randomized Clinical Trial
Randomized Clinical Trial

A randomized clinical trial is an experiment that randomly assigns members of a defined population into two or more groups that receive different treatments or interventions. The outcomes of interest in the groups are then compared.


The basic design of a randomized clinical trial is outlined below. A defined population (e.g., people with a specific disease) are randomly assigned to two or more study groups or “arms” that are to receive different treatments or interventions. Often, one arm receives the standard of care or a placebo. Assigning individuals randomly to the arms optimizes the likelihood that the groups are similar and therefore comparable at the end of the study. After a follow-up period, the rates of outcomes among the assigned study arms are compared. If a specific outcome is more likely in one study arm than another, the intervention has had an effect.


Brown et al (Med Inform (Lond). 1997 Jan-Mar;22(1):77-89) conducted a randomized clinical trial to evaluate whether an educational role-playing video game can improve self-care behaviors (measured by questionnaires) in teenagers with diabetes. Through random assignment a total of 59 teenage diabetic patients received either the educational game (intervention group), the objective of which was to take care of several animated diabetic characters, or a video game with no diabetes-related content (control group). After six months of intervention, the improvement of self-care behaviors in the treatment was significantly greater than that in the control group (p=0.003), thus suggesting that diabetic teenagers may benefit from an educational role-playing game.


A more favorable design of randomized clinical trial, called double-blinded randomized clinical trail, makes sure that neither the study subjects nor the observers who record the outcomes know the groups to which the study subjects are assigned. This ensures that subjective judgment does not bias study outcome. Subjects who switch study arms during the course of the investigation should be considered as if they didn’t switch but remained in their initially assigned group (intention-to-treat analysis). That is because what one is really evaluating in an RCT is not the treatments themselves, but whether being assigned one treatment has an advantage over being assigned another treatment. This is the "real world" effect; patients change treatments all the time for reasons that can’t be understood by investigators.

The randomized clinical trial is considered the gold standard of studies in human subjects.

Further reading

Randomized clinical trials - a user’s guide (BMJ publishing group)

June 4, 2004 v0.20
Copyright © 2004 Stanford School of Medicine