In our Research 101 blog series, we have discussed how clinical trials are designed and the strengths and weaknesses of these designs. We learned that Randomized Clinical Trials (RCTs) provide the most compelling evidence that a treatment causes an observed outcome, so let’s talk a bit about what randomization is and why it is so important in clinical trials.
When talking about clinical trials, randomization means that chance alone (like a flip of a coin) determines what group a person is put in to. By leaving it up to chance, we help ensure that the groups start off on equal footing. How does this work? Let’s work through a study and find out.
What do we want to study?
We want to know if eating apples keeps us healthy.
How do we want to find out?
Group A will be required to eat two apples a day. Group B will be banned from eating any apples. The study will last one year because we think that one year is long enough to see results. We know that people with lower heart rates tend to be healthier, so both groups will have their heart rates checked at the end of the study.
How are we going to pick people for Group A and Group B?
We need a large number of people for the study, and we need them to stick with it for an entire year.
One way to decide is to say – anyone who regularly eats apples should be in Group A and anyone who never eats apples is in Group B. This is a natural division in the population. The problem is people who already eat apples are likely to be very healthy at the start of the study. Therefore, the groups would not start off on an equal footing.
To ensure that Group A and Group B are equivalent at the beginning of the study, we can flip a coin for each person. That way each person has a 50% chance of being in either group – regardless of initial eating habits. This will then help us be sure that our results were a product of our treatment.
In this case randomization helped divide healthy and unhealthy people equally into Group A and Group B. Since Group A and Group B started off on equal footing, any difference in heart rate between the groups will be a result of eating apples. In other words the treatment (eating apples) caused the outcome (lower heart rates).
This is a simple example of how randomization helps equalize factors that may muddy, or confound, the results. In real-life studies, these confounding factors may not be obvious, making it difficult to determine if the treatment truly caused the observed outcome.