A new drug medical treatment or test comes on the market. It sounds promising. But is it safe? Is it effective? Is there a way to know?

In the words of W. Edwards Deming, “In God we trust, all others [must] bring data.”

One way to “bring data” in testing the effectiveness of a new medical intervention is a randomized controlled trial (RCT). An RCT study evaluates the effectiveness of a medication or medical treatment by encouraging a bias-free evaluation.

If you conduct an Internet news search for “randomized controlled trial,” you’ll find tens of thousands of trial results being shared on everything from vaccinations for shingles to whether eating protein will benefit elderly men. These trials often receive media attention because they are considered the “gold standard” of clinical trials.

In fact, according to BMJ, RCTs are “most rigorous way of determining whether a cause-effect relation exists between treatment and outcome and for assessing the cost effectiveness of a treatment.”

What makes these trials so highly regarded? The answer lies in the design of the trials themselves.

Components of a Randomized Controlled Trial

From setup to completion, a high-quality RCT is dependent on its design to ensure quality and impartiality. These are some of the main components of an RCT:

• Endpoints: Primary and secondary endpoints are determined before the study starts.

• • The primary endpoint is the main result measured at the end of a study to determine if a treatment worked.

Secondary endpoints evaluate other effects of the intervention.

• Study Population: The RCT needs to define the exact criteria for who is eligible to participate in the trial and should reflect those who suffer from the condition or disease. Additionally, a large sample size from diverse areas can add to the strength of the study.

• Randomization: Once a representative sample of patients are selected, they are randomly assigned into different study groups (experimental and control) to ensure the groups are comparable.
  • The experimental group will receive the new treatment or medication that is being tested; the control group will receive treatment as usual. This enables researchers to evaluate which group demonstrated the most improvement in symptoms and/or disease markers.
  • According to the National Institute of Health’s Office of Behavioral & Social Science Research, randomization is important to achieve balance: “patients in different groups are similar for characteristics such as age, gender, social class, time of year of presentation, country of presentation, and type of hospital.”
  • For example, randomization means that both the experimental and control group have equal representation in men, or people aged 20-45, or from the US.
• Blinding: Blinding means hiding the treatment from either patient or the physician/test administrator. So, for example, in a medication RCT, the patient may be given a placebo and neither the patient or the physician would know that it is a placebo.
  • Double-blind study: neither the patient nor physician know which treatment is being given.
  • Double-blind pharmacogenomics study: Since the physician uses the genetic test to guide medication selection, the patient and the “central raters” are blinded – meaning it’s not clear who had taken the genetic test and who had not. The central raters administer the depression rating scale to assess the patient’s depression severity.

All these steps work to ensure the most unbiased test results. By eliminating the placebo effect (in which a patient has a response to a placebo) and isolating the impact of the intervention through blinding, RCTs can lead to great advancements in the medical arena.