The randomized controlled trial (RCT) is the gold standard of clinical research, providing the highest level of scientific data for evidence-based medicine, and is often referred to as “Level I” in literature reviews and meta-analyses. RCTs are usually the most expensive studies, often in the millions of dollars, so it’s critical to plan your strategy wisely. In this blog post, we’ll help you determine why and when an RCT may be the best approach to collect your clinical data.
What is an RCT?
To begin, let’s ensure we’re all on the same page about what we mean by RCT:
- Randomized: Randomization is the process of assigning patients to treatment groups using an established recognized statistical method that employs an element of chance to determine the unforeseeable assignment in order to reduce bias.
- Controlled: The study is performed under an established protocol by which two (or more) groups of people in the trial receive different treatments, one of which is considered the control group. The control group may be subjected to standard practice (e.g., medical management, an existing surgical procedure, or a comparable marketed device), a sham treatment procedure, or no intervention at all.
- Trial (aka clinical investigation or study): Systematic evaluation in one or more human subjects to assess the clinical performance, effectiveness, or safety of an intervention.
Why go through the rigor and expense of conducting an RCT?
The most important benefit of an RCT is the ability to assess treatment outcomes versus control in a statistically unbiased fashion. An RCT also offers the opportunity for study blinding, including of the patient, the endpoint assessors, or sometimes even the treating physicians and site staff. Blinding can be just as important as randomization in interpreting study outcomes, since unblinded studies introduce evaluation bias. Effects on judgment and interpretation can creep in when study personnel know how patients have been treated.
Sometimes a single-arm, treatment-only study may be a plausible alternative to an RCT. How do you know which one is best for your situation? Perhaps the overview of pros and cons provided below will help with your strategy development.
When will an RCT be necessary?
The following situations make an RCT likely, if not certain:
- When seeking to show benefit (e.g., less invasive, simpler, safer, more effective) over an established standard of care. Examples include device therapy compared to medical management as a control, or sometimes even “watchful waiting” for conditions where medical management is not relevant or has proven fruitless. This sort of RCT is the classic superiority comparison that attempts to prove the novel therapy offers significant advantages in terms of study outcomes over the control group.
- When comparing a novel treatment to an established active therapy, which in device studies usually means the control is a device as well. In this case, the established therapy forms the control group, and the objective of the trial may be noninferiority, meaning that the novel treatment need only be as good as the established, active control.
In these situations, the need for a control group is obvious; a trial cannot compare treatment to control without a control group. It is naturally possible to conduct a trial with a concurrent control group but without randomization (e.g., when it is impossible to assign treatments to individuals, which sometimes happens in the social sciences when behavior is being observed rather than treatments assigned).
In a medical device framework, though, it rarely makes sense to have a concurrent control group enrolled without randomization, since randomization is fundamental to balancing the groups and, therefore, permitting a clean evaluation of study outcomes.
When dealing with FDA, there is no explicit statutory requirement for RCTs. For instance, FDA’s mandate on premarket approval (PMA) submissions states that the agency is to rely upon “valid scientific evidence,” which includes the following, per FDA guidance:
- Well-controlled studies
- Partially controlled studies
- Objective trials without matched controls
- Well-documented case histories
- Reports of significant human experience
Over time, FDA has agreed with study sponsors on the use of a vast number of nonrandomized studies, even those intended to support PMA submissions, and this will continue. Sometimes, it’s simply impractical for operational or even ethical reasons to conduct an RCT. Unsurprisingly, though, the FDA’s Center for Devices and Radiological Health (CDRH) will prefer a randomized controlled trial over other options if the opportunity exists. FDA’s list of evidence sources is presented in the order that it is (top to bottom, strongest to weakest) for a reason.
Conclusion
The RCT is the only kind of trial that provides a concurrent group of patients against which the novel treatment can be compared, while also minimizing the possibility of systematic differences between patients in the treatment and control groups. Rather than just evaluating treatment outcomes, the use of a concurrent control group gives a fully valid estimate of treatment benefit over the control. And thanks to randomization, the study groups are similar in terms of subject demographics and medical conditions, so the study results can be compared across the groups. Data from RCTs can be used to demonstrate product performance, effectiveness, and safety, all of which can be compelling to the market decisions made by regulators and clinical experts.
