“As a clinician, I sometimes get tired of hearing that randomized controlled trials are the gold standard for examining outcomes. There are other types of research that are also important if we want to learn about the effects of therapies. Why should we pay so much more attention to randomized controlled trials?”
All research designs have value
There are many types of studies that can help us understand therapies and their effects — and they are all important. Qualitative research can give essential insight into patients’ experiences and generate hypotheses about outcomes, mechanisms, and priorities for research. Studies that follow patients over time without a control group can tell us how many patients improve and by how much. Single-case studies provide unique insights into the processes patients go through during therapy and can reveal potential mechanisms that other designs cannot capture.
These studies — and many others — are essential if we want to understand how therapies work, how well they work, for whom, and how we can implement them in routine care. There is no simple hierarchy that says one type of research is “the best”. Each design has its own strengths and limitations, and each is well-suited to specific kinds of research questions.
The one thing that sets RCTs apart
Randomized controlled trials (RCTs), however, have one characteristic that no other design shares: they can tell us whether a therapy caused an improvement — not just that improvement happened alongside therapy.
In an RCT, a group of people with a mental health problem is split into two subgroups at random. One subgroup receives the treatment; the other does not. Because the split is random, both groups end up — on average — identical in every other way: same distribution of age, same symptom severity, same proportion of men and women, same likelihood of having other health conditions. When the treatment ends and differences between the two groups emerge, those differences can logically only have been caused by the treatment.
This is the power of randomization: it balances all known and unknown characteristics between groups — without the researcher needing to know in advance what those characteristics are.
Why this matters for patients
In healthcare it matters enormously that we only offer treatments that have genuinely been shown to work. We do not want to give patients treatments that may be ineffective or even harmful — whether for physical or mental health conditions. And the most reliable way to demonstrate that a treatment works is through a randomized trial. No other study design can make the same claim with the same certainty.
This does not mean that RCTs are “more important” than other types of research — only that they are uniquely suited to the question of whether a therapy causes improvement.
Other designs come close — but not quite
Some study designs approximate the logic of RCTs. Observational studies using propensity score methods, for example, statistically control for a wide range of factors that could influence outcomes, bringing comparison groups closer together. Other designs take multiple measurements before and after an intervention, making it possible to detect whether a change occurred specifically during the treatment period.
But none of these approaches fully eliminates the possibility that some unmeasured factor — something the researcher did not think to control for — is responsible for the observed results. With true randomization, no such factor can systematically favor one group over the other.
RCTs have real limitations too
That said, RCTs are far from perfect. Many trials have some degree of risk of bias — methodological weaknesses that can undermine the confidence we place in their findings. They are also often conducted with carefully selected patient samples, which raises legitimate questions about how well the results translate to the full range of people seen in routine clinical care.
These are real limitations, and they are why the field works hard to improve how trials are designed and reported — and why systematic reviews and meta-analyses, which pool and critically evaluate evidence across many studies, are so valuable.
The bottom line
When we want to know whether a treatment works, there is no substitute for the randomized controlled trial. That does not diminish the value of other research designs — it simply recognizes that each design has its domain. For the specific question of whether a treatment causes improvement, the RCT stands alone.
Pim Cuijpers is professor emeritus of clinical psychology and scientific director of Metapsy. He has been involved in more than 1,100 scientific studies, mostly on psychological treatments of mental health problems. This is one of a series of evidence summaries in which Prof. Cuijpers tries to answer questions from patients and clinicians, based on what is known in science about treatments. The knowledge is mostly drawn from collective work of the Metapsy collaboration of at least 15 years. Do you have other questions you would like Prof. Cuijpers to discuss? Feel free to contact us.
Literature
- Harrer M, Cuijpers P, Schuurmans LKJ, Kaiser T, Buntrock C, van Straten A, Ebert D. (2023). Evaluation of randomized controlled trials: a primer and tutorial for mental health researchers. Trials, 24, 562. https://doi.org/10.1186/s13063-023-07596-3