In research, randomized controlled trials (RCTs) are seen as the best method for looking at interventions effect. In RCTs, the preference of patients (or their health care providers) for a treatment is not taken into account. This could lead to those patients refusing to be randomized and will never even enter the trial. Because these patients will not enter the trial, the generalization of RCTs could be a problem. Read more about the external validity of RCT.
Another problem that can occur in RCTs is low adherence for treatment in patients who do not receive the preferred treatment after randomization. On the other hand, patients who get the preferred treatment may adhere to this treatment better than average. In the last case, it could be that there is treatment effect which results from a patient’s preference and not from therapeutic efficacy.
One example of a study including the preferences of patients is on home-based or hospital based rehabilitation in cardiac patients. Next to the traditional RCT design used in this study, they also included patients and let them choose their preferred intervention. In this study, it was shown that the patients in who choose their intervention, had a higher adherence to their rehabilitation than patients who were randomized.
A method that takes the preferences of patients into account is preference randomized control trial and maybe more suitable to non-blinded controlled clinical trials. In this design, patients who have a strong preference will get the treatment they prefer and patients who do not have a strong preference will be randomized. In a trial with two interventions (A & B) you will end up with four groups:
Group 1. Randomized to A
Group 2. Prefer A
Group 3. Randomized to B
Group 4. Prefer B
What are your opinions on taking preference of patients into account when designing a study?
Leonie Klompstra is a Nurse Scientist at the Linköping University in Sweden. Her primary focus is on heart failure and rehabilitations.