Some potential sources of bias in randomised controlled trials

P. Cuijpers and I. A. Cristea. 2015. How to prove that your therapy is effective, even when it is not: a guideline. Epidemiology and Psychiatric Sciences, Epub.

This is an interesting article about bias within randomised controlled trials (RCTs) of the efficacy of psychological therapies. The article is particularly relevant because one of the authors (Cuijpers) is well established in conducting meta-analysis of RCTs in the field of psychological therapies.

The authors identify a number of factors that might serve to increase the likelihood of positive research outcomes but that do not necessarily mean that the therapy intervention is efficacious. These factors include: the researcher holding a belief that the therapy works; the participants holding the belief that the therapy works; poor methods of randomization; poorly blinded studies; exclude participants who dropped out of the study from the final analysis; cherry pick outcome measures; small sample size and the use of a wait list control group.

This is of particular interest to me because I am thinking a lot about subjectivity/objectivity within research design at the moment and considering the case for the existence of the researcher within research methods such as RCTs. I suspect that over time it will become a ‘go to’ paper for me when considering the strengths and limitations of clinical trials.


No comments.

Leave a Reply

Your email address will not be published. Required fields are marked *