How can good randomized controlled trials in leading journals be so misinterpreted?

J Vasc Surg. 2013 Feb;57(2 Suppl):3S-7S. doi: 10.1016/j.jvs.2012.04.076.


The results of good randomized controlled trials (RCTs) published in leading peer-reviewed journals have been deemed the best possible basis for good medical practice. However, several limitations may decrease their value. These include flaws and weaknesses in the design and the timeliness of RCTs. Progress in a treatment method or control arm may invalidate a trial. So too can defects in patient selection, physician competence, randomization, applicability, end points, and the population being studied. Idiosyncratic flaws can also invalidate an RCT. Examples of these flaws and weaknesses are presented. Another problem with articles describing RCTs is the potential for the conclusions of the trial report to be misleading because of error or bias. This plus subsequent misinterpretation of the trial results or conclusions by others can make the effect of the trial misleading with an unintended detrimental result on medical practice. Guidelines based on such errors or bias-based conclusions and misinterpretations can further compound the problem. This article provides examples of misleading conclusions and/or misinterpretations (spinning) of trial results in articles describing RCTs in leading journals. All physicians should recognize these value-limiting processes so that RCTs can be evaluated adequately and fairly. In that way, they can be used along with good physician judgment to optimize the care delivered to individual patients and to society at large.

MeSH terms

  • Bias
  • Clinical Competence
  • Comprehension*
  • Data Interpretation, Statistical
  • Endpoint Determination
  • Evidence-Based Medicine* / standards
  • Evidence-Based Medicine* / statistics & numerical data
  • Humans
  • Patient Selection
  • Peer Review, Research*
  • Periodicals as Topic*
  • Practice Guidelines as Topic
  • Randomized Controlled Trials as Topic* / standards
  • Randomized Controlled Trials as Topic* / statistics & numerical data
  • Reproducibility of Results
  • Research Design
  • Treatment Outcome