The one-person randomized controlled trial

Qual Manag Health Care. 2005 Oct-Dec;14(4):206-16. doi: 10.1097/00019514-200510000-00002.


Currently, the gold standard for collection of clinical evidence is the randomized controlled trial (RCT), preferably with large, multicenter samples of subjects. Although this approach provides valuable information, many clinicians find it difficult to translate RCT results to the individual patient level. In this report, a statistical approach called Design of Experiments (DOE) is described as a method of applying the principles of RCTs one person at a time. An overview of the method, with a simple clinical example, is presented. As shown, DOE is a more efficient method than the sequential approach often taken by clinicians and their patients when evaluating various treatment choices. Further, the effect of multiple interventions can be assessed, alone or in combination with each other. In this way, DOE can be an important addition to the field of evidence-based medicine, although further studies are needed.

MeSH terms

  • Data Interpretation, Statistical*
  • Evidence-Based Medicine
  • Humans
  • Randomized Controlled Trials as Topic / statistics & numerical data*
  • Sleep Wake Disorders
  • United States