Fool's gold: Examining the use of gold standards in the production of research evidence

Research designs, and the methods that they dictate, normally arise from the nature of the questions being asked. However, in most medical research, a hierarchy of evidence is in common use that influences choice in favour of research designs that include methods with a higher ranking. The acceptance of superior research methods consequently limits the types of question that can be asked. The randomised controlled trial (RCT) has enjoyed a long reign as the gold standard in health research methodology. The RCT is used here as a case example of how hierarchies of research methods can distort the production of evidence relating to health care. The acceptance of the RCT as a superior method in the hierarchy of evidence production has been promoted by claims that RCTs offer reliability, objectivity and levels of certainty about treatment effectiveness (inferred through statistical significances) that other methods do not. These statistical significances depend upon mathematical probabilities that imply a level of uncertainty. Mathematical significance has been confused with the substantive or practical importance of research results. The importance of research findings cannot be determined by hierarchies, nor by statistical techniques alone; other types of method can provide findings that are equally valid, true or useful to evidence production in occupational therapy.