Controversy over 'Contradiction': Should Randomized Trials Always Trump Observational Studies?

摘要

A RECENT ARTICLE REPORTED THAT OBSERVATIONAL studies of antioxidants or hormone replacement therapy in heart disease continue to be cited "despite strong contradictory evidence from randomized trials."1 The authors suggest this reflects strongly held irrational beliefs (so-called wish bias). Underlying their viewpoint is the assumption that randomized trials (RCTs) are more valid than observational studies, and therefore, when there is a contradiction between the results of the 2 study designs, the evidence from the RCT replaces, rather than complements, that from the observational studies. Opponents argue that the assumption of the primacy of the RCT leads to the loss of important evidence from observational studies