Evaluating treatment effects reliably

BMJ 2002; 325 doi: https://doi.org/10.1136/bmj.325.7377.1372 (Published 14 December 2002) Cite this as: BMJ 2002;325:1372

Although principles are well known, they are ignored too often

  1. Martin J Landray, clinical research fellow (martin.landray@ctsu.ox.ac.uk),
  2. Gary Whitlock, research fellow
  1. Clinical Trial Service Unit and Epidemiological Studies Unit, University of Oxford, Radcliffe Infirmary, Oxford OX2 6HE

The introduction, more than half a century ago, of properly randomised trials in which the treatment allocation was rigorously concealed was a watershed in the evaluation of treatment effects.1 But this major advance did not come out of the blue: as was highlighted at a recent conference in Oxford (Beating biases in therapeutic research: historical perspectives, www.wuhmo.ox.ac.uk/docs/BeatingBiases.html), attempts to combat bias in therapeutic evaluation had in fact been made during the preceding few centuries.24

In the 18th century, the traditional practice of claiming therapeutic achievement on the basis of pathophysiological theories and anecdotal “successes” started to be challenged by medical non-conformists who wrote careful, prospective, analytical accounts of medical treatments, some of which included comparison with a control group.4 These “arithmetic observationists and experimentalists”4 recognised the need …

View Full Text

Log in

Log in through your institution


* For online subscription