Assessing methodological quality of published papersBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7125.151a (Published 10 January 1998) Cite this as: BMJ 1998;316:151
Pre-allocation bias in randomised controlled trials must be taken into account
- Michael Gossop, Head of researcha,
- John Marsden, Research coordinatora
- a National Addiction Centre, Maudsley Hospital, London SE5 8AF
- b Department of Paediatrics, Northampton General Hospital, Northampton NN1 5BD
Editor—In her article on assessing the methodological quality of published papers, Greenhalgh does not draw sufficient attention to one of the main sources of bias found in randomised controlled trials.1 This can occur before the allocation procedure itself and is, therefore, easy to miss.
Recruitment to the trial usually requires that subjects accept the offer of either of two forms of treatment. In many cases, one of the treatments may be so strongly preferred to the other on a …