‘Randomisation is to minimise bias and ensure that the patients in each treatment group are as similar as possible in all known and unknown factors. This will ensure that any differences found between the groups in the outcome(s) of interest are due to differences in treatment effect and not differences between the patients receiving each of the treatments.

It removes the chance that a clinician will consciously or unconsciously allocate one treatment to a particular type of patient and the other treatment to another type, or that a certain kind of patient will choose one treatment whilst another kind will choose the other.’

Harrison J. Presentation to Consumers’ Advisory Group for Clinical Trials, 1995.

From: Chapter 3, More is not necessarily better

Cover of Testing Treatments
Testing Treatments: Better Research for Better Healthcare. 2nd edition.
Evans I, Thornton H, Chalmers I, et al.
London: Pinter & Martin; 2011.
Copyright © 2011 Imogen Evans, Hazel Thornton, Iain Chalmers and Paul Glasziou.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.