Premium
Allocation concealment and blinding: when ignorance is bliss
Author(s) -
Forder Peta M,
Gebski Val J,
Keech Anthony C
Publication year - 2005
Publication title -
medical journal of australia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.904
H-Index - 131
eISSN - 1326-5377
pISSN - 0025-729X
DOI - 10.5694/j.1326-5377.2005.tb06584.x
Subject(s) - blinding , statistician , bliss , medicine , clinical trial , ignorance , library science , family medicine , political science , computer science , law , pathology , programming language
NHMRC Clinical Trials Centre, University of Sydney, Sydney, NSW. Peta M Forder, BSc, MPH, Statistician; Val J Gebski, BA, MStat, Associate Professor, and Senior Research Fellow; Anthony C Keech, FRACP, MScEpid, Deputy Director. Reprints will not be available from the authors. Correspondence: Associate Professor Anthony C Keech, NHMRC Clinical Trials Centre, University of Sydney, Locked Bag 77, Camperdown, Sydney, NSW 1450. enquiry@ctc.usyd.edu.au The Medical Journal of Australia ISSN: 0025729X 17 January 2005 182 2 87-89 ©The Medical Journal of Australia 2005 www.mja.com.au EBM: Trials on Trial known before the patient is entered into the study. Bl to the masking of the treatments after randomisation patient, the investigator or the outcomes assessor. W tion, allocation concealment is achievable in all clinical trials. In contrast, it is not always possible to to study treatments received. The CONSORT statem encourages detailed reporting of the allocation conc oo of to G d study design involves minimising all possible sources bias. Two important sources of bias arise through failure mask (ie, conceal), first, the randomisation process and, second, the treatments after randomisation. Allocation concealment is the term used to describe the procedure for protecting the randomisation process so that the treatment to be allocated is not inding relates — from the ithout exceprandomised blind people ent strongly ealment process and the measures taken to preserve blinding (Box 1).