Dear Allstat, I am working on a multi-centre clinical trial which uses minimisation on various factors to randomise patients. One of the factors that is minimised on is centre, and we realise that this is likely to increase predictability of treatment within centre, as when centre is included in minimisation algorithms, the treatment allocation just tends to alternate. The reason for including centre in the minimisation algorithm in this trial is that the primary outcome is measured by Doppler, which will be of variable quality between centres, so we want treatment to be balanced within centre. I'm looking at ways to improve on this. We want to retain the minimisation method as the trial is underway, but we want some way to tweak it, so that it remains as balanced as possible, but is less predictable. I have found an abstract in Controlled Clinical Trials by Gerrit Waverling and Martin Prins, looking at 'fixed P minimisation' and 'shifting P minimisation'. However, it is just an abstract at a meeting and it doesn't give many details. Does anyone out there have any papers, experience or advice they can give me on this issue? I have to decide what to do about this within a couple of weeks, so I will provide a summary once I've decided what to do. Thanks, Steff Lewis. --------------------------------------------------- Stephanie C. Lewis, PhD MSc CStat Medical Statistician Bramwell Dott Building Department of Clinical Neurosciences Western General Hospital Crewe Road EDINBURGH Tel: +44 (0) 131 537 2932 EH4 2XU Fax: +44 (0) 131 332 5150 UK Email: [log in to unmask]