Dr Kate Lee presents ‘Clustering in individually randomised trials: How much of an issue is it?’, on Friday 23 May from 11am to 12pm.
Even in trials which are individually randomised, there is often the potential for clustering of the outcomes, for example when the intervention is actively delivered by a health professional (e.g. surgeon, nurse, GP, therapist).
Each professional may be more or less effective, hence observations on patients treated by the same professional may be correlated or ‘clustered’.
Dr Lee’s presentation examines how often clustering arises in individually randomized trials and presents a number of methods to allow for clustering. She will provide random effects models used in the re-analysis of two major UK trials which highlight the importance of allowing for clustering to obtain accurate results and conclusions of a trial.
Dr Kate Lee is with the Clinical Epidemiology and Biostatistics Unit, Murdoch Children’s Research Institute, Melbourne.
This seminar takes place in the Menzies Seminar Room, John Mathews Building, Royal Darwin Hospital Campus.