Centre-by-treatment interaction: What is the issue?
Anders Källen
Astra Draco, Lund
Abstract
When analyzing multicenter studies, one of the issues is how to handle
the treatment-by-center interaction. This is an issue regardless of
whether the interaction is statistically significant or not. If it is
not statistically significant, should we analyze without the interaction
term in the model or should it be included? If it is statistically
significant, what, if any, conclusions can we draw from the study?
It will be argued from a clinical perspective that the
center-by-treatment
interaction is a non-issue and that the statistical analysis should
reflect this.
A mode of analysis of multicenter studies is proposed that focuses on
the units studied, patients, and still acknowledges the statistician's
which to adjust for
centers and their interaction with treatment. The analysis is not
conditioned on the outcome of the interaction test.