THE CENTER EFFECT IN MULTICENTRIC STUDIES - FIXED OR RANDOM

The multiplication of multicentric clinical studies leads to a general ization of the use of a center effect in statistical analyses. In the simple situation where the dependent variable is quantitative and ther e is only a simple factor of interest, the formalisation of such a pro tocol corresponds to a two way ANOVA (with the factor of interest and the center factor) with generally a term of interaction. Even such a s imple situation leads to interrogations, one of them being To determin e whether the centre effect is fired or random. In the first situation , the conclusion of the study will concern only the centres evaluated in the study, while in the second, the conclusion will concern the who le population from which the centers have been randomly selected. Such a generalizability has a cost : a loss of power in determining the ef fect of the factor of interest (the loss being as important as the int eraction factor of interest X centre factor is large) and the necessit y of sophisticated statistical software. This paper will focus on succ essively : the notion of ANOVA, of fixed or random effect (with the co nceptual, formal and numerical differences that distinguish these two approaches). A numerical example is proposed.