ON MIETTINENS MULTIVARIATE CONFOUNDER SCORE

Consider a regression model for the effect of a treatment on an outcom e variable in the presence of potential confounders. It is common to t est for main effects and interactions of the treatment variable; if th e confounding variables are discrete, one might then compare the treat ment groups within strata formed by suitable covariate patterns. An al ternative, due to Miettinen, is to stratify according to a ''multivari ate confounder score'' and test for treatment effects within strata. T his test has been shown to be flawed, and the method appears largely t o have fallen into disuse. Here we show that such stratification never theless provides a sound and often useful graphical comparison of trea tment and control groups. (C) 1998 Elsevier Science Inc.