The multiple propensity score for analysis of dose-response relationships in drug safety studies

In order to detect adverse drug reactions, large observational drug safety studies are necessary as randomized clinical trials rarely have enough powe r. However, in order to obtain reliable results the issue of confounding, e specially confounding by indication, should be addressed. We proposed a mul tiple propensity score, which is an extension of the propensity score, to r educe the bias in a dose-response analysis in a drug safety study. The mult iple propensity score has: similar properties to the propensity score in Ro senbaum and Rubin.(1) Using the propensity score for bias reduction of the risk ratio was considered. We used the multiple propensity score in a study of the dose-response relationship between diclofenac prescriptions and hos pitalization for gastrointestinal bleeding and perforation using a record l inkage database. We found that the unadjusted risk ratios were biased downw ards due to confounders and that this bias was reduced by using stratificat ion based on the multiple propensity score. Copyright (C) 2001 John Wiley & Sons, Ltd.