Conclusiveness of rechallenge in the interpretation of adverse drug reactions

We here consider the extent to which the presumed correlation between an adverse event and the administration of a particular drug can be reinforced by rechallenge. At first question of terminology is: what is a rechallenge? Rechallenge is often accepted too readily as proof of a causal relationship and clinical examples give illustrations of common misinterpretations. Definitions are proposed to characterize: the outcome of rechallenge; the conditions under which rechallenge is performed. In discussing causality, a sharp distinction is drawn between the outcome per se and the establishment of a causal relationship. Finally, the simple concepts proposed here should permit to establish a typology of rechallenge and to assess, by further experimental or retrospective research, the conclusiveness of rechallenge in interpreting adverse drug reactions.