Abstract
Blood dyscrasias account for only a minor fraction of all adverse drug reactions (ADRs), but are relevant because of their relatively high morbidity and mortality. For the majority of drugs, the magnitude of risk is low enough to remain undetected until wider distribution of the drug takes place. Thus, only post-marketing studies, carried out with appropriate methodology and sufficient statistical power, will allow the risk of serious haematological side-effects of new drugs to be ascertained. Publication of carefully studied and thoroughly described single case studies and reports to registries are necessary to detect new associations between drugs and blood dyscrasias, while only large cohort or case-control studies are suited to quantify the risks. Physicians managing a newly detected blood dyscrasia should be aware that it may be drug-induced. They should assess the exact diagnosis, obtain and thoroughly document a detailed exposure history and follow the blood counts after withdrawal of all potentially relevant agents. The recognition and appropriate management of the problem in individual cases is the basis for both effective patient care and the quality of subsequent pharmaco-epidemiological evaluation.