Introduction

Data on the epidemiology of aplastic anemia predominantly serve several primary purposes: to detect genetic predispositions, to discover external inducing factors, to quantitate the risks of exposure and to generate working hypotheses for the pathophysiology of the disease. There are several specific difficulties of ascertaining and interpreting the data, some of which are discussed in recent reviews (Gordon–Smith and Issaragrisil, 1992; Kaufman et al., 1991; Mary et al., 1996). The most relevant are:

1. The imprecision of diagnostic criteria. Stringent criteria, including adequate bone marrow cytology and histology, are used in collections of cases from single institutions or cooperative therapeutic trials. However, such series include selected cases referred to research institutions. In contrast, stringent criteria have not been applied to older epidemiological studies. As shown by the IAAS (The International Agranulocytosis and Aplastic Anemia Study) and the French registry on aplastic anemia, reviewing notified cases leads to the exclusion of other disorders that involve pancytopenia (Mary et al. 1990; The International Agranulocytosis and Aplastic Anemia Study, 1987), which were most probably included in less well-controlled studies.

2. Differentiating between chronic aplastic anemia and transient pancytopenia. The latter term refers to an entity not generally recognized in the classification of hematological disorders. Cases of pancytopenia with hypo- or normocellular bone marrow and spontaneous recovery within 10–90 days have been described under terms such as transitory bone marrow failure, transient aplastic anemia, bone marrow aplasia or bone marrow suppression, or reversible bone marrow aplasia (for literature see Keisu et al., 1990a). […]