Thrombocytopenia in children as in adults has a pleiomorphic expression, which may vary from a life-threatening hemorrhagic accident, to a biological finding in an apparently healthy child. In many patients clinical expression is absent or mild and thrombocytopenia will be fortuitously detected by blood cell counts during the course of a disease or prior to surgery. Neonatal thrombocytopenias also have special features requiring special diagnostic approach and management.

Platelet counting by blood cell automatic counters usually is reliable. However, in some situations, platelet counting can be erroneous; for example, circulating particles whose volume is within the platelet volume range can spuriously increase the platelet count and mask a thrombocytopenia; this can occur with extreme microcytosis or red cell fragmentation. On the contrary, in vitro platelet consumption, due to hemostasis activation during sampling or inefficient anticoagulation of the sample or to EDTA-induced platelet aggregation, can lead to spurious thrombocytopenia. In any case, an initial diagnosis of thrombocytopenia must be verified by phase contrast microscopy after microsampling, and by examination of the platelets' size, morphology and density on blood smears. There is no close correlation between clinical presentation and platelet count. Most patients will have minimal bleeding symptoms even with platelet counts below 50 × 109/l. Others can exhibit bruising and bleeding in spite of moderate thrombocytopenia due to associated platelet functional abnormalities. The search for a precise etiology, which can be either constitutional or acquired, needs a rigorous approach, as therapeutic decisions have to rely on certainties (Table 37.1).

In this chapter drug-induced thrombocytopenia and thrombocytopenia associated with malignancies will not be considered.