Twelve patients with severe malaria treated with partial exchange transfusion. Comparison between mathematically predicted and observed effect on parasitaemia.

Imported malaria is increasing in Western countries, which results in considerable morbidity and mortality, the latter mainly due to delayed diagnosis and treatment. Partial exchange transfusion has been proposed as a therapy for very severe falciparum malaria, but the utility of this procedure has not been proven. We report on 12 patients with severe Plasmodium falciparum malaria, treated with exchange transfusion in 5 Belgian hospitals between 1987 and 1991. The mean parasitaemia before exchange was 17.2% (range 6-35%). An average of 3.1 l was exchanged within 3 to 7 hours. Ten of the 12 patients survived. One patient recovered initially, but died as a consequence of a cascade of complications of dialysis. Three patients developed the adult respiratory distress syndrome (ARDS), two had less serious pulmonary involvement, and five had temporary renal failure; none of the survivors had sequelae. A formula is proposed to calculate the expected reduction in parasitized erythrocytes by exchange transfusion in function of the initial parasitaemia, the initial haemoglobin level and the volume of blood exchanged. Comparison between the mathematically predicted and the observed decline in parasitaemia shows on average a 25% excess of observed over predicted efficacy per unit of blood exchanged. After introducing this correction the formula enables the clinician to estimate roughly the volume of blood that has to be exchanged in order to bring the initial parasitaemia down to a desired level.