Severe imported malaria in children in France. A national retrospective study from 1996 to 2005

Backgrounds Malaria is a leading cause of imported febrile illnesses in pediatric travelers, but few studies have addressed severe imported pediatric malaria. We aimed to determine the risk factors and the features of imported pediatric severe malaria. Methods We conducted a retrospective, descriptive study using the French National Reference Center for Imported Malaria database, in children aged 0-15 years who were hospitalized with a falciparum malaria from January 1st 1996 to December 31th 2005. Uncomplicated and severe cases of falciparum malaria were compared to identify risk factors for severe cases. In the hospitals that reported more than five severe cases during the study period, we evaluated severe cases for prognostic factors and assessed the accuracy WHO criteria for predicting severity. Given the rarity of deaths, adverse outcomes were defined as requiring major therapeutic procedures (MTPs)-e.g., sedation, mechanical ventilation, nasal oxygen therapy, blood transfusions, hemodialysis, fluid resuscitation-or pediatric intensive care unit (PICU) admission. Results Of 4150 pediatric malaria cases included in the study, 3299 were uncomplicated and 851 (20.5%) were severe. Only one death was recorded during this period. Predictors for severe falciparum malaria were: age <2 years (OR = 3.2, 95% CI = 2.5-4.0, p < 0.0001) and a travel in the Sahelian region (OR = 1.7, 95% CI = 1.3-2.0, p = 0.0001). Of 422 severe malaria cases, a stay in a Sahelian region, lack of chemoprophylaxis, age < 2 years or thrombocytopenia < 100 x 10<^>3/mm<^>3 predicted adverse outcomes. Except for the hyperparasitemia threshold of 4%, the main WHO 2000 criteria for severe malaria reliably predicted adverse outcomes. In our study, the threshold of parasitemia most predictive of a poor outcome was 8%. Conclusion In imported pediatric malaria, children younger than 2 years deserve particular attention. The main WHO 2000 criteria for severity are accurate, except for the threshold of hyperparasitemia, which should be revised.