| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dear Sir:
In our series of children living in a low, seasonal transmission setting for malaria (less than one infective bite per person per year), thrombocytopenia less than 100,000/mm3 was strongly associated with both severity and mortality.1 We have also proposed that thrombocytopenia can be a prognostic factor and independent predictor for death in P. falciparum malaria in African children living in areas of low transmission.1 However, in a large retrospective study from Kenya, Ladhani and others found no such associations and platelet counts were actually lower in mild cases than in severe ones.2
Given the extremely significant prognostic value of thrombocytopenia in our study, we doubt the differences are due to population size. Similarly, different care seems unlikely given that the superiority of loading doses of quinine on survival has not been proven,3 and that all severe cases in our series were treated in emergency wards, where mechanical ventilation and other intensive care supplies were available.1 We think that other reasons could explain the differences observed between our respective data.
First, in the series of Ladhani and others, the high upper limit of the normal platelet count reference range, i.e., 700,000/mm3, suggests confounding factors, e.g., iron deficiency (ruled out in our study) might have affected platelet counts and introduced a bias.
Second, we agree with Ladhani and Newton that geographic and patient population differences could explain the differences between the two studies. In Kilifi, Kenya, malaria transmission is higher and the patients are younger. This results in severe anemia predominating among severe forms of malaria and a much lower ratio of cerebral malaria to severe anemia than in Dakar, Senegal.4 In our series in Dakar, platelet counts were higher in those with severe anemia than in those with other manifestations of severe malaria, and the association between thrombocytopenia and death was not significant among severe cases of anemia (odds ratio [OR] = 3.55, 95% confidence interval [CI] = .3–46.4, P = 0.197; Table 3 in our paper), while it was highly significant in the other main severe forms of malaria, i.e., cerebral malaria and respiratory distress.1
Third, severe anemia has a better prognosis in hospitals where blood transfusions are available,5,6 which could explain the 2% death rate observed among their severe cases, a rate much lower than the 10–30% usually reported.7 However, we agree with Ladhani and Newton that fatality rate in their study was likely to be too low to detect significant differences in platelet counts between children who died and those who recovered.
Fourth, in Dakar, people are weakly immune against malaria, and severe forms occur irrespective of age.5 We have reported that cerebral malaria is more frequent among older children and has a greater inpatient mortality risk than severe anemia.5 In the present series, an admission thrombocytopenia less than 100,000/mm3 was associated with increased mortality, particularly in those with cerebral malaria (OR = 9.4, 95% CI = 2.3–54.4).1 We monitored platelet counts from the first to the seventh days in 208 children with severe malaria, and a retrospective analysis showed a significant association between improvement in the level of consciousness and an increase in the platelet count (P < 0.01, by chi-square test and P < 0.005, by Kruskal-Wallis test) (Table 1
). Among survivors, an increase in platelet counts between admission and the first control was associated with a +0.64 mean improvement in the Blantyre coma score. Conversely, a decrease in platelets counts was associated with a smaller improvement (+0.34) in the coma score among survivors (P < 0.04, by Mann-Whitney test). No patients with a worsened level of consciousness had an increased platelet count (Table 1). These clinical results, as well as other findings from a recent pathologic study,8 suggest that the role of platelets in cerebral malaria is likely to be more important than previously thought.
|
Moerman and others have rightly underlined the association between infection with human immunodeficiency virus (HIV) and thrombocytopenia in childhood both in developed and developing countries.9,10 In our series, children had no history of HIV and were not systematically investigated for antibodies to this virus. However, prevalence rate of infection with HIV among children hospitalized in the Pediatric Unit of Hôpital Principal in Dakar was less than 0.2% (i.e., less than 10 new cases each year) during the period of the study. In Senegal, this rate was 0.05% among children (0–15 years old) at the end of 2001.11 Thus, infection with HIV cannot be considered a confounding factor in our Senegalese study.
What about other settings? In the absence of preventive interventions, the risk of mother-to-breastfed child transmission of HIV is generally estimated to be approximately 30% in developing countries. Most HIV-infected infants develop disease during the first year of life and have a high mortality rate,12 up to 89%, by three years of age, as was shown in a prospective study in Malawi.13 Consequently, it is unlikely that the prevalence rate of HIV infection among children 0–5 years of age exceeds 1% in countries where this prevalence is high among adults (e.g., 10%). Indeed, it has been estimated to be less than 0.1% in sub-Saharan Africa.12 Thus, it is unlikely that infection with HIV could have a significant confounding effect in the association between thrombocytopenia and poor outcome of P. falciparum malaria in children in most sub-Saharan African countries. However, we agree with Moerman and others that the validity of using thrombocytopenia as a prognostic indicator for P. falciparum malaria should be tested in settings other than the hypoendemic conditions prevailing in Dakar.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
