Delacollette C, Van der Stuyft P, Barutwanayo M
Projet de Lutte contre les Maladies Transmissibles et Carentielles (LMTC), Section de Lutte contre le Paludisme, Bujumbura, Burundi.
Rev Epidemiol Sante Publique. 1993;41(5):416-21.
To evaluate malaria control measures one needs to document morbidity trends over time. Malaria morbidity incidence rates (MMIR) can be estimated from multi-round surveys that use specially designed questionnaires, but they are cumbersome and very expensive. Parasitological incidence rates (PIR), in contrast, can be estimated from malariometric indices, which are simple to collect, but are difficult to interpret in terms of morbidity. We present an extension of Muench's mathematical model which overcomes this drawback. Between June 1990 and May 1991, data were collected in Nyanza-lac, South Burundi, in two parallel studies of children under 6 years old. A fortnightly morbidity recall survey documented MMIR by age. Thick blood films were collected bimonthly to calculate PIR according to Muench's model. The proportion of high parasitaemia (PHP) was defined as the number of slides with 100% of the examined microscopic fields positive for at least one asexual form of Plasmodium falciparum over the total number of slides examined. To test the hypothesis that parasitological data can provide an accurate approximation of malaria morbidity, age specific PIR were multiplied by the corresponding PHP to obtain the incidence rate of high parasitaemia (HPIR). Close agreement was found between MMIR and HPIR (regression coefficient = 1.15; r2 = 98%). Thus, this method offers a simple, accurate and cost-efficient alternative for questionnaire-based morbidity surveys.
