[An epidemiological survey to discover the probable places of infection with sleeping sickness in the Central African Republic].

The sleeping sickness focus at Nola-Bilolo in the forest region was the subject of a survey of 142 patients in order to discover their probable places of infection. Seventy of these patients had been detected by active case-finding during a survey conducted in January-February 1991; the other 12 had been discovered by passive case-finding during 1990. The sample of actively detected patients is significantly younger (average 22 +/- 3 years) than the sample detected passively (30 +/- 3 years). This survey highlights the geographical heterogeneity of transmission in the focus. The places where infection is highest are M'Poyo, which with 18% of cases forms the epicentre of the focus and the main reservoir of infecting flies, followed by Bilolo (9%), Mékara (8%), Modigui-Kouna, Ziendi and Domissili (7%). The patients, 95% of whom belong to the M'Bimou tribe, all engage in agriculture; the main activities are coffee-growing for men and the steeping of casava in the river for women. These two activities determine the main places and times of human-tsetse contact. The information obtained by listing the cases detected passively from medical registers is inadequate for locating the places of transmission. The information gathered during active case-finding campaigns by the mobile teams offers a better approach. The method used here, which is recommended, consists of matching the replies given by the patients about their points of contact with the tsetse flies, not just with their places of residence but also with their successive movements and their places of work. The rigour demanded by this kind of approach can only be ensured by a specific survey. This survey also made it possible to demonstrate substantial movements from village to village, which appear to be characteristic of this focus. The results indicate that poorly targeted vector control is likely to be doomed to failure. The study makes it possible to specify priority areas for vector control by trapping and to avoid a "blind" control strategy whereby the whole focus is swamped with traps. The latter strategy is not financially feasible for many countries of tropical Africa.