[Evaluation of control measures for visceral leishmaniasis (kala azar) in an endemic area in Bahia, Brazil (1995-2000)].

This study evaluated the Control Program for Visceral Leishmaniasis in the city of Feira de Santana, Bahia, Brazil. Information was collected on control measures from 1995 to June 2000, and focused on periodicity, coverage, and incidence trends in the target areas. The dependent variable was the incidence rate for human cases, and independent variables were: percentage of buildings visited in the canine survey and buildings sprayed, number of cycles in the canine survey and spraying, and canine prevalence rate. To evaluate the correlation between control measures and the human visceral leishmaniasis incidence rate, we calculated the Pearson s correlations coefficients and performed linear regression analysis. No association was observed between the prevalence of dogs with positive serology and the incidence rate of human cases. The variables that remained in the final model and that proved most effective in controlling the disease were: percentage of buildings visited during the survey and spraying and number of cycles in the canine survey, capable of explaining 40% of the incidence of human cases. The findings constitute relevant material for the discussion of effectiveness in control measures for visceral leishmaniasis.