Complexity of implementing canine visceral leishmaniasis control measures at a population level: use of impregnated deltamethrin collars and culling

Background In Brazil, dogs are considered the main reservoir for Leishmania infantum , the main causative agent of visceral leishmaniasis (VL) for dogs and humans. The goal of this study was to evaluate the effectiveness of deltamethrin impregnated collars for controlling of canine L. infantum infection. Methods A prospective case-control intervention study was performed in paired neighborhoods, implementing deltamethrin-impregnated collars in one area and culling VL dogs in both. Four cross-sectional serosurvey of canine, and one of human L. infantum infections were performed. Sand flies were monitored along 19 months. Bayesian model and Moran’s index of canine L. infantum infection rates were used to detect spatial autocorrelation. Results A total of 11,285 dog evaluations were performed over 27 months. At baseline, rates of canine L. infantum infection differed between areas: 10% in the intervention and 19.7% in control area [odds ratio 0.454 (95% CI: 0.364, 0.566)]. Human L. infantum infection also varied, with rates of 4% and 14.3% in the intervention and control areas, respectively. Comparing dogs and humans L. infantum infections within each area, the OR of infection were 2.405 [95% CI: (1.720, 3.363) in the control area and 1.459 [(95% CI: 1.117, 1.906) in the intervention area. The pooled OR across both areas was 1.776 [95% CI: (1.441, 2.189)]. Large scale implementation of insecticide collars was effective at 6 months [ OR 0.448, 95% CI: (0.343, 0.584)], but the effect was not sustained thereafter. Nonetheless, dogs wearing collar had lower of L. infantum seroconversion (p = 0.044), and households with collared dogs had reduction in sand fly abundance. Collar coverage remained below 80%, and approximately 40% of collars were lost before the scheduled replacement at 6-month interval. Reduction in VL dog culling was observed in both areas at 6 and 12 months. Spatial analysis revealed outlier blocks of canine infection, forming clusters that influence infection dynamics in neighboring areas. Conclusions There was reduction in Leishmania seroconversion in dogs that used collars. The use of collars reduced sand fly density in the household. Culling of VL dogs were not systematic because of loss of follow up or non-consent from the owners.