Department of Medical Parasitology, London School of Hygiene and Tropical Medicine, Instituto de Medicina Tropical Alexander von Humboldt, Universidad Peruana Cayetano Heredia, London, United Kingdom, Peru
Risk factors for cutaneous leishmaniasis were identified from a comparative study of transmission rates in 27 villages in the Departments of Lima, Ancash, and Piura in Peru. To evaluate regression analysis as a tool for the incrimination of sand fly vectors in the absence of other biologic evidence, univariate and multivariate logistic regression analyses were used to identify which of 14 variables (the abundance of nine sand fly species, four social factors, and region) predicted transmission rates in villages (incidence, active prevalence, or cumulative prevalence). In general, suspected or proven vectors (e.g., Lutzomyia peruensis) had the strongest associations with transmission rate, indicating that regression is a useful supplementary method of incriminating vectors. Regression was then used to quantify the importance of suspected risk factors. Transmission rate increased with the abundance of Lu. peruensis, Lu. ayacuchensis, Lu. noguchii, and, to a lesser extent, Lu. verrucarum and transmission was higher among villagers who slept more frequently in temporary shelters in crop areas. There were also weak effects of the number of dogs/person (negative) and the number of persons/household (positive). Linear regressions failed to detect a threshold sand fly density below which transmission ceases. The minimal adequate multiple regression model explained 82% of the variance in village incidence rates. This model was used to predict the effect on incidence of reducing each of the four suspected vectors in northern and southern Peru. The results indicate that vector control programs in the south should aim at Lu. peruensis, Lu. verrucarum, and Lu. noguchii, but focus on Lu. ayacuchensis in the north.