Epidemiological implications of the use of various methods for the diagnosis of canine leishmaniasis in dogs with different characteristics and in differing prevalence scenarios.

Our aim was to establish the influence on the prevalence of canine leishmaniasis (CanL) of the following: (1) the use of different diagnostic techniques; (2) different positivity thresholds; (3) selection of animals either at random from a population or focused on symptomatic individuals, (4) the function which the dog performs; and (5) scenarios with differing epidemiological characteristic. Three groups of dogs were analysed (416 sampled at random from an endemic area, 71 with symptomatology compatible with CanL and 15 from a non-endemic area) using three serological techniques (indirect fluorescence antibody test (IFAT), Kalazar Detect(TM) and Q Letitest ELISA) and a PCR-ELISA. The diagnostic technique had a considerable influence on the CanL prevalence value obtained. Uncertain antibody titres were more representative in dogs sampled at random and with the IFAT technique. Although employing different capture antigens, correlation of results was higher between the two commercial techniques in the group of dogs with symptomatology compatible with CanL. The sensitivity and specificity values of the different diagnostic techniques were affected by the epidemiological characteristic of the area under study, the presence of clinical signs and the function which the dog performs. This must be taken into account when comparing endemicity in different geographical areas, such as in studies carried out for the construction of risk maps. Using more than one technique, and adopting the criterion of considering an animal to be positive only when it has been diagnosed as such by more than one technique, considerably raises the prevalence values but maintains the differences between areas with different characteristics.