[Quality of diagnosis of canine visceral leishmaniasis in epidemiological surveys: an epidemic in Belo Horizonte, Minas Gerais, Brazil, 1993-1997].

In Brazil, visceral leishmaniasis poses a serious endemic and epidemic problem. In recent years the disease has spread to several States that had not previously reported any human or animal cases. For canine surveys, the Ministry of Health currently recommends use of the indirect immunofluorescence assay (IIF), with a sensitivity of 90-100% and specificity of 80%. Use of IIF may decrease the effectiveness of the Kala Azar Control Program by not detecting, and thus failing to sacrifice, false-negative infected animals. On the other hand, the program incorrectly identifies and leads to the unwarranted sacrifice of false-positive uninfected animals. Such uncertainties have already led to disagreement over the Program's objectives, such as the recommendation by some veterinarians that infected animals be treated. This work analyzes the quality of diagnosis of canine epidemic surveys against the backdrop of a visceral leishmaniasis epidemic in Belo Horizonte, Minas Gerais State, from 1993 to 1997.