Performance indexes of a dot-enzyme-linked immunosorbent assay (dot-ELISA) and an enzyme-linked immunosorbent assay (IgG-ELISA) for field surveys of New World leishmaniasis.

Abstract

Diagnostic performance indexes of sensitivity, specificity, positive predictive value and efficiency were determined for dot-ELISA and IgG-ELISA tests in 340 leishmaniasis sera. Sensitivity of the dot-ELISA was significantly lower than IgG-ELISA's; the two tests had indexes of specificity and positive predictive value of the same magnitude. Seventy-eight sera gave a negative dot-ELISA test result and a positive IgG-ELISA test result. When sera were classified according to different criteria as how to interpret this diversity, the kappa statistic did not corroborate the classification indicating that the two tests display a substantial strength of agreement. The results presented indicate that performance indexes accrued in a survey where variables are well known may be extrapolated to other population studies if the disease presents itself as highly prevalent (due to a selection bias or not) and may be expected to discriminate a disease status among test positives.