Abstract
This paper examines the sensitivity and specificity of two ELISA assays for IgM antibodies to Mycobacterium leprae, one employing natural phenolic glycolipid and the other employing a synthetic disaccharide glycoconjugate as antigen. Estimates of sensitivity and specificity are derived, based on a panel of sera from leprosy cases in Malawi and various non-leprosy controls from the UK. Though both assays were able to identify a high proportion of multibacillary patients, neither was able to detect a high proportion of paucibacillary patients without considerable loss of specificity. The implications of the inverse relationship between sensitivity and specificity are discussed with reference to the predictive value of such tests in such areas as Malawi, where the large majority of cases are paucibacillary.