Abstract
In this study, we examined the diagnostic accuracy of the InBios Scrub Typhus Detect IgM enzyme-linked immunosorbent assay (ELISA) and determined the optimal diagnostic optical density (OD) cutoffs for screening and diagnostic applications based on prospectively collected, characterized samples from undifferentiated febrile illness patients in northern Thailand. Direct comparisons with the serological gold standard, indirect immunofluorescence assay (IFA), revealed strong statistical correlation of ELISA OD values and IFA IgM titers. Determination of the optimal ELISA cutoff for seroepidemiology or screening purposes compared to the corresponding IFA reciprocal titer of 400 as previously described for Thailand was 0.60 OD, which corresponded to a sensitivity (Sn) of 84% and a specificity (Sp) of 98%. The diagnostic performance against the improved and more-stringent scrub typhus infection criteria (STIC), correcting for low false-positive IFA titers, resulted in an Sn of 93% and an Sp of 91% at an ELISA cutoff of 0.5 OD. This diagnostic ELISA cutoff corresponds to IFA reciprocal titers of 1,600 to 3,200, which greatly reduces the false-positive rates associated with low-positive IFA titers. These data are in congruence with the recently improved serodiagnostic positivity criteria using the Bayesian latent class modeling approach. In summary, the InBios Scrub Typhus Detect IgM ELISA is affordable and easy-to-use, with adequate diagnostic accuracy for screening and diagnostic purposes, and should be considered an improved alternative to the gold standard IFA for acute diagnosis. For broader application, regional cutoff validation and antigenic composition for consistent diagnostic accuracy should be considered.