Abstract
Detection of anti-Strongyloides IgG in urine by enzyme-linked immunosorbent assay (ELISA) for diagnosis of strongyloidiasis reportedly has comparable performance to conventional serum assays. Initial comparisons of urine assays using commercial ELISA kits designated for serology have shown its diagnostic potential but sub-optimal accuracy. In the present study, we optimized urine ELISA protocols based on different antigen types and evaluated their accuracies in determining the epidemiology of strongyloidiasis in Northeast Thailand. Paired urine and fecal samples of 966 individuals from the study community were collected for three consecutive days and tested for strongyloidiasis. We compared three ELISA protocols using different antigens including crude S. stercoralis antigen (Ss-ELISA), crude S. ratti antigen (Sr-ELISA) and recombinant NIE antigen (NIE-ELISA) and fecal examination by agar plate-culture (APCT) technique and formalin-ethyl acetate concentration technique (FECT). The optimized ELISA protocols using three different antigen sources yielded significantly higher prevalence rates of strongyloidiasis (58.9-65.1%) than those by fecal examination methods (19.7%). The prevalence of strongyloidiasis determined by ELISA protocols significantly increased with age (p value < 0.0001) and males had higher prevalence than females (p value < 0.0001). Diagnostic agreements between ELISA protocols were moderate (κ = 0.461-0.586) and the agreement between each ELISA protocol and fecal examinations were slight (κ = 0.139-0.210). The results obtained by urine ELISA protocols using three different antigens showed comparable diagnostic performances, provided further supports for the utility of urine as an alternative clinical specimen for diagnosis of strongyloidiasis.