Abstract
Agreement between Helicobacter pylori IgG antibodies measured using the Pylori-set EIA-G kit in serum, plasma and successively thawed specimens was studied and the implications for epidemiological and clinical studies assessed. Plasma titres may differ from serum titres by -6% to +8% and therefore may be substituted for serum. The change in titre around the cut off value was -0.31 (se = 5.7, p = 0.96) per thaw. The estimated maximum drop after three thawings, 34.5, would result in only a small decrease in sensitivity (1.3%). For qualitative epidemiological studies, this additional misclassification rate is relatively small. However, positive titres did reduce over successive thawings, with the estimated maximum drop being 11.4% per thaw. Therefore, thawing does need to be considered as a contributing factor when interpreting titre drops in eradication trials. Baseline and follow up specimens from clinical studies should be thawed once only and tested concurrently.