Agreement of EV-A71 neutralization assay: serial 4-fold versus 2-fold dilution comparison

Abstract

Objective: To evaluate consistency between 2-fold serial and 4-fold serial diluted neutralization tests against Enterovirus A71 (EV-A71) in estimating titer, Geometric mean titer (GMT), seroprevalence, and seroincidence. Methods: Based on a prospective cohort of 1-9 years old children, mothers and infants established in Anhua County, Hunan Province, during 2013-2018, from which 92 participants with a total of 386 blood specimens were sampled and tested with a 2-fold serial dilution and a 4-fold serial dilution neutralization tests against EV-A71 at the same time. Agreement was estimated using the Bland-Altman method. Stratified analysis was conducted to estimate effect dilution approach on GMT, seroprevalence and seroincidence. Results: The mean difference (0.04, 95%CI:-0.02-0.10) between the two dilution approaches was not significant. However, the limits of agreement (LOA) (-1.12-1.21), with the 95% confidence interval of upper LOA (1.10-1.31) and of lower LOA (-1.22--1.02), significantly exceeded the Clinic accept interval (-1, 1) indicating insufficient agreement between the two approaches in practice. While the dilution approaches did not affect estimates of GMT of the total population and the positive population, and seroincidence with seroconversion only, the differences were 2, 6 and 2%, respectively (P>0.05). Estimates of seroincidence with at least 4-fold increase and seroconversion/4-fold increase were significantly higher using a 4-fold dilution neutralization test compared to the 2-fold dilution neutralization test with 8% (95%CI: 1%-12%) and 9% (95%CI: 1%-17%), respectively. Conclusion: The 2-dilution and 4-dilution neutralization tests yielded comparable results when estimating the population's GMT; however, the difference between the two is not negligible when assessing the seroincidence.