摘要
Concordance study between two laboratories testing clinical samples is important to a clinical program to assess the comparability of test results, especially when the two laboratories performed the same test at different locations. Assays that measure immune response should not be compared using the simple linear regression model; instead literature publications recommend an ＂errors-in-variables＂ model that accounts for the variability in these assays （Home, 1998）. In this paper, concordance slope and intercept was estimated using the statistical linear relationship （Tan and Iglewicz, 1999） for various tests between laboratories. The confidence bounds for the concordance line were used to assess the impact on vaccine booster thresholds when a consistent bias was observed between the results generated by two laboratories.
