Multilaboratory assessment of threshold versus fold-change algorithms for minimizing analytical variability in...

Pneumococcal vaccination is frequently used to assess a patient's humoral immune function. The comparison of pre- and postvaccination levels of antipneumococcal antibodies is widely held to be the gold standard for documenting a response. However, many of the published criteria for defining an adequate response are based on assays that are no longer widely available. We compared the clinical classification of patient response by multiplex pneumococcal assays currently performed at three large reference laboratories using a variety of published criteria for defining responses in adults. The classification of responders agreed for 79% of the patients when using a threshold-based algorithm compared to 57 to 96% of the patients when using various fold-change-based algorithms. The highest rate of discordance was seen when the most stringent criteria for response were used (4-fold increase postvaccination in 70% of serotypes). The discordant samples tended to show similar patterns of response across all three assays, with small variations in the final number of serotypes converting postvaccination. We conclude that the use of published cut points for documenting response to pneumococcal vaccination can be affected by interlaboratory differences in pneumococcal assays, particularly for algorithms that require large fold changes for a response to be documented. However, the overall patterns of response were similar in virtually all samples, regardless of the assay used.