2.5. Data and statistical analyses
To analyse the results, descriptive statistics were performed. The
linear correlations between the results of the clinical laboratory and
the results of the MSD multiplex immunoassay were evaluated by the
Pearson correlation coefficient (r ). This coefficient was also
used to determine the intra- and inter-assay reproducibility and the
difference in incubation time of the MSD multiplex immunoassay. To
assess the comparability of the ImmunoCAP assay and the MSD multiplex
immunoassay, a Bland-Altman analysis was performed.16A Bland-Altman plot is a scatter plot of the difference between the two
methods’ means (y-axis) against the mean of the two methods (x-axis).
The calculated mean difference is shown and the 95% confidence limits
of agreement are obtained by calculating the mean difference plus or
minus two times the standard deviation of the differences. The
sensitivity and specificity of the MSD multiplex immunoassay were
determined, both in comparison to the skin prick test and the ImmunoCAP
assay of the clinical laboratory. The threshold set for a positive test
in both the MSD and ImmunoCAP assay was 0.1 kU/L. When checking for
outliers, subject 3 was eliminated from the analyses: positive scores
were observed for all MSD multiplexed measurements (all sIgE-levels
between 0.3 and 0.7 kU/L) and consistent negative scores for the other
assays. Moreover, this person reported to be diagnosed with essential
thrombocythemia (ET), which could explain the divergent results.