92 Chapter 5 Figure 1. Weighted Deming regression for cross-validation. Emicizumab concentrations using the modified, calibrated one-stage clotting assay (mcOSA) are plotted against emicizumab concentration using liquid chromatography-tandem mass spectrometry (LC-MS/MS) method in patient samples (n = 77). Purple line is the regression fit (-1.61 + 1.02*X; Pearson’s r = 0.986); purple area represents the 95%-confidence interval (jackknife method) of the fit; dashed red line is line of identity. A Bland−Altman analysis was performed on absolute and relative differences. The absolute differences had a mean bias of 0.03 µg/mL (SD = 4) with 95% LoAs ranging from -9 to 9 µg/mL (Figure 2A). No trends or outliers were observed. The relative differences (Figure 2B) had a mean bias of 2% (SD of 11), with 95% LoAs from -20 to 25%. The mean difference between methods was <20% in 71 of 77 samples (92%), which is well within the acceptance criterion of >67% of samples. The six samples with >20% difference had a mean emicizumab concentration ranging between 4 and 35 µg/mL ; the mean absolute difference of these six samples was 3.9 µg/mL. Four of six samples retained a difference of >20% after re-analysis with LC-MS/MS. Figure 2. Bland−Altman difference plots for cross-validation. Absolute (A) and relative (B) differences in emicizumab concentrations obtained by modified, calibrated one-stage clotting assay (mcOSA) and liquid chromatography-tandem mass spectrometry (LC-MS/MS) method in patient samples (n = 77) were plotted against the corresponding mean emicizumab concentration (µg/mL). Straight line is mean bias; purple area (between the dotted lines) represents 95%-limits of agreement.
RkJQdWJsaXNoZXIy MTk4NDMw