The method was validated with regard to its intra- and interassay
precision and accuracy by analyzing beverages with a BVO
content of 5, 15, and 25 lg/mL on two separate days. These beverages
were prepared by spiking a commercially obtained citrus soft
drink devoid of BVO with the appropriate levels of BVO. On each
day 15 independently spiked samples and 15 unspiked samples
were prepared for each BVO concentration and analyzed by
LC–MS, for a total of 45 spiked samples and 45 unspiked samples
in each of the two days. These analyses revealed excellent intraand
inter-assay statistical performance of the method (Table 1)
with accuracies ranging between 97.3 and 103.4% and precision
errors under 3.6%. It should be noted that a crucial element to
achieve this level of statistical performance was the consistent
use of appropriately calibrated digital analytical syringes in all
sample preparation procedures. There were no observable matrix
effects noted in the analysis of samples, probably due to the considerable
dilution that the samples undergo in their preparation
(20-fold dilution) and the very high retention of BVO, thus,
enabling an early elution of interfering matrix components.