|
5 Ensuring reliability of analytical results
Variance of analytical results from different laboratories
increases with decreasing analyte concentrations, as W. Horwitz1
pointed out. For the analysis of BPA, phthalic acid esters or
nonylphenol in biological samples or animal feeds, which are at
concentrations of ppb order, this is one of the most important
issue for the reliability of analytical results.
Another problem is that these substances exist in the laboratory
environment and readily contaminate the samples resulting in
gross errors in analytical results. The analytical procedures
should therefore be so designed as to preclude contamination.
The guidelines described in this document have been developed
with these problems taken into account.
(1) Evaluation of reliability of analysis
As stated in the Interim Report, evaluation of analytical
results of different laboratories on the same samples is an
indispensable step for determining the reliability of an
analytical procedure.
Unfortunately, however, it was impossible to find laboratories
which would participate in systematic inter-laboratory cross
checks of the present guidelines which deal with microanalysis
of very small amounts of biological samples for components
readily confounded with contaminations. Although cross checks
for some samples was carried out among the laboratories which
took part in the development of the present guidelines, a more
general project involving more laboratories is still needed.
(2) Measures for maintenance and assurance of precision
In order to confirm that a laboratory provides sufficient
sensitivity and precision in analysis of a specific sample, it
is recommended to perform beforehand a cross check with one
(preferably more than one) reference laboratory using the same
sample.
1 J. Assoc. Off. Anal. Chem., 63, 1344(1980)
|