Agriculture Reference
In-Depth Information
of the prepared bulk test material. It also provides essential information on how
participant
s results should be statistically treated in order to provide an unbiased
assessment of performance.
Laboratory performance in a pro
'
ciency test is dependent on (i) the skill and
experience of the analyst, (ii) the analytical method used, and (iii) the equipment/
instrumentation used. It should also be remembered that performance in any particular
PT round pertains solely to the speci
c test material provided and the analyte(s) it
contains. Participating laboratories are expected to use the same method that they
would normally use to analyze the samples on a daily routine basis.
Once the PT provider has received all the participant
s results and assessed them
statistically, they will provide a report that allows the laboratory to see how they
performed in comparison with the other laboratories. Within the report, the results
from each laboratory are coded so that anonymity is retained. If in the report the
performance is not deemed to be satisfactory, then the cause of the poor performance
must be investigated and established. Remedial action must then be taken in order to
negate the particular problems associated with the poor performance.
Most pro
'
ciency tests demand the use of quantitative methods and numerical
results are required to be reported. However, for qualitative screening methods the PT
provider will only request information on the identities of each analyte that has been
detected in the test material. In this case, performance will be dependent on the
robustness of identification technique used and the scope of the method. Any analyte
that was present in the test material but was not detected and therefore not reported
would be deemed to be a false-negative result. Obviously, the more the false-negative
results, the poorer the performance. Reporting a false-positive result (reporting the
presence of an analyte that was not present in the test material) is also an indicator of
poor performance, but this may be less important as if an analyte is tentatively
detected during screening analysis it will normally be followed up by a second
analysis using a quantitative method.
3.6 VALIDATION OF QUALITATIVE SCREENING MULTIRESIDUE METHODS
FOR VETERINARY DRUG RESIDUES IN FOODS
3.6.1 EU Legislation Covering Method Validation for Veterinary Drug
Screening
Chapter 3.1.3 of the Annex to Commission Decision 2002/657/EC [18] describes a
comprehensive approach for the initial validation of methods for different matrices
(e. g., muscle, liver, and kidney) and different species (e.g., bovine, porcine, ovine,
and poultry). Such a comprehensive approach makes the task of initial method
validation time consuming and hence expensive. However, a more recent publication
from the Community Reference Laboratories [3] provides a more pragmatic approach
to method validation. This document provides practical guidance on how to validate
screening methods based on biological, biochemical, and physicochemical detection
principles.
In this section, only the validation of multiclass methods using
