PT as a tool to point out criticalities in the strategy for control of antibiotic residues in milk: the Italian experience

Abstract

Screening represents the first step for antibiotic residues control; thus, screening methods play a key role in the entire control procedure. The Italian National Reference Laboratory (NRL) for veterinary antimicrobial residues in food from animal origin organized the first proficiency test (PT) for the screening of antibiotic residues in milk to assess the competence of the official control laboratories. The 38 participants received two series (respectively, at 1.8 % and 3.5 % of fat mass fraction) of eight blind lyophilized milk samples, spiked with cloxacillin (CLX), benzylpenicillin (PEN), sulfadiazine (SDZ) and oxytetracycline (OXY) at once and twice the respective maximum residue limit. In addition, an unspiked sample (blank) was also included in each series. Since the vast majority of false-negative results was associated with OXY, leading to low sensitivity and accuracy, laboratories performances were evaluated separately for OXY (sensitivity 16.44 %, false-negative rate 62.25 % and accuracy 44.54 %) and, respectively, for CLX, PEN and SDZ assessed together (sensitivity 99.56 %, false-negative rate 2.53 % and accuracy 99.63 %); the false-positive rate resulted equivalent to 0 % and the specificity to 100 %. The criticality identified by the PT was that OXY was not detectable under the routine screening analysis conditions adopted because of the use of only two screening methods (Delvotest SP-NT and CMT) both unfit for its detection at mass fractions of interest. As a consequence, the NRL provided guidance to participants to implement different and complementary screening methods. The effectiveness of the corrective actions undertaken by participants will be evaluated with the next PT.