Low false response rates in screening a hundred veterinary drug residues in foodstuffs by LC-MS/MS with analyte-specific correction of the matrix effect

Abstract As a response to the demand for better efficiency and reliability in the monitoring of veterinary drug residues in food, liquid chromatography-mass spectrometry (LC-MS), either in tandem with quadrupoles (LC-MS/MS) or with high resolution systems (LC-HRMS), has become an attractive tool for screening purposes. However, the lack of reproducibility of the ionization yield observed by electrospray remains a major limitation for having powerful LC-MS platforms able to screen more than a hundred substances in a broad range of food on a reliable manner. False responses are not scarce, and require insightful approaches to mitigate them. This study describes a screening with analyte-specific correction of the matrix effect (SACME), as a strategy based upon the comparison of the chromatographic signals between the sample analyzed ‘as is’ and a replicate that has preliminarily been spiked at the action level. Emphasis was put on the high level of confidence in the results achieved with SACME, providing a self-quality control of each sample tested. The significant decrease in the variability of the responses compared to a screening with absolute abundance (SAA) of the signals is shown, and the perspective of screening a wide range of food matrices (raw materials, processed ingredients and finished products) without tedious technical precautions is highlighted. In the end, false response rates are compared, and data demonstrate unambiguously the benefit of the matrix effect attenuation against the common approach using absolute areas. No false positive (FP) was obtained with SACME, whilst a minimum of 2–14 compounds (over 105 substances in the scope of the method) exhibited false positive responses with SAA. With SACME, false negative (FN) rates as low as 1 to 5 compounds were obtained, whilst remained in the range 15–31 compounds with SAA.