Measurement uncertainty of chloramphenicol in food products by LC-MS/MS

摘要

One of the requirements of the new ISO norm 17025 for accredited test laboratories concerns the measurement uncertainty of results. This latter has been estimated for chloramphenicol for food products containing honey. As a consequence of the zero tolerance level of choramphenicol in food, the calculation addressed the content of chloramphenicol at the lowest detectable quantities. The measurement uncertainty value obtained in repeatability conditions was compared to intermediate reproducibility i.e. long term precision of the method. Since the values are not significantly different we can state that, for this method, intermediate reproducibility corresponded to an adequate estimation of measurement uncertainty. Furthermore, measurement uncertainty was used for the determination of both decision limit CCα and detection capability CCβ, criteria introduced by the European Union to replace detection limit. These limit values were not different from those calculated under intermediate reproducibility conditions as recommended by the european guidelines. Hertce limits can be obtained before having carried out tests under intermediate reproducibility conditions, which needs more time. Moreover, the estimated measurement uncertainties were compared to the empirical approach by the Horwitz's formula. All the values calculated with this formula were higher than the values obtained from validation data. This means that using the Horwitz approach for a first measurement uncertainty estimation there is no risk of its underestimation for the chloramphenicol analysis.