engleski

Quality assurance and control of methods to examine visually recognisable substances in feed and food : In cooperation with International Association for Feedingstuff Analysis, Section Feedingstuff Microscopy

Visual examination of visually recognisable substances, including microscopy, focus on targets orcontaminants such as particles of animal origin, plant seeds, spore bodies of moulds, sclerotia, packaging material, microplastic and‘Besatz’(everything that differs from the norm). The twoprincipal results are counts (numbers) and weights for macroscopic methods, or presence/absence for microscopic methods. The level of detection equals at least the size of one unit, usually with a weight exceeding 1 mg, which is in the range of parts per million (ppm). Theseparameters do not follow a normal distribution but Poisson (counts), lognormal (weights) orbinomial (Booleans) distributions, with effect on the interpretation of validation parameters. Asfor other domains, examination methods for visual monitoring need to be properly validatedand quality control during actual application is needed. In most cases procedures for validationof visual methods are based on principles adopted from other domains, such as chemical ana-lysis. A series of examples from publications show inconsistent or not correct implementationsof these validation procedures, which stress the need for dedicated validation procedures.Identification of legal ingredients and composition analysis in the domain of visual examinationrelies on the expertise of the laboratory staff, therefore validation of a method usually includesthe validation of the expert. In the view of these specific circumstances, a Guidance for qualityassurance and control of visual methods has been developed, which are being presented anddiscussed in this paper. The general framework of the Guidance is adopted from ISO standards(17023, 17043, 13528). Part 1 of the Guidance includes the general background, theory and prin-ciples. Part 2 presents the actual validation procedures with experimental designs and equationsfor calculating the relevant parameters, and can be used as blueprint for a SOP in a qualitymanagement system. An EURL and NRL network for physical hazards is strongly recommended

Microscopy ; visual methods ; particles ; in-housevalidation ; quality assuranc

nije evidentirano