Measurement uncertainty evaluation of a HPLC method for niacin quantification in foodstuffs

Abstract nº 13, publicado no livro de resumos do "3rd IMEKO TC8, TC23, TC24 - Symposium on Traceability in Chemical, Food and Nutrition Measurements".Measurement uncertainty estimation is a required tool to evaluate analytical method accuracy in order to obtain accreditation by ISO 17025. Our goal was to estimate measurement uncertainty of an RP-HPLC method, for niacin quantification, in foodstuffs. An intralaboratory approach, using in-house validation data, obtained in precision and trueness studies was used, to include all contributions. To estimate precision associated uncertainty (µP), five matrices were tested in four different dates, in intermediate precision conditions. Reference material, FAPAS 2166 was tested to estimate trueness associated uncertainty (µT). Combined standard uncertainty (µc) was calculated from the square root of the quadratic addiction of µP and µT. To calculate expanded uncertainty (U), a coverage factor of 2 was applied to (µc). Additionally, uncertainties for the individual matrices were estimated. The estimated analytical measurement uncertainty, U, was 19%. Calculated U for individual matrices were 17, 18, 16, 19 and 18% for canned tuna, green peas, roasted peanuts, wheat bran and powder milk, respectively. The major contribution to the analytical measurement uncertainty estimated was the one associated to trueness, due to the high uncertainty associated to the reference material (14%). Accordingly to the results obtained, niacin content in foodstuffs should be reported with 2 significant figures