Statistical approach on retrospective validation process of polyvitaminic mixture

Abstract

The regulatory demands related to the labeling of food, mainly the one constituted of polyvitaminics and minerals destined to lactants and children in their first infancy, require manufactured process that assure the labeled content. Therefore, the suppliers of polyvitaminics have to demonstrate that the process of powder homogeneity leads to expected results, according to the specifications. Thus, the objective of this work was to validate the process of the mixture of the following vitamins: pyridoxine hydrochloride, nicotinamide, ascorbic acid and calcium pantothenate, based on retrospective approach. Control charts were employed for the statistic analysis of the data related to the assay of each vitamin targeting the evaluation of the stability of the process. The capacity indexes, Cp and Cpk, were calculated after the evaluation of the data distribution using Anderson-Darling test. All data showed normal distribution, except for nicotinamide assay. This data was treated by a method proposed by Box-Cox (lambda equal -4.24) aimed at obtaining the normality of the data. The capacity indexes (Cpk) obtained were 1.01, 1.14, 1.21 and 1.96, respectively, for pyridoxine hydrochloride, nicotinamide, ascorbic acid and calcium pantothenate. Therefore, the process of polyvitaminic mixture can be considered validated.