Measurement uncertainty evaluation from correlated validation data: Determination of elemental impurities in pharmaceutical products by ICP-MS

Abstract

Pharmaceutical products as well as active pharmaceutical ingredients (APIs) are checked for levels of elemental contaminants to guarantee medicines administration will not involve the consumption of level of contaminants greater than their maximum admissible intake. However, the conformity decision is affected by the measurement uncertainty function of analytical steps performance, used standards quality and how measurement performance is assessed during method validation. When an ingredient is considered conform, since the measured concentration is lower than the maximum limit, the risk of a false acceptance depends on how close the measured concentration is from the limit and on the measurement uncertainty. The analytical methods used for pharmaceutical analysis should be validated by ICH and USP recommendations, in order to prove measurements are fit for purpose. The validation must also be economically feasible and have an acceptable duration. This work discusses how to evaluate the uncertainty of elemental analysis in pharmaceutical ingredients from data collected during the validation of the analytical method by following ICH guidelines and USP chapters. A top-down uncertainty evaluation based on results from the analysis of a model API intermediate, with the native analyte after spiking at three concentration levels, where analyses are performed by two analysts in two different days, is presented. The impact of the correlation of some uncertainty components of collected results on the uncertainty evaluation is discussed and considered in the calculations. The developed measurement model was checked by a cross-validation procedure where some validation data was randomly removed and used for an independent model control. The developed uncertainty evaluation methodology was successfully applied to the analysis of Pd in a model API intermediate by ICP-MS after a micro-wave assisted acid digestion, where the risk of a false acceptance of the pharmaceuticals is determined. The measurement performance data and used spreadsheet are made available as Supplementary Material.