[en] An innovative versatile strategy using Total Error has been proposed to decide about the method’s validity that controls the risk of accepting an unsuitable assay together with the ability to predict the reliability of future results. This strategy is based on the simultaneous combination of systematic (bias) and random (imprecision) error of analytical methods. Using validation standards both types of error are combined through the use of a prediction interval or β-expectation tolerance interval. Finally, an accuracy profile is built by connecting, on one hand all the upper tolerance limits, and on the other hand all the lower tolerance limits. This profile combined with pre-specified acceptance limits allows to evaluate the validity of any quantitative analytical method and thus their fitness for their intended purpose.
In this work, the approach of accuracy profile was evaluated on several types of analytical methods encountered in pharmaceutical industrial field and also covering different pharmaceutical matrices. The four studied examples depicted the flexibility and applicability of this approach for different matrices ranging from tablets to syrups, different techniques such as liquid chromatography, or UV spectrophotometry and for different categories of assays commonly encountered in the pharmaceutical industry that are content assays, dissolution assays and quantitative impurity assays. The accuracy profile approach assesses the fitness of purpose of these methods for their future routine application,. It also allows to select the most suitable calibration curve, to evaluate adequately a potential matrix effect and propose efficient solution and to define correctly the limits of quantification of the studied analytical procedures.