HOW EXPERIMENTAL DESIGN CAN IMPROVE THE VALIDATION PROCESS. STUDIES IN PHARMACEUTICAL ANALYSIS.

A critical discussion about the possibility of improving the method validation process by means of experimental design is presented. The reported multivariate strategies concern the evaluation of the performance parameters robustness and intermediate precision, and the optimisation of bias and repeatability. In particular, accuracy and precision improvement constitutes a special subset of experimental design in which the bias and the relative standard deviation of the assay are optimised. D-optimal design was used in order to plan experiments for this aim. The analytical methods considered were capillary electrophoresis, HPLC, adsorptive stripping voltammetry and differential pulse polarography. All methods were applied to real pharmaceutical analysis problems.