RELEASE OF METOPROLOL FROM SOLID DOSAGE FORMS. CHOICE AND VALIDATION OF THEORETICAL MODEL

The study presents the results of modeling the release kinetics from a metoprolol extended release formulation and comparison of models using information, statistical and phenomenological criteria. Following the fact that extended release was obtained by embedding the active substance in a hydrophilic matrix, plausible models were considered Higuchi, Noyes Whitney and Weibull models. Application of mathematical criteria indicated Higuchi's model as optimum in fitting the experimental data with a minimum number of parameters (only one parameter). Examination of residuals indicated only slight differences among, experimental and predicted data. Differences in the extrapolations based on models were greater than differences among interpolated values. An "outlier" prediction appearing in the case of Noyes Whitney model. The correction of the Noyes Whitney model by renounceing, to the restriction that the solution intercepts x-axis in origin improved predictions and criteria classified the model as optimum. Penalization of the model by considering the elimination of restrictions as addition of a second parameter situated again Higuchi model in first position. As a Final conclusion it was considered that the comparison of models has to take into consideration mathematical, statistical and phenomenological criteria as well as plausibility of interpolated and extrapolated data.