The ability to predict the results of bioequivalence studies based on a consistent statistical analysis of informative factors

Abstract

Predicting the results of bioequivalence studies is an urgent and important task in developing protocols for clinical studies of bioequivalence. Objective: to develop an algorithm for predicting bioequivalence results based on factors with a sufficient degree of informativeness. Materials and methods: the analysis of the main factors associated with conducting bioequivalence studies on the data of 290 bioequivalence studies; the informativeness of the factors is characterized and the corresponding points (prognostic factors) of each gradation of each factor are determined; a forecast table and forecast calculation algorithm have been developed; The sensitivity and specificity of the developed prognostic algorithm are compared with the control sample (data from 65 bioequivalent studies from open sources). Statistical processing was performed using SSPS Statistics v. 25 and Microsoft Office Excel 2016. Results: The results of the analysis performed to determine the informative factors by various assessment methods are described. Equations are given for performing the corresponding calculations. The key and most informative features are shown, on the basis of which a forecast table is created and a forecast algorithm is developed. The results of the analysis of the sensitivity and specificity of the algorithm for predicting the results of bioequivalence studies are presented. Conclusion: an algorithm and a calculator for predicting the results of bioequivalence based on factors with a sufficient degree of information content have been developed. The algorithm showed high values of sensitivity and specificity, as well as a low percentage of errors.

Keywords

• исследования биоэквивалентности
• прогноз результатов
• информативные факторы