Comparison on three methods for identification of critical process parameters / 中草药

ABSTRACT

Objective: To investigate the characteristics of multiple linear regression method, stepwise regression method, and standardized partial regression coefficient method for the identification of critical process parameters. Methods: The acid precipitation and alkaline dissolution process of Scutellariae Radix (SR) water extract was investigated as an example. Results: Similar results were obtained using three different methods. The concentration of SR water extract, pH value of acid precipitation, pH value of alkali dissolution, and refrigeration temperature were considered as critical process parameters. Conclusion: The selection of threshold values of the three methods all are subjective. Multiple linear regression method is simple, but is not as sensitive as stepwise regression method. The effects of a parameter on several process indices can be considered simultaneously and easily weighted using standardized partial regression coefficient method.