Application of a forward design-based multi-attribute decision-making method in quality assessment in pharmaceutical...

Area of Science:

• Pharmaceutical Technology
• Quality Control
• Computational Modeling

Background:

• Pharmaceutical tablet quality, especially for chewable forms, faces challenges due to stringent regulations and unique sensory requirements.
• Assessing mechanical properties and sensory attributes of chewable tablets necessitates advanced quality control methods.

Purpose of the Study:

• To implement a forward design-based multi-attribute decision-making (MADM) framework for assessing pharmaceutical chewable tablet quality.
• To develop a robust weighting strategy and a predictive model for comprehensive quality evaluation.

Main Methods:

• Utilized batch-based tableting technology to produce chewable tablets, evaluating critical quality attributes (weight variation, friability, porosity, hardness).
• Implemented a game theory (GT)-based combinatorial weighting strategy within the MADM framework.
• Developed a backpropagation neural network (BPNN) model to predict a comprehensive quality index.

Main Results:

• The GT-based weighting strategy outperformed traditional methods (AHP, EM) in evaluation robustness and accuracy.
• MADM results demonstrated strong agreement with external validation using tensile strength data.
• The BPNN model achieved high prediction accuracy for the comprehensive quality index.

Conclusions:

• The proposed MADM framework with GT weighting provides a reliable and accurate method for multi-criteria quality assessment of pharmaceutical chewable tablets.
• The BPNN model offers quantitative support for process optimization and intelligent quality control.
• This approach contributes to the advancement of smart pharmaceutical manufacturing through data-driven optimization and automation.