[Critical quality attribute identification and consistency evaluation of traditional Chinese medicine mixing unit for...

To promote the transition of traditional Chinese medicine(TCM) preparation production towards intelligent and controllable manufacturing, this study addressed the challenges in practical production, such as the lack of quantitative evaluation system for physical quality attributes and the difficulty in determining batch-to-batch variability in mixing units. Based on samples from actual production batches of Tongren Niuhuang Qingxin Pills, a digital characterization strategy for the physical quality attributes of the mixing unit was established, followed by critical quality attribute(CQA) identification and batch-to-batch consistency evaluation. Sixteen physical quality attributes, including particle size, specific surface area, density, and color, were digitally quantified from 30 batches of mixed raw medicinal powder. The results showed that the particle-size-related indicators and specific surface area exhibited relative standard deviations(RSDs) greater than 10%, indicating pronounced batch-to-batch variability in the mixing unit. Subsequently, Euclidean distance was employed to evaluate sample consistency and identify discrepant batches. Combined with orthogonal partial least squares discriminant analysis(OPLS-DA) and a variable importance in projection(VIP) threshold greater than 1, the contribution of physical quality attributes was assessed. Ultimately, twelve attributes with substantial impact on batch-to-batch variability were identified, including relative homogeneity index(Iθ), particle size range(Width), particle size corresponding to 90% cumulative particle size distribution(D_(90)), and percentage of particle size<50 μm(%Pf). These can serve as critical physical quality attributes for quality control of the mixing unit. In summary, this study established a digital characterization and consistency evaluation approach for the physical quality attributes of TCM mixing units. It identified critical physical quality attributes and batch-to-batch variations, providing a basis for consistency analysis and variation elucidation. Furthermore, these findings support process optimization and quality stability of subsequent processing units, thereby facilitating the transition of TCM mixing units toward a digital and intelligent process control paradigm.