Understanding the Sources of Error in PAT Methods That Monitor Continuous Processes

All analytical methods are subject to random and systematic errors. PAT methods are not an exception. This conference will summarize the lessons learned in monitoring the drug concentrations of flowing powder blends in continuous mixing systems, feed-frames, and a recently patented stream sampler.

Sampling errors occur in PAT applications, and are much larger than analytical errors. However, they are not estimated in most PAT systems. Sampling errors can be easily estimated through variographic analysis in real time monitoring of continuous manufacturing. They may also be determined in batch manufacturing by monitoring the drug concentration in a feed frame, or a stream sampler. A comparison of sampling and analytical errors in various PAT systems will be presented. Through this understanding, it will be possible to validate sampling in PAT methods.

References

1. Esbensen, K. H.; Romañach, R. J., A Framework for Representative Sampling for NIR Analysis–Theory of Sampling (TOS). Handbook of Near-Infrared Analysis 2021, 415.
2. Sierra-Vega, N. O.; Romañach, R. J.; Méndez, R., Real-time quantification of low-dose cohesive formulations within a sampling interface for flowing powders. Int. J. Pharm. 2020, 588, 119726.
3. Sánchez-Paternina, A.; Sierra-Vega, N. O.; Cárdenas, V.; Méndez, R.; Esbensen, K. H.; Romañach, R. J., Variographic analysis: A new methodology for quality assurance of pharmaceutical blending processes. Computers & Chemical Engineering 2019, 124, 109-123.