Analytical methods for real time monitoring of pharmaceutical blending processes are thoroughly validated but not the sampling methods. Variographic analysis was investigated as a method to determine the sampling and analytical errors when the drug concentration of pharmaceutical powder blends is determined by near infrared spectroscopy. Variograms provide an estimate of the true process variance and the minimum possible error (MPE) defined as the sum of the total sampling error and the total analytical error under the specific sampling scenario. In this study, the MPE was 50–186 times greater for the 97.00 [%w/w] blends and 75–342 times higher for the 95.00 [% w/w] blends than the short-term precision of the analytical NIR method. This study shows that variographic analysis may be used along with concentration vs. time profiles to evaluate a blending process, and also to discriminate between the sources of sampling, analytical and process errors.

• University of Puerto Rico at Mayagüez
• KHE Consulting

• PAT
• Oral solid dose

1. MPE (minimum possible error) / True process variance
2. Multivariate data analysis
3. Near-infrared spectroscopy (NIRS)
4. oral solid dose
5. Partial least-squares regression (PLS-R)
6. pat
7. Research Article
8. Theory of Sampling (TOS)
9. Variographic analysis