Monitoring the performance of pharmaceutical assays and identifying sources of variation is difficult. Developing a system to detect and predict potential issues allows for early intervention, reducing both costs and risks. Factors such as equipment, chemical batches, documentation, and personnel must be considered to determine sources of variation.
Combining these elements effectively reduces "time to detect" and "time to understand." Multivariate analyses allow comparing multiple inputs at once for a more holistic view. Enhancing assay detection is achieved by monitoring the stability index and Cpk across multiple runs, especially during development phases. Annotating with contextual parameters simplifies identifying variation sources, enabling management to recommend training or address supplier issues.
This paper presents an application developed using JMP, which trends results, detects shifts, and helps determine root causes of variation. It also offers reporting on assay validity and lab errors over time, providing a high-level overview of the state of deployed assays.
