The growing conversation on the Food and Drug Administration’s Process Analytical Technology (PAT) initiative continues. My persistent RSS search on PAT pointed to another great article, this time in Pharmaceutical Technology magazine. The article, The Five Steps to Starting PAT by Jacob Cook, discusses simplifying the process of getting started with a PAT initiative.
The five steps discussed were:
- Pick simple.
- Understand all the details and nuances.
- Evaluate the instrumentation you already have, and the information you can easily collect.
- Understand the appropriate intervals for collecting that data.
- Evaluate the tools available for reading and synchronizing the data.
Just last week we discussed the benefits of applying a structured approach to a PAT initiative to improve opportunities for initial success.
I passed this article by Christie Deitz, whom you may recall from earlier posts on PAT and ISA-88 (S88) projects. Like most initiatives, Christie believes having good data (step 3) is very important. The Life Sciences industry project teams use DeltaV Batch which integrates in a single location the data required for this analysis. This data includes: alarming, continuous and batch history, operator actions and other events. Having this information organized together around batches and campaigns can help identify PAT opportunities.
Where manufacturing execution systems (MES) like Compliance Suite are also used, exception-based reporting can also help with this process of analyzing the data. We discussed using XSL style sheets to do these reports in an earlier post. An example of this exception-based reporting is showing the batch reviewers only the alarm data that occurred during any particular batch run or campaign.
Christie also points out that where manufacturers have already implemented PAT analyzers, they can make decisions in electronic work instructions (EWIs) based on the analyzers’ real-time data values to help verify its correct operation. For example, if a PAT analyzer is not reading the expected value based on other operating data, the work instruction can be to have the operator take a manual sample against which to compare the analyzer data value.
Whether you “pick simply” as a starting point or apply a structured methodology to assess the best opportunities to begin, analyzing your existing data is extremely important. The analysis process is less manually intensive when this data is either centralized or logically organized together in some manner to help better identify these opportunities.