Better Process Understanding Improves Quality, Lowers Risk

Published on: 
Pharmaceutical Technology, Pharmaceutical Technology-06-02-2006, Volume 30, Issue 6

Predictable outcomes lead to greater manufacturing efficiency and speed time to value.

To improve quality, and thus lower risk to consumers, the fundamental goal of an extended quality team (i.e., everyone involved in manufacturing) should be to have sufficient process understanding so that adjustments can be made in real time to achieve the intended quality outcome of the current batch and every subsequent batch. Today's reality is that manufacturers often have insufficient process knowledge to achieve this goal because they do not measure all critical quality attributes while the batch is running. A process is well understood when:

  • all critical sources of variability are identified and explained;

  • variability is managed by the process;

  • product-quality attributes can be predicted accurately and reliably.

Justin O. Neway, PhD

Accurate and reliable predictions are a reflection of process understanding, which is inversely proportional to risk. Conventional process measurements often don't provide enough information about the root cause of unacceptable outcomes. Investigational analysis using data from the processes that lead to unacceptable results is required to prevent recurrences and to introduce improvements that reduce future variability. Conversely, a well-understood process reduces the need for final product testing because the process is under control while it's running (i.e., in real time or relevant time). Parametric release, or real-time release, is based on the idea that the more you understand about your process, the more predictable the outcome and the lower the risks to consumers.

In 2003, the US Food and Drug Administration started emphasizing process understanding as a way to improve quality. Part of the solution lies in process analytical technologies (PAT), which include not only real-time measurements, but also software systems for continuous and discrete data collection as well as data access and analysis systems that enable better process understanding and control.

As part of its risk-based inspections program started in 2004, FDA is encouraging improved process understanding and control to lower variability and to improve quality and productivity. A site risk potential (SRP) score, developed by FDA as a way to prioritize plants for inspection, is made up of facilities risk (e.g., establishment type and defect history), product risk (e.g., prescription, injectable, or over-the-counter drugs), and process risk (e.g., process controllability and contamination potential). The date of the last inspection also is factored in.

A manufacturer can't change its fundamental business or production processes or its record of past inspections readily, but it can improve its understanding of current processes and use that information to improve the control of future variability. Making better choices when designing future manufacturing processes also can reduce the process risk portion of the SRP score and improve future inspection records.

Science-based quality systems that solve data-intensive manufacturing problems are the key to better process understanding that reduces process variability and thus the SRP score. Four recommended areas of focus include:

  • Immediate process development and manufacturing batch and production data review and trending with widespread reporting and review. This, of course, requires close collaboration between process development and manufacturing teams;

  • faster investigation and reporting of process-development experiments and manufacturing-batch deviations;

  • fast access to and investigational analysis of continuous and discrete data in a combined form (enabled by PAT);

  • practical ways to capture paper data for use in combination with electronic data by providing a more comprehensive view of the entire manufacturing process.

These choices lead to better control of process variability, which is good for business, beyond pleasing FDA and reducing the likelihood of inspections. More predictable outcomes lead to greater manufacturing efficiency and can speed time to value for new products. Accepting responsibility for quality improvements, therefore, requires the extended quality team to examine critically all the data from its manufacturing processes throughout the product life cycle and use the best technologies to enable process understanding and control.

Justin O. Neway is executive vice-president and chief science officer of Aegis Analytical Corporation, 1380 Forest Park Circle, Suite 200, Lafayette, CO 80026, tel. 303.926.0317,