Aseptic Processing Risk Management — Challenges & Common Misconceptions

By Hal Baseman and Amanda McFarland, ValSource , LLC

In 2006, the Parenteral Drug Association (PDA) task force writing Technical Report No. 44¹ sent out a survey on aseptic processing risk.² Shortly after, some expressed concern that the survey might imply that aseptic processing is risky, even though a well-controlled aseptic process may be less risky than a poorly controlled tableting operation. Others found fault with this assertion, pointing out that because the impact of an aseptic processing failure could be loss of product sterility — resulting in serious patient safety issues — aseptic processing is, by its very nature, high-risk.

However, is this correct? Risk is not based strictly on severity. Risk is the combination of impact of a potential failure (if it occurs) and the probability of that failure occurring.³ If risk was based on severity alone, aseptic processing risk could not be reduced, since the impact of sterility loss on the patient cannot be reduced. To apply this logic further, if there is no way to reduce severity, then there is no way or reason to try to improve the process. The problem is that once a process is identified as “risky,” we tend to accept that risk and do little to try to uncover ways to reduce that risk and improve the process.

What’s Stalling Aseptic Process Improvement?

Aseptic processing of sterile drug products can be improved. Strategies for controlling process and process risk have changed little in past decades. The result, at times, has been a failure to adequately control risk — the similarity of FDA 483 observations and warning letter citations associated with aseptic process control from year to year is evidence of such failures.⁴ While other technology-based industries have moved forward, ours seems to be mired in acceptance of the status quo.

There are several reasons for this lack of improvement:

1. Delays associated with regulatory approval of post-approval change notifications for the use of new technologies — The effort and time required to obtain post-approval changes can be a significant and real challenge when considering process improvements.⁵ However, we believe that justifying process improvement as a means to reduce process risk — thus increasing product quality assurance, and patient safety, as presented in the third reason (below) — is a key to clearing a path for regulatory reform.
2. Cost and return on investment of new technology for existing processes — Justifying the investment needed to modernize aging facilities used for existing product lines can also be a challenge. However, it is interesting that ours is perhaps one of the only industries that fails to recognize the cost/benefit of process failure reduction. We tend to see a direct relationship between increasing process quality/control and increased cost, or perhaps better put, an inverse relationship between quality and efficiency. It seems the opposite would be more logical. As process quality improves, reject levels decrease, batch yields increase, downtimes decrease, plant utilization increases, failure investigations decrease, supervisor floor time increases, and efficiency should improve.
3. Failure to realize how process improvements can control that risk — The key to process improvement is process understanding and properly controlling process variables that may pose a risk to product quality. More effective process control begins with using critical thinking to better evaluate process risk, and using sound scientific principles to design processes that effectively reduce and control that risk.

Aseptic processing quality risk management may be more challenging than assessing risks of other processes, due to one or more of these factors:

• High severity: The unwanted result of the process failure is ultimately “loss of sterility,” which is always considered severe in relation to patient safety. Severity, as it relates to the loss of sterility, cannot be reduced.
• Weak correlation: The loss of sterility is a rare occurrence that is difficult to observe using most current methods, including standard USP <71> sterility testing. The correlation between what can be observed (environmental monitoring, smoke study results, and media fills) and the undesired outcome of a failure (loss of sterility) is not easy to quantify.⁶
• Variation and uncertainty: Human behavior is relatively unpredictable, and environmental microbiological monitoring is imperfect and subject to much variation. Therefore, even where correlation can be developed, predictability may be difficult.

5 Common Misconceptions in Evaluating Aseptic Processing Risk

We also have a tendency to be influenced by certain misconceptions when evaluating aseptic processing risk, and we believe this hampers our ability to objectively control such risk.

1. If something has not happened yet, it likely will not happen.

Despite evidence of process failure risk, we have a sense that failures from faulty aseptic process design will not result in harm, because we rarely see evidence of contaminated product. Sterility test failures do not occur all that often. We see no evidence of sterility failure, because the evidence may be hard to see. Therefore, we do not take the steps to analyze the chance of contamination. This phenomenon results, in part, from the availability heuristic, a cognitive shortcut wherein an example that can be easily recalled takes prominence over alternative solutions that are not as readily recalled.⁷ The method helps solve problems more quickly, but it also causes us to overestimate the probability of a recent outcome recurring.

2. If something has not been cited during an audit or inspection, it must be acceptable.

Passing regulatory inspections is essential, but it may also provide a false sense of security. Regulatory inspections are not comprehensive investigations of the process aimed at finding any and all weakness. They are audits designed to capture vulnerabilities in processes, either known or unknown to the organization. Just because an inspection fails to identify a weakness does not mean the process is OK.

3. People-related problems are the result of people making mistakes.

People do make mistakes from time to time, but often a flawed process is the reason for the errors. It is more effective to redesign a process than it is to fix people; however, it is easier to blame people than to fix the process. In the end, it would be wiser and more effective to look first at the process as the potential culprit, before blaming the person.

4. Detection is a highly effective way to reduce risk.

In this industry, we often rely too heavily on current detection methods to mitigate suboptimal process designs. No level of detection, especially microbiological environmental monitoring, is practically perfect. Efforts to reduce risk by uncovering the source of failure and then reducing the occurrence of that failure is better than increasing detection.⁸

5. Process qualification testing by itself proves that the process is under control.

Process qualification and testing are not enough to assure process control. Sound process design is the key to assuring effective process control. Process requirements must be identified early in the process/facility design stage of any new manufacturing operation. Process risk assessment results should drive process qualification and operational controls. Aspects and functions of equipment that affect process and product quality should be the focus of qualification testing. Acceptance criteria should be aligned with the requirements of the process. Data acquired from continuous process monitoring (CPV) should be carefully evaluated and the interpretation of such data used as a basis for process improvement, where warranted.

Conclusion

Aseptic processes involve significant hazards that must be properly controlled to avoid patient safety consequences. Much has been and can be learned from using critical risk-based decision making to properly evaluate, analyze, and design effective control strategies. The results of this approach can and should be used to improve the process, by recognizing and better controlling process risk. Monitoring, testing, and qualification can provide useful information regarding performance, but true process control is best achieved through careful analysis (i.e., risk assessment), process understanding (i.e., knowledge management), and thoughtful design.

References:

1. PDA Technical Report #44: Quality Risk Management for Aseptic Processes, Vol. 62, Issue No. S-1. April 2008.
2. Ahmed, R., et al, “PDA Survey of Quality Risk Management Practices in the Pharmaceutical, Devices, & Biotechnology Industries.” PDA J Pharm Sci Technol, 62:1-21, Jan./Feb. 2008.
3. ICH Q9: Quality Risk Management, International Conference on Harmonization, 2006.
4. U.S. Food & Drug Administration Inspection Observations: http://www.fda.gov/ICECI/Inspections/ucm250720.htm.
5. Vinther, A., “Drug Shortage is a ‘Wicked Problem’”, PDA Letter, May 4, 2016.
6. Sutton, S., “The Environmental Monitoring Program in a GMP Environment”, Journal of GXP Compliance. Vol. 14. No. 3, Summer 2010.
7. Esgate, A., and Groome, D., An Introduction to Applied Cognitive Psychology, p. 201, Psychology Press, 2005.
8. Bishop McFarland, A., and Waldron, K., Considerations Regarding the Use of Detectability In Risk Assessments, IVT Network, Dec. 21, 2015.

About The Authors:

Hal Baseman is chief operating officer and a principal at ValSource, LLC and Concordia ValSource, LLC. He has over 38 years of experience in pharmaceutical operations, validation, and regulatory compliance. He has held positions in executive management and technical operations at several drug manufacturing and consulting firms. He is the immediate past chair of the Parenteral Drug Association (PDA) Board of Directors, former chair of the PDA Science Advisory Board, former co-leader of the PDA Process Validation Interest Group, and co-leader of the PDA Aseptic Processing Points to Consider task force, as well as a long-time member of the PDA Training Research Institute faculty. He is a frequent presenter and instructor on subjects related to quality risk management, validation, and aseptic processing. Hal holds an MBA in management from LaSalle University and a B.S. in biology from Ursinus College.

Amanda Bishop McFarland, M.S., is a quality risk management and microbiology consultant with Concordia ValSource, LLC. She specializes in the creation and implementation of risk management programs and in developing risk-based strategies for use in microbiology. Amanda is an active member of the Parenteral Drug Association (PDA) and a faculty member for PDA’s Training Research Institute and an instructor for the PDA course on QRM implementation. She has a B.S. in entomology and an M.S. in mycology, both from the University of Florida.