Following these guidelines in an ideal world, a drug would progress from early development through commercial production within processes that are well designed and controlled. In reality, however, because of the lack of detailed process design plans and complete understanding of any given drug's molecular form and function, there will be imperfections that currently can be corrected only at certain lifecycle stages.
Thus, despite the progress shown in these guidance documents, the Q9 guideline on risk management and the OOS guidance still fail to properly address two important supporting quality elements for validation. First, these guidance documents do not set clear expectations for how to set risk-based acceptance criteria in validation protocols. If validation protocols don't contain appropriately set limits, future risks to patients or firms may not be controlled. Second, and even more important, these documents do not include any instructions on how to deal with validation failures. It is as if validation failures simply do not exist or are "not allowed."Current industry practice often demands that risks be managed by the extent of validation studies executed but not so much by actually using risk-based acceptance criteria. The lack of a quality system that allows a firm to openly deal with validation failures, similar to the way OOS test results are handled, has several consequences, none of which is good for the patient or the firm. In contrast, openly accepting validation failures and having a pre-set plan to address them would encourage firms to set risk-based acceptance criteria.
The OOS guidance document provides excellent instructions on how to deal with OOS results and how these results should trigger process improvements. A similar system for dealing with validation failures would add great value to the validation process. Let us first start by accepting the fact that validation failures—like OOS results—do occur. Then, we can talk about how to set challenging validation acceptance criteria. Establishing an open system to address validation failures will allow the important concepts contained in the ICH Q8–10 guidelines to be used more fully. And that would benefit the industry, and more importantly, patients.
Stephan Krause, PhD, is the director of quality control at Favrille, Inc., San Diego, CA, 858.526.8026, firstname.lastname@example.org
1. Krause SO. Validation of analytical methods for biopharmaceuticals—A guide to risk-based validation and implementation strategies. Bethesda, MD: PDA/DHI Publishing; 2007 Apr.