What Level of Validation Does Each of My Systems Require?

At Kymanox, we have managed a great deal of validation projects, from manufacturing equipment to computer systems and software. As with any validation, one of the decisions that needs to be made is determining the appropriate level of validation for each system in the systems inventory. As discussed in my last post, there are a number of different levels of validation that can be performed on a given system, and the VMP must specify the level of validation that each system will receive. However, a common question that we encounter when managing validation projects is "How do we determine the appropriate level of validation since each system serves a different purpose?"

The appropriate level of validation for each system depends on the potential risk the system has on the product with regard to quality, safety, and efficacy. ICH Q7, Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients (1), and Annex 15 to the EU Guide to Good Manufacturing Practice (2) provide some general definitions and guidance for each type of qualification. Systems that indirectly and minimally impact the overall quality of the product would possible only require an IQ or a simplified combined IQ/OQ. Examples of these systems include:

• Utility systems (e.g., glycol chiller)
• Secondary packaging equipment (e.g., tape dispensers for final shipping boxes)

Any issue with these types of systems will have no direct impact to product, and ultimately the patient. Therefore, minimum validation is performed to qualify the system for use.

If a system could directly impact the quality of the product, the system will require more validation. However, if the system is not directly involved in a process method, only an IQ and OQ may be necessary to ensure the system is capable of performing as intended within normal operating parameters. In some instances, an IQ, OQ and equipment-based PQ may be required. Some examples of systems that might fall into this group are:

• A product holding tank with no critical operating parameters such as heating or cooling liquid to a specific temperature
• Mixers for simple tanks and solutions
• A remote printer for printing manufacturing trends

A fault in these systems may not directly impact the final product. However, these systems must be functioning properly for the overall process to function. As a result, more validation is performed to qualify the systems.

If the system has a direct impact to the safety, quality, or efficacy of the product, the system will require a much more extensive validation. This also applies to systems directly involved in a portion of a process method. There are many validation level combinations that can be performed on each system, including those systems with direct impact to the product. Examples of such systems may include:

• Environmental chambers
• Chromatography systems
• Bioreactors
• Packaging and labeling equipment

An error in one of these systems could directly affect the product and patient - and therefore require more extensive validation.

When making the final decisions about the level of required validation, the Regulatory and Quality Assurance groups must be consulted. This will ensure that the overall validation approach is minimally sufficient, defendable, and in line with current industry practices and FDA expectations. The examples provided herein are simplified hypothetical scenarios; your scenario may be different and require a different approach.

Medical device systems have an additional decision that must be made when determining the required level of validation. Validations for medical devices do not always need to be performed. Process verification – usually by 100% visual inspection – can be performed in place of validation for some types of processes. The GHTF guidance on process validation (3) provides detailed information on validation versus verification decisions.

The decision to verify rather than validate is centered on the process output. If the process output is fully verifiable by inspection, verification can be performed in place of validation. However, it must also be sufficient and cost effective to verify rather than validate. If the process output cannot be fully verified by inspection or it is not sufficient and cost effective to verify, then process validation will most likely be performed. The systems inventory should contain a column for verification for any medical device systems to ensure the validation has been effectively planned.

It is also essential that there is a general understanding of the overall process in order to avoid over-qualifying the systems. It is always “better to be safe than sorry”. However, over-qualification due to a lack of knowledge adds a great deal of time and cost to validation. When there is a detailed understanding of the process, reduced validation can be justified. This risk-based approach is consistent with guidelines outlined in ICH Q9, Quality Risk Management (4), and GHTF guidance on implementing risk management principles (5). Specifically, a lot of software packages are over-validated as compared to their actual use and impact to the product and patient; in many cases, a black-box approach to testing software is sufficient and can save considerable time and money which can be spent on systems needing additional validation attention.

A sampling strategy with statistical justification must also be described in the VMP for the systems in the systems inventory. The pre-determined sampling plan is combined with the pre-determined acceptance criteria in the specific validation protocols. Gone are the days of N=3 as acceptable default sample sizes. The new expectation is that a statistically justified sampling strategy is implemented for all validation activities and documented a priority.

There are several methods for determining appropriate sampling strategies with statistical justification. The methods include:

• Performing pre-validation activities (engineering runs)
• Utilizing commissioning data
• Applying previous knowledge or experience with similar equipment, processes, etc.
• Using estimates of variability (standard deviation) based on data from pre-validation activities and previous knowledge of the systems
• Employing documented risk assessments for reduced sampling

Performing pre-validation activities is the most useful method for determining an appropriate sample strategy, and therefore sample sizes. Pre-validation helps identify behavior and variability in the systems. Also, performing pre-validation activities will save time, money, and various quality-related issues during validation execution. Whatever method is utilized to determine an appropriate sample strategy for validation, the determination process must be documented in the VMP.

My next post will move on to discuss the second planning focus of VMPs, which is planning the validation schedule. Please feel free to contact me with any questions or comments you have regarding this blog. Thanks!

Best regards,

Megan K Gladfelter

References:

1. ICH, Good Manufacturing Practice Guide for Active Pharmaceutical Ingredients Q7, International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use 10 November 2000.
2. Final Version of Annex 15 to the EU Guide to Good Manufacturing Practice, Qualification and Validation July 2001.
3. Quality Management Systems - Process Validation Guidance, Document Number GHTF/SG3/N99-10:2004 (Edition 2) January 2004.
4. ICH, Quality Risk Management Q9, International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use 09 November 2005.
5. Implementation of Risk Management Principles and Activities within a Quality Management System, Document Number GHTF/SG3/N15R8 20 May 2005.