Writing a validation procedure
While validation may be part of FDA requirements, the ultimate goal is to collect data to aid the manufacturer in producing a quality product.
By Karen von Holtz, CSI Testing Inc.
The quality of a product depends upon balancing engineering and personnel control measures. Quality engineering and a well-written quality control plan, rigidly followed, will result in quality products. This premise is recognized by the FDA and is detailed under Subpart G of the agency`s Quality System Regulation (QSR) document. The FDA says manufacturers must develop, conduct, control, and monitor production processes to ensure conformance to specifications.
The trick is in trying to understand what this means and how to achieve it in a cost- effective manner. The road to achieving QSR is convoluted and often difficult. Many compliance and documentation issues must be dealt with: interpretation is not standard. Numerous experts, each with their own interpretation, provide manufacturers with advice that may not be consistent, and each expert and regulator has their own idea of QSR compliance strategies.
The reality of the standard is that the FDA does not require the manufacturer to have cleanrooms. However, it is well known that having a cleanroom or controlled environment improves the quality of the finished product and it shows the FDA that extra measures have been taken to produce a quality product. Although cleanrooms are a significant expense, in the long run manufacturers can produce a product whose quality is more consistent because they don`t have to deal with the added variable of an uncontrolled environment. The best advice on this is: expensive is not better, and, don`t cut corners on the air handling system. The key to an effective cleanroom system is balancing and integrating its components: the maintenance of the room and its mechanics and controlling the actions of the people working within it.
The first part of this process is the physical manufacturing environment and the environmental controls. The FDA`s QSR, Subpart G 820.70 Environmental control states: “Where environmental conditions could reasonably be expected to have an adverse effect on product quality, the manufacturer shall establish and maintain procedures to adequately control these environmental conditions. Environmental control system(s) shall be periodically inspected to verify that the system, including necessary equipment, is adequate and functioning properly. These activities shall be documented and reviewed.”
This portion of the QSR is a discretionary requirement. This means that the degree of environmental control should be consistent with the intended use of the product, and the details of how to achieve these controls are left for the manufacturer to decide. The manufacturer must have the proper environmental control measures necessary to ensure the quality of the finished product.
The cleanroom/controlled environment (CR/CE) needs to be tested initially to ensure that it meets its design specifications and to ensure that contaminants will not enter through its envelope. A program of expanded testing should then be designed to ensure that the CR/CE can consistently meet these requirements and maintain its integrity when subjected to distress. Therefore, the CR/CE needs to be tested during periods of activity.
Because manufacturing in a CR/CE is a dynamic process, conditions cannot be duplicated. This is why validation testing of the CR/CE is done in phases, over a period of time, and under various production conditions. This method will yield sufficient data so that patterns and trends to indicate environment changes can be identified. Depending upon how much historical data is necessary or useful, these data may be collected during different stages of production, during different shifts, or at different times during shifts. Also, the FDA would like to see this historical data collected over a period of time that includes seasonal changes. Without a defined period and program of data collection, it may be impossible to set realistic alert and action limits.
Environmental control factors and test specifications: When initially setting up the environmental control parameters, manufacturers should consider what conditions will impact the quality of the product. From there, they must determine the specifications, tolerances, frequency of testing, and action/alert limits. Caution should be taken in setting limits before knowing the ability of the controlled environment to meet those limits. Some examples of environmental control and specifications may be:
HEPA filtration will filter the air to produce an initially clean environment. The number of HEPA filters will affect the amount of clean air available, the number of room air changes and the room pressure differentials. Filters are best tested using an aerosol challenge. The test methods and specifications are found in IES RP-006.2.
Proper filter velocities will ensure that enough clean air reaches the critical work surfaces and will provide enough air changes to regularly “wash” the room of contamination. Filter velocities are best taken with a multi-point digital measuring device or capture hood. Test methods and specifications are found in IES RP-006.2.
Effective use of positive pressure will prevent ambient contamination from entering the environment. The amount of positive pressure is the direct result of sufficient supply air volume and dampering controls on the room returns when it does not escape through unsealed construction joints. Test methods and specifications are found in IES RP-006.2.
Particle counts are controlled initially through effective HEPA filtration and, thereafter, by contamination control practices. Definitions for cleanroom classes, determination of sample locations and sample times, and acceptance criteria are found in Federal Standard 209E.
Microbial counts are taken on surfaces, in the air, and of water systems to determine the amount and types of contaminants in order to evaluate the potential effect on the product.
Proper gowning and rules of conduct will control personnel contamination and protect the product.
A well defined and implemented program of cleaning and maintenance will assure uniform environmental performance.
Testing phases to be considered
Installation qualification: In this phase, the components that control the environment are tested and evaluated to assure performance to design specifications. In some instances, this is also called the commissioning phase. The air handling/air supply system is tested for rated performance and ability to handle an additional load. The HEPA filters are leak tested to ensure the media is leak free and is properly installed by verifying the absence of bypass leakage. The velocity of the air through the filters is tested to see if it meets specification and that each of the filters within the environment has velocities within a uniform range of each other to reduce turbulence within the room. The room pressure differential is tested to verify proper positive pressure and the magnehelic gauges are calibrated to the measured pressures. Particle counts are taken and baseline bioburdens are established.
Operational qualification:Performing operational state qualification establishes that the environment is capable of operating within intended limits and tolerances while in production.
It is strongly suggested that operational qualification testing be performed in three phases because this is the area of the qualification where manufacturers need to evaluate all the operating parameters of the CR/CE. A three phase approach is recommended:
Static: The room in final production clean, with all production equipment in place and running but without production personnel. The information gathered from this testing will give you a valuable baseline from which to judge the variable production stress results. Testing may include filter velocities, pressure differentials, particle counts, microbial testing and other established environmental controls as necessary.
Dynamic/normal: This testing phase requires normal production for the CR/CE with a full complement of production personnel. Testing requirements in this state would be the same as in the static phase, but these results allow the added load of the normal production process to be evaluated.
Dynamic/stress: In this phase, tests are conducted under what the manufacturer deems to be “worst case condition” for the CR/CE. These conditions may exist during or at the end of a special part of the production cycle, at the end of the last shift of a full work week, or at any time during which a stress period is identified or predicted. The test requirements would be the same as performed in the other phases. It is critical to know that the CR/CE is capable of controlling the environment within established tolerances and limits while under adverse conditions and that the product manufactured during this time of stress continues to meet product specifications.
Performance qualification:On-going performance qualification confirms that the CR/CE is consistently operating within established limits and tolerances, is continually effective in maintaining the controlled manufacturing environment, and to detect changes in performance during seasonal changes. Test requirements would be the same as performed in the other phases and the product continues to be evaluated during normal production.
An essential part of the validation program is recording test conditions. These would include time of day, activities, number of personnel present, meteorological conditions, room temperature, test equipment used and their calibration documents, test personnel, and deviations from normal conditions. These notations are essential in the successful evaluation of test data.
Other considerations: Equipment: The validation plan might also list the approved test instrumentation and that all critical equipment is calibrated to NIST standards. Calibration documents should be part of each report.
References: Listing of all the standards and recommended practices from which the testing specifications are taken.
Acceptance criteria: This is the section where the manufacturer would list the specifications to be met. Some of the criteria have been established in the design of the room. For the remainder of the acceptance criteria that may not be established until the end of the validation study, the manufacturer would list the conditions of collection, how the data is to be evaluated, and when the limits are to be set.
Frequency of testing: Each test designated may not be performed during each phase of testing. For instance, the HEPA filter leak test need only be performed once a year unless certain predetermined conditions occur. These conditions may include unexpected damage, upgrade to the system, or long term power outages.
Particle counts and microbial testing are required more frequently than other testing parameters. Depending upon the class of the room, this testing may take place once a month for three months, then every other month until the end of the validation study. If the class of the room is low, the initial test cycle may be once a week until early baseline data is collected, then somewhat less frequently until the end of the validation study.
Karen von Holtz is the CEO of Certification Services Inc. and CSI Testing Inc. CSI specializes in the testing/certification/validation and repairing of cleanrooms and controlled environments. With 16 years experience in the clean manufacturing technology industry, Ms. von Holtz consults with controlled environment owners to find cost-effective solutions to the unique challenges involved with manufacturing in a clean environment. Her areas of specialization include design, construction protocol, facility startup, contamination control, validation protocols, and personnel training and education.
 |
 |
This paper was presented at CleanRooms West `98. For a copy of the conference proceedings from CleanRooms West `98, contact Libbey Duggar at 603/891-9462. For information on other CleanRooms shows, including CleanRooms East `99 (March 22-24, 1999, Philadelphia, PA), contact Nuala Kimball at 603/891-9267.