Laboratory Compliance: Making The Grade
Mon, 06/01/2009 - 11:13am
What is "laboratory compliance"? It is a general term used by many people, but which means different things to different people. There is no definitive answer or even one in Wikipedia the "on-line" encyclopedia! For me, it is the name applied to the procedures, policy and general approach used in a laboratory to ensure that the analytical results generated are meaningful, valid and suitable for the decisions which will be made from the data. This fundamental statement applies to all laboratories and can be represented by the quality pyramid shown in Figure 1. What differs about laboratory compliance between organizations and industries is:
* The approach they use to demonstrate that the data is valid
* The level of rigor they apply to the documentation
* The level of compliance monitoring used
* Where most emphasis is placed in the pyramid in Figure 1
Even the laboratory compliance terminology used can differ in meaning – another potential source of confusion. Many of the terms encountered are used in an interchangeable manner. For example, calibration, qualification, verification or validation may mean something very similar or something very different – at present it is necessary to clarify the exact meaning of the phrase by asking the person who uses it for examples (e.g. "what do you mean by" …, "can you give me a specific example"…? ). The clarification provides a context which enables us to interpret the meaning of the phrase and compare this meaning with our own understanding. This level of detailed understanding is critical when comparing qualification services between organizations.
One of the key secrets to robust compliance which supports value driven analysis and a low risk audit defense strategy lies in understanding the inter-relationship between the laboratory compliance levels of the pyramid in Figure 1. The levels must not be considered in isolation from each other and all of the levels are fundamental parts of the laboratory Quality Management System (QMS). However, any level will have a different general interpretation in a given industry. Additionally, independent of industries the approach used by suppliers to qualify their instruments varies considerably. The impact of this variation means that presenting a harmonized instrument qualification approach, as part of a robust regulatory audit defense is unexpectedly difficult – because of the different and fragmented qualification approach by suppliers. The remainder of this article will focus on Analytical Instrument Qualification (AIQ), acknowledging that this should not be considered in isolation from the other levels shown in Figure 1.
Analytical Instrument Qualification
A detailed history of trends in Analytical Instrument Qualification (AIQ) has been published previously1. The purpose of this article is to provide a framework to help people understand some of the implications of this history against laboratory compliance needs, explain the origin of variation in supplier qualification and provide an update on more recent information / guidance. Fundamentally, analytical instruments must be qualified (to show they are working properly) before analytical methods are developed and validated using the instrument. If the instrument is not working properly, the method development and validation are invalid.
Qualification of analytical instruments was first implemented in the pharmaceutical industry, although the general principles of needing a framework to document that an instrument is suitable for its intended use are applicable to all laboratories and all industries. Would a laboratory knowingly buy an expensive instrument which was not appropriate, use it in a way which was not valid and knowingly make decisions from data which simply not suitable? The answer is NO. However, by not qualifying the equipment, this is exactly the risk which laboratories operate.
The initial qualification has to include consideration of what the instrument will be required to do (the URS - User Requirement Specification). A simpler way of understanding this is to consider things such as:
* What samples will be tested
* How will the samples be tested (what methods and operating ranges / accessories will be needed)
* What results the instrument be used to generate
* What level of uncertainty is needed (especially important for trace analysis)
* What level of confidence is needed in the results (in relation to the process capability2)
The original FDA validation Guidelines were developed for Pharmaceutical manufacturing process validation3. When analytical instrument qualification was in its infancy, the process of interpreting these guide lines and then re-applying them to analytical instrumentation resulted in significant "process validation" bias and variation in interpretation. Hence, the approaches taken by organizations to AIQ, from multi-vendor qualification suppliers (independent of manufacturer), suppliers of analytical instruments through to industry and compliance consultants varies significantly. Typically, variation is modest for the IQ, but can be significant for the OQ and PQ, even down to the scope of what these sections should contain.
In addition to this historical fragmentation - in absence of a more authoritative guide from regulators, the Pharmaceutical industry in particular turned to Good Automated Manufacturing Practice (GAMP) for a validation framework, which introduced a software-driven approach. Earlier this year, GAMP 5, "A Risk Based Approach to Compliant GxP Computerized Systems", was published4.
However, in August 2008, the United States Pharmacopeia (USP) monograph on Analytical Instrument Qualification <1058> became effective5, so the scope and content of this monograph has now been defined and it is available for all who have access to the USP.
GAMP 5 Vs. USP<1058>
Both GAMP 5 and USP<1058> employ risk-based thinking and increasingly, the general principles of risks based thinking and also being applied by regulatory auditors. Therefore, there is an increasing expectation that laboratories will use risk based thinking. One way to overcome IQ/OQ/PQ variation is to move to a support model for the laboratory where all of the equipment maintenance and qualification needs are supported by the same organization, harmonizing the qualification rationale.
In principle, these differences could be resolved by defining and agreeing what the DQ, IQ, OQ and PQ should contain, as well as the other differences between USP<1058> and GAMP 5. One of the most common areas of difference is the terminology. Table 1 shows a definition of common terms to help reduce confusion in this area. See Reference one for more information on the qualification options laboratories face.
Such a harmonization process is beyond the scope of this article. USP<1058> and GAMP 5 are both examples of independent consensus derived documents, therefore any harmonization, even if the two independent groups which supported their development agreed to it, would need to be consensus based. Table 2 shows a comparison of definitions from USP<1058> and GAMP 5. Even where the definitions are similar, they still may be different to those defined in individual laboratory QMS or company policy, therefore laboratories need to be aware of both GAMP 5 and USP<1058> for potential impact on their policy and procedures. Additionally, the definitions must not be read in isolation from the parent reference document (which provides the context). Fundamentally, there are many similarities to what both are trying to achieve, some of which include:
* Definition of roles and responsibilities
* Outline of the qualification process
* Implementation of risk based thinking
* Potential simplification of the qualification process
* Maximize use of supplier derived information
Because they have been developed in isolation, they approach qualification from different perspectives. GAMP 5 is a rigorous project management approach to qualification derived from a software development focused perspective which can be applied to the most complex bespoke software and/or instrument. The GAMP 5 book is 352 pages of well developed and well written text and is supported by a number of Good Practice Guides (GPG) developed by industry GAMP Special Interest Groups (SIG). If a laboratory was designing a new Laboratory Information Management System (LIMS) or some other bespoke software package, GAMP 5 would be excellent.
For most laboratories however, the process may simply be too rigorous for analytical instrument qualification and non-pharmaceutical laboratories in particular will be deterred from the "apparent" complexity of the approach. On the other hand, USP Monograph USP<1058> (which was developed from the American Association of Pharmaceutical Sponsored (AAPS) meeting "A Scientific Approach to Analytical Instrument Validation"6 is 5 pages long and provides a very flexible, pragmatic and much simplified approach to laboratory equipment qualification, particularly common sense and attractive for the more basic laboratory instruments. It also generally considers software as a core part of the instrument – so in qualifying the instrument, the software is also qualified. An integral part of USP<1058> is the sub-classification of analytical instruments into Groups A, B or C, where the qualification process is potentially simplifies and scales down: View the chart
The driving force behind this was the need to simplify the qualification approach for more basic laboratory equipment. It is not known how the different approaches of USP<1058> or GAMP 5 will impact on a regulatory audit. Although they have a different approach, the two documents do not contradict each other – largely because of the flexible way USP<1058> has been written, stating "performing the activity is far more important than the phase under which the activity is performed". Both are guidance documents (<1058> is in the section of the USP which is not mandatory) and therefore, until a number of regulatory inspections have been carried out after August 2008 (when USP<1058> became effective), there is a potential level of uncertainty about how strongly and quickly organizations need to align and how the wording in USP<1058> will be interpreted by regulators such as the FDA.
Using an HPLC system as an example, each module of the HPLC system will be subject to tests during the instrument qualification. The full system must also be subject to holistic testing, where confirmation is required that the components of the system perform correctly when connected together.
Fundamentally, what level of testing should be included in the qualification? Additionally, should the calibration tools (e.g. flow meter and temperature probes) be used "as is" – where the reading from the device is taken as an "absolute" result. Or, should a metrology approach be used where the reading can corrected for measurement bias and the uncertainty of measurement is also reported? Clearly the metrology approach to tool calibration and qualification is better, the results are more accurate. One of the risk of using "as is" calibration results, is that "true" reading may be sufficiently different to fail a "passed" setting, or, worse still, "pass" an instrument which fails!
A metrology based approach to qualification naturally overcomes these difficulties – as calibration of the tools automatically corrects for measurement bias and the uncertainty of measurement is designed in and reported with the results. This leads to greater accuracy of results and a higher level of confidence that the instrument is working correctly.
Ideally, the full operating range of the instrument must be qualified as extrapolation of results is not good practice and results in greater risks. Additionally, as well as spanning the range of operation, instead of simply measuring the actual minimum and maximum setting, it is desirable to include settings which are close / at the most commonly used values.
If analytical instrument qualification was being introduced now for the first time, many of the causes of variation which are historically related would probably not exist. However, we do not have that luxury, therefore, we have to manage the variation and the differences and decide how laboratories achieve their compliance. Out of the two main qualification approaches introduced this year4,5, the impact of these on regulatory audits is currently unknown and will only become better understood following a number of FDA audits, when interpretation of these two documents is understood. Both GAMP 5 and USP<1058> have advantages. GAMP comes into its own for large bespoke software systems (such as LIMS or custom robotic automation). However, for most laboratory equipment, it is may be too complex to be practical (with the exception of laboratories which are already aligned).
USP<1058> is much simpler than GAMP – but only provides a general framework for qualification, it does not provide the detail. In doing so, it simplifies the process, removes the mystique and therefore makes it easy for all laboratories to understand and follow.
The variation in qualification approaches is a cause of frustration because it makes the strategic job of audit preparation difficult. There is no strong driving force within the service industry to harmonize approaches. Therefore, if a laboratory wants to implement a harmonized consistent approach to analytical instrument qualification, it will either have to qualify all the equipment itself (the "DIY" approach1) or move to an outsourced multi-vendor qualification model.
When considering the qualification approach a supplier is using, it is essential to understand the detail of the approach, especially when prices are being compared. Out of the qualification approaches used, a metrology driven qualification approach which includes both qualification of the operating range of the instrument and settings that are commonly used in the applications (methods) presents the ideal option in the opinion of the author.
1. BioProcess International; Vol. 5, No 10, Trends in Analytical Instrument Qualification, Paul Smith, November 2007.
2. Statistical Process Control, J. Oakland, Butterworth Heinemann; 6E edition, ISBN 10-978-0750669627.
3. CBER/CDER, Guidelines on General Principles of Process Validation. US Food and Drug Administration, May 1987; www.fda.cder/guidance/pv.htm.
4. GAMP 5: A Risk-Based Approach to Compliant GxP Computerized Systems, ISPE Publications, Florida USA, January 2008
5. USP31-NF26 through First Supplement, General Chapter <1058> Analytical Instrument Qualification, 3587-3591, Effective 1st August 2008.
6. Qualification of Analytical Instruments for Use in the Pharmaceutical Industry: A Scientific Approach, AAPS PharmSciTech 2004; 5 (1).
7. 1058 Analytical Instrument Qualification, Horacio N. Pappa, AAPS Workshop on Pharmaceutical stability testing to support global markets, September 2007.
Paul Smith can be reached at firstname.lastname@example.org or +44 (0) 1494 679 292