Because of the traditional emphasis on quality in the industry, I find the panic over validation perplexing. Historically, the medical industry has maintained its quality through a mixture of good manufacturing practises and validation. Validation and GMP should be at the very heart of manufacturing philosophy and therefore deeply embedded in the culture of the industry. Given this history, the recent substantial emphasis on validation and validation documentation should be approached with confidence, especially given the performance in the past.
A simple answer to the question, "Why suddenly the emphasis on documentation?" is that the regulatory authorities need proof that the validation process has been properly conducted. This may seem bureaucratic, and can indeed become so, but there is an added benefit of ensuring that the documentary record forms the basis of ensuring the machine or process can be maintained within validated parameters. This helps to ensure the quality of the product is as is expected, and enables manufacturing efficiencies to improve. Indeed a good guide as to the successful completion of validation is whether the manufacturing efficiency has improved - if it hasn't, you have probably got it wrong.
Even with this benefit of potential improved quality and efficiency, the whole new emphasis placed on documentation has created a massive problem. Many companies have scrambled to put documents in place without stopping to think what is really necessary or relevant. The preoccupation with documentation has also meant that many companies have concentrated on the methodology of generating the documents, instead of ensuring that the content of the documentation is appropriate. Of course, methodology in documentation is clearly important, and following the correct systems is vital. Yet, obviously, without the correct content, the documents mean nothing.
Methodology is eminently transportable and is therefore often the sole focus of the consultant. The same methods can usually be used on different machines and in different companies, and be applied regardless.Conversely, ensuring that the content of the document is correct requires a deep understanding of the process and the machine, an understanding that will only be found in a professional who has years of experience and an instinctive understanding of the process.
A company recently sought my advice on a sealing machine that was giving continuous problems in production. The machine had completed its process qualification and I was given a copy of the validation documentation developed as a result. The documentation correctly identified that the speed (i.e. sealing time) temperature and pressure were key factors in ensuring seal integrity. Whilst superficially recording many aspects of the machine, the document failed to record many fundamental factors that affect seal quality. A good example is the hardness of the rubber when using a steel to rubber sealing media. A further example is the flatness of sealing platen or the concentricity and correct diameter of sealing rollers. This record of material hardness and dimensional compliance is absolutely vital in the validation of any sealing machine. Within the validation documentation there were no records of initial settings, no challenging of spring pressures, and, most importantly, no detailed record of material specification. It was impossible to return the machine to the condition under which the validation tests were conducted.
These factors that had not been documented were not simple temperatures or pressure settings. Instead they were machine conditions which, without deep knowledge and understanding of the machine, could not have been anticipated. Clearly, whilst understanding the methodology required, those responsible for producing the document did not understand the machine well enough to properly construct and conduct the validation process.
MACHINE VALIDATION
Fundamental to the validation process (and strangely missing from most methodologies) is a risk analysis of the machine to forecast the areas of criticality. A risk analysis helps identify all areas in the machine that contribute to product quality, and is the catalyst to blend the correct mix of methodology and content. It should be used to differentiate ethical factors (which could also be defined as avoidance of recalls) form the internally focused marketing and performance issues. The methodology used in the process of risk analysis is nothing more than applied common sense. The flow chart shows how the risk process challenges the design and forces consideration of mandatory maintenance and operating procedures.
The use of risk analysis helps focus tests on what the machine or process should not do and what can go wrong, rather than what it should do. Most machines operate satisfactorily when all procedures and manufacturer's instructions are followed. It is important to go beyond this functional testing to challenge the machine or process for all possible events. If the machine does not pick up the fault, or is not expected to pick up the fault, write an SOP to ensure manufacturing procedures are designed to anticipate or eliminate the error.
There is a possibility that the risk analysis will highlight fundamental machine problems - especially on machines that are being retrospectively validated. In the worst case, you may decide that the machine is not fit for purpose. Remember that bad manufacturing practice cannot be validated. If the machine is fundamentally unsuitable, then correct it or replace the machine before attempting to validate.
Small modifications can often be made to machines to make them both validatable and improve the repeatability of operation in production. I have never successfully validated a machine without making some modifications to the machine or the control system.
Validation of software seems the most difficult of the disciplines to grasp. Regulatory requirements are open to interpretation and much has been written in an attempt to clarify the topic. Once again, risk analysis seems to be strangely missing from all the methodologies and systems. In order to improve visibility and understanding of the software, a first step in the risk analysis process is to ensure that the software is constructed in modules, with flow charts.
Then, as in general validation, each module within the software must be examined to determine the possibilities of risks and the consequences of failure. Once again, be sure to separate the functionality (i.e. how the machine works) from the criticality (i.e. those areas that will cause product failures). There are many events within the software which may cause the machine to stop working, but, provided you can be assured that the machine will not adversely affect the product in the event of these faults, they should not be classified as critical elements from the validation perspective. Critical factors are only those which allow the machine to run in an incorrect manner and produce product out of specification.
With such a complex topic as machine validation it is always difficult to try to summarise key elements. In principle I have concentrated on the dangers of omitting experienced people from the validation process and the desirability of using risk analysis as the basis of test construction, but there are some key points that any manager should observe.
The first is that validation should not be conducted without input from members of your production team who have an in-depth knowledge of the machines and processes. Ironically, these same people should not be those who conduct the validation. Often it is better to use a scientific or engineering based professional who makes no inherent assumptions about machine performance, but does understand the process in general.
The second is that more emphasis should be placed on content of validation process than method used. A good challenge is to ask a qualified person, with relevant disciplines, if they are absolutely sure they can return the machine to its validated settings in two years time, and guarantee the results. If the answer is no, then you do not have a machine that is validated.
The third point is that a good validation process with risk analysis will highlight many GMP issues that require standard operating and maintenance procedures to ensure the risks are correctly managed in the production environment.
The final recommendation is that the focus in the validation operation should be on the points that are important, not those that are unimportant or irrelevant. The term risk analysis seems to be missing from most methodologies. Yet, in my view, to attempt to validate machines without it both over-complicates the process and creates the danger of missing the critical factors altogether.
Many regard validation as an unnecessary overhead. Since fully validating machines before shipment our installation times have reduced and the number of return visits following an installation have also fallen.
The investment that we have made in validation documentation on behalf of our customers is being recovered by efficiencies in other areas. Properly conducted, validation will both pay for itself and be a considerable benefit to the machine user. The blind bureaucratic approach being followed by some companies is an expensive process that will achieve nothing more than a temporary reprieve from the regulatory authorities.
Click on link to view article on Process Optimisation and Validation.
|
