Within pharmaceutical manufacturing, process validation builds quality into the operation at every step. The result is consistent product quality. It seems straight forward but how does one achieve this when it involves a myriad of multiple systems (PLCs, HMIs, control systems, etc.) and requires a team effort that encompasses people from various disciplines within the plant?
For starters, it is important to design processes that ensure product quality, safety and efficacy. Today’s software systems can help validate pharmaceutical manufacturing and steer the most appropriate design of processes in the right way. Such systems can deliver clear documentation, provide version control for the multitude of systems in use and backup to ensure all risks are mitigated.
Part of the validation process is in the documentation -- there needs to be integrated support for documentation with 100% clarity and traceability. Version control of all software systems used within the pharmaceutical manufacturing process are necessary to ensure compliance with the current good manufacturing practices (CGMPs). These regulations state that manufacturing processes must be designed and controlled to ensure that in-process materials and the finished product comply with quality requirements that have already been predetermined and assure consistency and reliability. 21 CFR Parts 210 and 211 encompass requirements for approving or rejecting processes and specifications that impact drug quality.
In highly regulated industries like pharmaceutical, the four eyes principle is a requirement that mandates two individuals approve some action before it can be taken and another key procedure the validation system should ensure occurs.. It is also important that the validation software can integrate seamlessly with a full range of automation devices and equipment commonly in use by pharmaceutical manufacturers.
The validation system can also act an automated backup or scheduler. It can connect to the control network with zero human involvement. It automatically reaches out to the devices in the network to obtain the current copy that's running on a device and checks it against the central repository to ensure its correct. Nothing can happen that isn't allowed to happen.
There is also flexibility because the user can choose how frequently they would schedule a backup. Most will do it on a daily or on a shift basis, so that any and all changes will be noted. For example, perhaps somebody walked up to a controller and put some code on it without letting anyone know. A validation system would pick up on that and flag it. Maybe it is a malicious attack or some sort of cybersecurity issue, the system will catch vulnerabilities through regular scanning intervals.
In summary, Validation is instrumental to ensure the process is repeated exactly step by step and is achieving the end result within the allowed tolerances. For instance, temperature set points for certain producing such drugs as stimulants (methamphetamine and Ritalin or antidepressants) or perhaps tolerances for thicknesses of insulin bags as an example, would have its processes validated and assigned validation thresholds. For the threshold, each individual step would be assigned a certain range such as the temperature could go from five to six or anywhere in-between but if it crosses over and gets to seven, then that would cross the threshold. The process would be flagged and product considered bad. This can be costly and why a lot of engineering work goes into figuring out what are the validated process variables.
A lot of time and money is spent on certifying one individual way of making a particular drug and why software systems have become key in the solution to help pharmaceutical companies get to that validated state. Then once they are running, the software ensures that if they need to audit or provide a guarantee, they have one. The system helps to set a protocol that requires review, signoff and signatures. The protocol is always checked and authenticated before the process can go into production.