Investigations (also known as GMP Failure Investigations) are a requirement of 21 CFR 820. It is expected that medical device manufacturers immediately conduct an investigation on all non-conforming or out-of-specification (OOS) products. There are three phases to an effective investigation process:
As soon as a non-conformance happens, the individual performing the task should immediately work with their supervisor to identify where the process went wrong. If there is not an apparent reason for the deviation, the supervisor reports the issue to the quality assurance department. This will trigger the formal investigation, with the intent to establish a corrective action. A checklist should be utilized to identify the source(s) of the non-conformance. This will evaluate all aspects of production leading up to that point: sampling, handling, instrument performance, test procedure, etc. All procedures and root causes should be documented. The QA department will work to identify potential corrective actions to prevent a future recurrence.
If the cause of the non-conformance cannot be identified, a more detailed evaluation of the manufacturing process should be conducted. If you are unable to establish a manufacturing error, and the QA department determines the deviation has no identifiable cause, then it may be decided that the issue is not representative of the whole batch or process.
In the final phase, your goal is to identify the most probable cause of the non-conformance. The final report should include reference to the procedure(s) and/or process(es) that need to be improved, along with identification of the person(s) responsible for implementing the corrective actions needed to prevent recurrence of the situation.
There should be a predetermined time allotted for completion of the investigation process. If this time frame is not sufficient, the FDA requires documentation with a valid justification for the delay.
Let’s talk data loggers!
Talk with one of our experts to get your questions answered and see how we can help you solve your data logger pain points.