There’s a phrase in EU GMP Volume 4 that was a pleasant surprise the first time I read it. It struck me as a very good idea, and I wondered why other GxPs hadn’t done something similar. From Part 1, Chapter 4: “The various types of documents used should be fully defined in the manufacturer’s Quality Management System.” Fully defined in the QMS. That seems like a solid approach.
EU GVP modules require records to be created to report and verify all activities, with a clear description of which documents are considered records, how they are stored, and for how long. This comes close to the same idea, though without stating it outright.
ICH E6 R2 alludes to it with the term “essential” documents, but rarely has so much confusion stemmed from such a vague definition. Sponsors and their teams move closer with their TMF indexes, yet meaningful, study-specific tailoring remains more honoured in the breach than the observance. Over time, I’ve come to believe this task belongs in the vendor engagement process, not just at study start-up.
While clinical trials involve more unforeseeable events than manufacturing, making complete definition a challenge, there’s value in setting out procedures clearly to meet legal and regulatory requirements. SOPs and work instructions should not only detail what is done but also how it’s recorded, by whom, when, and what to do when standard records aren’t feasible.
ICH E6 R3 brings us further along, stressing the need to incorporate quality and risk management early in trial design. Quality by design asks us to identify critical quality factors prospectively. It suggests a “structured content list for storage repositories may be used to prospectively identify essential records,” though it stops short of treating record-keeping as a critical factor itself. Still, its definition of essential records is an improvement.
On the subject of clarifying which records are to be created and maintained, it’s equally important to define how they are to be created. Does your organisation have a set of ‘good documentation practices’? It’s curious how often people refer to ‘GDocP’ as if it were a formally recognised GxP. Some companies have in-house versions, which is a start, but many do not. Since no global regulatory body has formally defined it, we shouldn’t overstate its importance—but some risk-based guidance from qualified professionals wouldn’t hurt, would it?
In another newsletter, I’ll argue the same case for metrics: it’s sensible to describe performance metrics explicitly in procedural documents. And yes, you should have some good performance metrics practices—GMetP.
