188 documents is enough that "check the PMS references" stops being a task and starts being a process. At that scale, the temptation to sample a subset and extrapolate is real, and it's exactly the wrong call for a record that has to hold up in an audit.
Why sampling isn't enough
A post-market surveillance reference record is only as strong as its weakest document. A sample check can tell you the record is probably fine. It can't tell you which specific document has an outdated reference source or a format that drifted from what the QMS expects. For a record that gets pulled apart during an MDSAP or notified body audit, "probably fine" isn't a defensible position.
- Checking each document against its reference source
- Checking each document against the expected format
- Logging every discrepancy found for correction
- Applying the same verification method consistently across all 188
What a consistent method buys you
Applying one verification method to every document, rather than a spot-check plus assumptions, produces a record you can actually defend. When an auditor asks how the PMS reference record was confirmed, "every document was checked against its source and expected format" is a complete answer. "We sampled a portion" is not.
The Real Takeaway
At scale, a PMS reference check only works with a consistent verification method applied to every document.
Spot-checking a sample isn't enough to defend the surveillance record in an audit, and the gap between "sampled" and "verified" is exactly what an MDSAP or notified body review is built to find.