If the signal is X, where does the risk live?
A generic, teaching-only framework for thinking about signal-based QMSR inspections: where an investigator is likely to enter the quality system, and the lifecycle path they may follow from there.
A few days ago I wrote about how the shift from QSIT inspections to QMSR inspection thinking may feel different in practice. Several people asked what I meant by a QMSR lifecycle inspection risk framework, so here is a simple generic example to illustrate the idea.
The point of the framework is not to predict exactly how an inspection will proceed. It is a way to visualise something that becomes more important under QMSR.
Inspections often begin with a signal
- Complaint trends
- Rising MDR reports
- A recall
- Recurring CAPAs
- Supplier failures
- Software anomalies
- UDI issues
Once the signal is identified, the investigator often enters the system at the place where that risk is most visible, and then follows the lifecycle of the issue across connected processes.
So instead of asking “which subsystem do we inspect next?”, the question becomes:
If the signal is X, where does the risk live, and where do we start?
Vioela’s note: A drill worth running before any inspection. Pick one real complaint and time how long it takes your team to put every linked record on the table: the complaint file, the decision on whether it was reportable, the investigation, the CAPA, the design change, the risk file update, the supplier record. If that takes more than a few hours, the inspection will feel long however good the records are, because slow retrieval reads as a system nobody actually uses.
What a mapping looks like
For each signal it is worth working out:
| Column | What it captures |
|---|---|
| Signal | What drew attention in the first place |
| Trigger source | Where that signal surfaced |
| Likely entry point | Which part of the QMS an investigator opens first |
| Risk hypothesis | What they are probably testing |
| Lifecycle path | The connected processes the thread may run through |
| Evidence requested | What they will ask to see along the way |
A common worked example is a software-enabled device with complaint and MDR trends. The thread can move across Measurement and Improvement, Design and Development, Change Control, Outsourcing and Purchasing, and Management Oversight. The preparation question is whether your evidence holds together along that whole path, not just within each box.
Vioela’s note: Follow a thread far enough and it usually ends in a regulatory question, not a quality one. A complaint trend leads to a design change, and the design change leads to: was that change significant enough to need a new submission? A field fix leads to: was that a correction that should have been reported to FDA? Teams prepare the quality records and forget that the investigator’s last question on each thread is whether something should have been filed. Write down the reasoning for each of those decisions at the time you make it, because a justification written during the inspection reads exactly like one.
One important caveat
Once inspections become more risk-based, investigator variation naturally increases.
A seasoned investigator may follow a signal differently from a newer one. Someone trained recently under QMSR may navigate differently from someone who spent years inspecting under QSIT.
So this framework is not a guarantee of inspection sequence. It is meant to illustrate how a risk-based inspection may unfold, and to help teams think about how risk is controlled across the device lifecycle.
Generic and educational only. Not tailored to any specific device or regulatory situation, and not consulting or legal advice.
Adapted from a post first published on LinkedIn.