What is post-market surveillance?
Post-market surveillance is the ongoing monitoring of a medical device after it reaches the market, to detect and act on safety problems that pre-market testing could not surface. It pulls together adverse event reports, customer complaints, published literature, and other signals to catch emerging risks before they scale.
The premise is simple: a clinical trial cannot reveal every problem. Rare complications, long-term failures, off-label use, and manufacturing drift only appear at real-world volume over real-world time. Surveillance is the system that watches for them.
Why it matters
Most consequential device safety issues are not visible at approval. A lead that fails after five years, a pump whose controller software misbehaves under a rare condition, a valve that migrates in a specific anatomy: none of these show up reliably in a trial of a few hundred patients followed for months. They show up in the field, in reports that trickle in one at a time. Surveillance is the difference between catching that pattern early and reading about it in a recall notice.
The U.S. regulatory framework
In the United States, post-market surveillance rests on a few pillars:
- Medical Device Reporting (21 CFR Part 803). Manufacturers and importers must report deaths, serious injuries, and certain malfunctions. User facilities report deaths and serious injuries.
- MAUDE. The public database where those reports become searchable. See our guide to the MAUDE database.
- Section 522 postmarket surveillance studies. The FDA can order a manufacturer to actively study a specific device after clearance or approval.
- Recalls and safety communications. The corrective actions that follow once a problem is confirmed.
The data sources teams rely on
Surveillance is only as good as its inputs. The common sources, roughly in order of how public they are:
- MAUDE adverse event reports. The largest public source, and the one Claripulse is built around.
- Internal complaint files. A manufacturer's own complaint handling, which precedes most MDR filings.
- Published literature. Case reports and studies that surface failure modes before they reach regulatory channels.
- Registries and claims data. Where they exist, these can supply the denominator MAUDE lacks.
No single source is complete. The art of surveillance is triangulating across them.
The core challenge: from collection to detection
Collecting events is the easy half. Many teams log complaints, file reports, and review trends quarterly, and assume that constitutes surveillance. It does not. Accumulating events only catches problems that are already obvious. The hard half is detection: separating a genuine safety signal from the noise of rising sales, reporting bias, and duplicate filings.
This is where most surveillance quietly fails, because raw counts have no denominator and the real signal often lives in unstructured narrative text. We unpack the methods that address this in our guide to safety signal detection.
Manual vs automated surveillance
A manual workflow checks a handful of devices when someone has time, eyeballs report volume, and hopes a pattern stands out. It does not scale past a few products, and it has no baseline for what normal looks like.
An automated approach monitors every device on a fixed schedule, compares current volume against each device's own history and its peers, and surfaces ranked signals for review. The goal is not to replace judgment but to point it at the right reports. See how the two compare in practice in Claripulse vs. manual MAUDE search, or browse live adverse event data in the free MAUDE lookup.