What Level 1 Post-Market Surveillance Maturity Looks Like in Medical Device Organizations

Your notified body asks for your PMS plan for the Class IIb device you've had on market for four years. What you hand them is a flowchart showing how complaints are received and processed. They ask about proactive data collection methods, literature monitoring protocols, PMCF endpoints, and trend reporting triggers. You have none of these. The PMS plan is complaint handling by another name.

This is not a hypothetical. It is the single most common finding in PMS audits under the EU MDR, and it reveals something deeper than a documentation gap. It reveals that the organization has never built a post-market surveillance function. What it has is a complaint handling system — possibly a good one — that someone relabeled as PMS when the regulatory requirement appeared. The relabeling changed nothing operationally. Complaints still come in through the same channels. They are still processed by the same people. The only difference is that a document titled "PMS Plan" now sits in the quality management system, and it describes exactly what the complaint handling procedure already described.

The problem is that EU MDR Articles 83 through 86 define PMS as something fundamentally different from complaint handling. PMS must be proactive, not reactive. It must encompass data sources beyond complaints — clinical literature, registry data, competitor events, user feedback, and planned clinical data collection through PMCF. It must produce analytical outputs — PSURs that compare actual device performance against expected performance and update benefit-risk determinations. None of this happens at Level 1 because the organizational capability to do it was never built.

What the Auditor Sees

The absence of product-specific PMS plans is the first indicator, but it is not the only one. At Level 1, the entire post-market data ecosystem is fragmented and passive. Complaints live in the QMS. Field service records live in a separate system. Sales data — needed to calculate complaint rates per units distributed — lives in a third system that no one in quality has access to. Literature monitoring, if it happens at all, consists of someone occasionally searching PubMed without documented search terms, without defined evaluation criteria, and without any record of what was found or what it meant.

PSURs at Level 1 are either nonexistent or empty of analysis. The organization may produce a document labeled PSUR that lists total complaints received, categorizes them by type, and notes any field safety corrective actions taken during the period. What the document lacks is the analytical core that gives a PSUR its regulatory purpose: comparison of observed adverse event rates against expected rates from the clinical evaluation, assessment of whether post-market data confirms or challenges the assumptions in the risk management file, and an explicit updated determination that the benefit-risk ratio remains acceptable. Without these elements, the PSUR is a data summary. It tells the reader what happened but not what it means.

PMCF is where Level 1 exposure is most acute. EU MDR Article 61 and Annex XIV Part B require proactive clinical data collection for most device classes. At Level 1, PMCF exists only as a sentence in the clinical evaluation report acknowledging that it should be done. There is no PMCF plan, no defined clinical questions, no data collection methodology, and no PMCF evaluation report. The gap between what PMCF requires and what Level 1 organizations do is often the single largest compliance deficit in their quality system.

Why Signal Detection Doesn't Exist

Signal detection requires three things that Level 1 organizations lack: normalized data, statistical methodology, and defined thresholds. Without complaint rates calculated against units in the field, raw complaint counts are meaningless for trend analysis — an increase in complaints might reflect increased sales rather than decreased safety. Without a statistical approach to distinguishing signal from noise, trending becomes subjective. Without predefined thresholds for escalation, even genuine signals depend on someone noticing a pattern during a periodic management review and deciding it looks concerning.

EU MDR Article 88 requires manufacturers to report any statistically significant increase in the frequency or severity of incidents. An organization without trending methodology cannot determine what constitutes a statistically significant increase. This is not a theoretical regulatory risk. It is a patient safety risk. Signals that systematic surveillance would detect early instead accumulate until they manifest as clusters of serious adverse events — at which point the response is reactive, expensive, and late.

The Path Forward

Advancing from Level 1 does not require a transformation program. It requires five specific actions, executed in sequence. First, create product-specific PMS plans starting with your highest-risk devices, using EU MDR Annex III Section 1.1 as the structural template even for non-EU products. Second, establish a literature monitoring protocol with defined search terms, databases, frequency, and evaluation criteria — this is the lowest-effort, highest-impact PMS data source beyond complaints. Third, produce your first analytical PSUR for your most established product, forcing yourself through the exercise of comparing observed performance against expected performance. Fourth, assess your PMCF obligations product by product, documenting which clinical questions require post-market answers. Fifth, integrate complaint and distribution data so that complaint rates per units in the field can be calculated, enabling meaningful trending.

None of these actions is technically complex. What they require is the organizational recognition that PMS is a distinct discipline with distinct capabilities — not complaint handling with a different title.

Take the Post-Market Surveillance Maturity Assessment to identify exactly where your PMS capability stands and build a structured improvement roadmap.