The launch date has passed. Devices are in use. Sales has moved on to the next quarter. Then the clinical or regulatory team asks a question that sounds simple and is not:
How are we following these patients over time?
The first answer is often a mixture of spreadsheets, site emails, warranty records, occasional surveys and a plan to call people later. That may be enough to discover that follow-up is difficult. It is rarely a durable evidence workflow.
In brief: post-market clinical follow-up, or PMCF, is a planned process for proactively collecting and evaluating clinical data after a CE-marked device is on the market. Patient-reported outcomes, device-use information, registries and structured follow-up studies can contribute, but the method must begin with the evidence question, not the software.
Smaller manufacturers feel this problem sharply. They still need credible follow-up, but they may not have a global registry team, a large clinical operations department or the budget to implement an enterprise platform for every product.
What is PMCF, and what is it not?
Under the EU Medical Device Regulation, PMCF is described as a continuous process that updates the clinical evaluation. Manufacturers proactively collect and evaluate clinical data from the use of a CE-marked device within its intended purpose to confirm safety and performance over the expected lifetime, monitor known effects and contraindications, identify emerging risks and assess whether the benefit-risk balance remains acceptable. Read Annex XIV of Regulation (EU) 2017/745.
PMCF sits inside a larger post-market system. It should connect to clinical evaluation, risk management, post-market surveillance and the manufacturer's quality processes. It does not replace them.
| Activity | Primary purpose | Where patient follow-up may contribute |
|---|---|---|
| PMCF | Proactively generate and evaluate clinical data after market launch | PROs, functional outcomes, device experience, long-term follow-up and targeted studies |
| Post-market surveillance | Systematically gather and review post-market information about quality, performance and safety | Aggregated trends, feedback and clinical evidence can inform the broader PMS process |
| Vigilance | Assess and report serious incidents and field safety corrective actions under applicable rules | A follow-up response may identify information that must enter the manufacturer's vigilance process |
| Complaint handling | Receive, assess, investigate and document product complaints | A patient or site may report a product issue during follow-up, but the complaint workflow remains distinct |
| Customer feedback | Learn about satisfaction, service and product experience | Useful for product improvement, but not automatically clinical evidence or PMCF |
Important boundary: a PMCF questionnaire should not become a place where possible incidents or complaints disappear. The protocol needs a defined route for responses that may require safety, complaint or vigilance review.
Start with the evidence gap
Buying "PMCF software" before defining the evidence question usually creates a polished form with no clear reason for collecting the answers.
The PMCF plan should identify the clinical questions that remain after launch. Depending on the device, those questions may include:
- Does performance remain consistent over the expected device lifetime?
- Do outcomes in routine use match the premarket evidence?
- How do patients function months or years after treatment?
- Are known side effects occurring at the expected rate or severity?
- Are there new patterns of use, misuse or off-label use?
- Do particular patient groups experience different outcomes?
- How durable is the benefit?
- What does the patient experience reveal that routine site records miss?
The European Commission publishes the MDCG 2020-7 PMCF plan template and MDCG 2020-8 PMCF evaluation report template as practical guidance for structuring the plan and reporting the findings. View the current MDCG guidance library.
Those templates do not dictate one universal data-collection method. They reinforce the need to define objectives, methods, rationale, timelines, analysis and links to the clinical evaluation.
How do patient-reported outcomes fit into PMCF?
Patient-reported outcomes, or PROs, describe how a person feels or functions directly from the patient's perspective. They can capture pain, fatigue, mobility, treatment burden, quality of life, confidence using a device and other experiences that may not be visible in a device log or clinic record.
FDA's Center for Devices and Radiological Health distinguishes PROs from clinician-, observer- and performance-reported outcomes and notes that high-quality clinical outcome assessment data can contribute to device benefit-risk evaluation and labeling. Read the FDA overview of clinical outcome assessments for devices.
A post-market program might combine several types of information:
| Data source | Example | What it may help answer |
|---|---|---|
| Patient-reported outcome | Pain, function, quality of life or symptom impact | Whether the clinical benefit remains meaningful to patients over time |
| Patient experience or usability report | Ease of use, comfort, setup difficulty or treatment burden | Whether routine use reveals problems not apparent in controlled settings |
| Device-use information | Frequency, duration, interruption or reason for non-use | How the device is used outside the clinic and whether exposure affects outcomes |
| Clinician or site report | Examination findings, reintervention or clinical assessment | Objective performance, complications and clinical context |
| Device or connected measurement | Telemetry, status, wear time or measured output | Technical performance and use patterns, subject to integration and data-quality controls |
| Administrative or registry linkage | Procedure, revision, claims or mortality data | Long-term events that may be difficult to capture through questionnaires alone |
Not every PMCF program needs every source. Collecting less data well is often more useful than collecting broad data that nobody is assigned to review.
Why does long-term PMCF follow-up fail?
The difficulty is rarely writing the first questionnaire. It is still reaching the right person at month 12, year three or year five.
Common failure points include:
- contact details that are never refreshed;
- follow-up schedules stored separately from participant records;
- forms that are too long for the question being asked;
- messages that do not clearly identify the study or manufacturer;
- no distinction between a missed response and a normal outcome;
- site staff manually maintaining reminder lists;
- patient responses arriving in one system and clinical data in another;
- no documented workflow for concerning free text;
- instrument licensing or translation questions discovered after launch; and
- a vendor contract ending before the device follow-up period does.
A workable program treats retention as part of the design. It explains why follow-up matters, keeps each interaction proportionate, provides more than one appropriate contact method and makes it easy for participants to get help.
Registry, ePRO, app or full clinical platform?
| Approach | Often a reasonable fit when | Main limitation |
|---|---|---|
| Forms and spreadsheets | The cohort is very small, the follow-up is short and the evidence question is narrow | Scheduling, retention, record history, reconciliation and long-term continuity remain manual |
| Focused ePRO or survey layer | The main need is recurring patient questionnaires and completion monitoring | May not manage site data, device events, consent, complaints or the complete registry |
| Patient registry platform | The program combines patient, site and longitudinal clinical data over a long period | Implementation can become broader than a small manufacturer needs for one evidence gap |
| White-label companion app | The manufacturer also needs onboarding, device-use support, reminders and a branded patient relationship | The PMCF protocol, safety routing and regulatory documentation still need separate definition |
| Full EDC/eCOA clinical platform | The post-market study has multiple sites, formal monitoring, complex data capture and extensive controls | Cost and implementation scope may be disproportionate for a targeted follow-up project |
The right answer can also be a combination. A participant-facing ePRO can feed an existing EDC or registry. A branded app can collect routine device experience while a separate safety process handles complaints and incidents.
How should a PMCF patient-reported outcomes workflow work?
A simple workflow can still be rigorous when each step has an owner.
1. Define the cohort and follow-up schedule
Specify who is being followed, how they enter the program, which device or version they used, the expected follow-up points and what happens when contact information changes.
2. Collect only what answers the plan
Map each questionnaire, field and device measure to an objective. Confirm instrument permissions, translations, recall periods and scoring before the first participant is invited.
3. Make completion visible
The team should be able to distinguish due, completed, missed, withdrawn and unreachable participants. Those states affect interpretation and follow-up.
4. Route responses that need another process
Define which entries may indicate a complaint, possible incident, new contraindication or safety concern. The software can route information; the manufacturer remains responsible for the appropriate assessment.
5. Reconcile patient, site and device data
Participant identifiers, device identifiers, follow-up dates and site data need a consistent relationship. A dashboard that cannot produce a usable export is not the end of the evidence workflow.
6. Review the findings against the clinical evaluation
PMCF is not complete when the last survey closes. Results need to be evaluated, documented and fed into the relevant clinical evaluation, benefit-risk, risk-management and post-market documents.
Do not make patients guess what is being monitored
A person receiving a post-market questionnaire may reasonably assume that a clinician, manufacturer or research team is watching the answers in real time. That may not be true.
The invitation and app should clearly state:
- which organization is conducting the follow-up;
- why the information is being collected;
- whether participation is voluntary;
- who can access the responses;
- how quickly entries are reviewed, if they are reviewed individually;
- how to report a product problem or seek urgent care;
- how long follow-up is expected to continue; and
- how to update contact information or withdraw where applicable.
For patients who find this article: PMCF is a way for a manufacturer or research team to keep learning about a device after it is in use. A questionnaire should explain who is collecting the information and what to do if you have an urgent medical concern or a problem with the device. Do not assume a survey response is continuously monitored.
AI can reduce review work, but it cannot own the conclusion
Long-term follow-up creates repetitive work. AI may help categorize free text, identify incomplete records, summarize a participant timeline or prepare a list of responses for human review.
It should not quietly decide that a possible complaint is harmless, alter what a patient reported, fill missing outcomes or determine whether a new risk is acceptable.
The useful question is not "Can AI automate PMCF?" It is "Which bounded task can be supported without weakening traceability, clinical judgment or the manufacturer's quality responsibilities?"
How can CareClinic support PMCF patient-reported outcomes?
CareClinic provides a participant-facing foundation for longitudinal tracking: symptoms, medications, measurements, mood, sleep, activity, journals, assessments, reminders, care plans and reports.
For a smaller medical-device manufacturer, that may support a focused post-market questionnaire program, device-use diary, patient-reported outcome workflow or branded follow-up experience.
CareClinic should not be presented as a complete PMCF, vigilance or complaint management system by default. Study-specific controls, participant enrollment, assessment schedules, locked submissions, audit history, long-term retention, device identifiers, EDC or registry integration and regulatory documentation need to be reviewed against the intended program.
The most credible first step is a protocol-fit discussion: what evidence gap needs to be addressed, which participants are followed, what they report, how often they are contacted, where the data goes and which responses must enter another quality or safety workflow.
For the broader patient product decision, see The Device Is Ready. The Patient App Is Not.
Related CareClinic guides
- Electronic Patient Diary Software Beyond Traditional ePRO
- The Trial Is Remote. The Burden Isn't.
- A Reminder Is Not a Patient Support Program
Bring the PMCF question, not just the questionnaire
Tell the CareClinic team what device is being followed, which evidence gap the program addresses, the expected cohort, follow-up duration, assessments, reminder schedule, review process and required exports or integrations.
Discuss a PMCF or Post-Market Follow-Up
Choose Patient Reported Outcomes (ePRO) when the primary need is structured post-market questionnaires or a patient diary. Choose OEM or White Label when the manufacturer also needs a branded companion experience. Do not include patient information in the inquiry.
Frequently asked questions
What is PMCF for a medical device?
Post-market clinical follow-up is a planned, continuous process for proactively collecting and evaluating clinical data after a CE-marked device is on the market. It updates the clinical evaluation and helps confirm ongoing safety and performance, monitor known effects and identify emerging risks.
Is PMCF the same as post-market surveillance?
No. PMCF contributes clinical data to the broader post-market surveillance system. PMS also includes other information and processes concerning device quality, performance and safety.
Can patient-reported outcomes be used in PMCF?
Yes. PROs may help evaluate symptoms, function, quality of life, treatment burden and other outcomes meaningful to patients. The selected measures and schedule should be appropriate to the PMCF objectives and intended analysis.
Does a customer satisfaction survey count as PMCF?
Not automatically. Customer feedback may be useful, but PMCF requires a documented clinical objective, method, data evaluation and connection to the manufacturer's clinical evaluation and post-market processes.
Can a PMCF questionnaire collect complaints or adverse events?
A response may contain information that belongs in a complaint, safety or vigilance process. The workflow should identify and route that information rather than treating the questionnaire as a replacement for those systems.
What is the difference between a PMCF registry and ePRO?
A registry may combine participant, site, clinical, device and administrative data over time. ePRO focuses on outcomes reported directly by participants and can operate as one data-collection layer within a registry or post-market study.
Can CareClinic be used for post-market patient follow-up?
CareClinic can be evaluated as a participant-facing layer for questionnaires, diaries, reminders and longitudinal patient-reported information. The required study controls, integrations and documentation must be scoped for the intended PMCF or post-market program.
Can AI review PMCF responses?
AI may support bounded tasks such as organizing free text or identifying incomplete records. Decisions about complaints, incidents, emerging risks, benefit-risk acceptability and clinical conclusions require appropriate documented human oversight.
How long should PMCF follow-up last?
There is no single duration for every device. The period should follow the device's expected lifetime, residual evidence gaps, risk profile, clinical objectives and applicable regulatory plan.
Educational information only. This article is not regulatory, legal, clinical, quality-system, statistical or medical-device advice. Manufacturers and research organizations should assess their device, intended use, classification, evidence gaps, jurisdictions and applicable requirements with the appropriate professional advisers.
