How to Identify and Escalate Medical Device Complaint Trends
Ten complaint files may each look minor when reviewed separately. Together, they may reveal a recurring failure affecting the same product, supplier, market, or use condition.
That is why reviewing medical device complaints one at a time is not enough. Complaint closure does not end the need to evaluate collective data. Manufacturers may need to look across complaints, service records, returns, nonconformances, supplier information, CAPA, and other post-market data to determine whether an emerging pattern requires action.
A trend can be based on an increased frequency, increased complaints, increased severity, repeated failure mode, or the focus on a specific lot or supplier. A high number of complaints does not necessarily indicate poor device performance, while low numbers of complaints do not mean there is no significant problem.
For manufacturers needing support with complaint patterns and downstream quality implications, see Medical Device Complaint Handling and Vigilance Review Services.
Key point: A trend is identified from the pattern in the data, not simply from the number of complaints received.
When Do Repeated Medical Device Complaints Indicate a Trend?
Repeated complaints are a reason to investigate a potential trend, but repetition alone does not establish that a trend exists.
A manufacturer should consider whether the complaints differ meaningfully from expected or historical performance.
Repetition Alone Is Not Always Enough
Potential signals include:
⦿ An increase above the normal or expected level
⦿ Recurrence of the same or related failure mode
⦿ Concentration within one lot or manufacturing period
⦿ Concentration around one supplier or component
⦿ A shift toward more serious outcomes
⦿ Emergence in a new market or user population
⦿ Recurrence after corrective action
⦿ A pattern involving the same use condition
⦿ A change from historical device performance
The appropriate assessment depends on the device, available data, exposure, and applicable requirements. Under EU MDR Article 88, manufacturers must assess statistically significant increases in the frequency or severity of certain non-serious incidents and expected undesirable side-effects using the methodology defined in their PMS plan.
Types of Complaint Trends
Frequency trend:
Similar complaints show increased trends over time.
Rate trend:
Complaints increase relative to an exposure measure such as units distributed, devices in use, procedures, tests or operating hours.
Severity trend:
Complaint volume may remain stable while outcomes become more serious.
Failure-mode trend:
A new or previously uncommon failure begins recurring.
Lot, supplier, geographic or use-condition trend:
Complaints cluster around a particular lot, supplier, market, site, clinical setting or operating condition.
Emerging Signal, Established Trend or Systemic Issue
For internal governance, it can be useful to In terms of internal controls, it could be helpful to differentiate between:
⦿ Emerging signal: Early indication requiring closer review or monitoring.
⦿ Established trend: An early indication calling for additional analysis or monitoring.
⦿ Systemic issue: Evidence of a broader product, process, supplier or quality-system problem.
These are practical internal categories, not universal regulatory classifications.
Repeated medical device complaints indicate a potential trend when their frequency, rate, severity, or shared characteristics differ meaningfully from expected or historical performance.
How Should Similar Medical Device Complaints Be Grouped?
Before assessing a trend, manufacturers need to determine which complaints should be considered together.
If categories are too broad, specific failure patterns may be hidden. If they are too narrow, related complaints may appear unrelated, and trends may be missed.
Useful grouping dimensions include:
⦿ Device family or model
⦿ Failure mode
⦿ Patient or user outcome
⦿ Component or supplier
⦿ Lot or batch
⦿ Software or firmware version
⦿ Market or geography
⦿ User type and clinical use
⦿ Servicing history
⦿ Complaint code
The customers might report the same issue using different terms: “device shut down,” “battery died,” or “screen went blank.” It may be the same mode of failure. Proper coding will help to avoid splitting the same problem into different categories.
Trend analysis should not rely only on confirmed root-cause codes. Some complaints remain inconclusive, and a pattern may need attention before every investigation has identified a definitive cause.
Complaint information should also be compared with service reports, product returns, nonconformances, supplier defects, CAPA, and PMS information. For a deeper look at investigation quality, see How Do You Know Whether a Medical Device Complaint Investigation Is Adequate?.
Is Complaint Count Alone Sufficient?
No, raw complaint count is useful, but it should not normally be the only basis for determining whether complaints represent a meaningful trend.
Complaint count can be misleading when:
⦿ Sales or distribution volume increased
⦿ The installed base increased
⦿ Complaint reporting improved
⦿ Product use changed
⦿ One customer generated multiple complaints
⦿ Complaint coding changed
For example, 15 complaints from 20,000 distributed devices may represent a lower rate than 10 complaints from 10,000 devices. Conversely, two complaints involving failure of a critical alarm may deserve more attention than many minor cosmetic complaints.
A useful review therefore considers count together with rate, severity, recurrence, historical performance and concentration.
Should Complaint Rate Be Considered?
Yes, where a reliable denominator is available.
A basic complaint rate can be expressed as:
Complaint rate = Relevant complaints ÷ Relevant exposure measure.
The appropriate denominator depends on the device. Units distributed may be useful for some products, while procedures, tests performed, installed devices or operating hours may be more meaningful for others.
Rate interpretation can be affected by under-reporting, delayed reporting, incomplete installed-base information, distributor data gaps and inconsistent coding. The rate should therefore be interpreted alongside other evidence.
A manufacturer may still need to act before a meaningful rate can be calculated. Death or serious injury, a new severe failure mode, failure of a critical safety function, ineffective risk controls or clustering around one lot or supplier may justify escalation based on the available evidence.
When Should a Medical Device Complaint Trend Be Escalated?
A potential trend should be escalated when routine monitoring may no longer be sufficient.
Possible triggers include:
⦿ Complaint frequency or rate rises above the historical baseline
⦿ Severity increases
⦿ A new failure mode emerges
⦿ Similar serious complaints recur
⦿ Events cluster around one lot, supplier or device version
⦿ The issue continues after corrective action
⦿ Risk controls appear ineffective
⦿ Complaints reveal a broader design, software or usability issue
⦿ The pattern affects several markets
⦿ Internal alert limits are reached
Escalation does not always mean immediate CAPA. Depending on the evidence, the response may include focused investigation, enhanced monitoring, supplier review, risk-management review, CAPA, engineering action, PMS escalation or regulatory assessment.
This is where complaint trending connects to risk management of medical devices. A recurring pattern may indicate that a hazard, probability estimate, severity assessment, risk control or residual-risk conclusion needs to be reconsidered. ISO 14971 provides the framework for medical-device risk management across the product lifecycle, including consideration of production and post-production information.
CAPA may be used where there is a system or repetitive problem which cannot be solved through complaint corrective action alone. The issue is not just whether there is closure of another complaint but whether the organization has reason to believe that it needs to take some corrective action for a larger problem.
Under the EU MDR, PMS systems must include methods for detecting and reporting trends, and Article 88 establishes specific requirements for certain statistically significant increases. Complaint trends should therefore feed the broader PMS process where applicable.
Practical Example: Recurring Battery Failures
Complaints about battery failure in the monitor measuring blood pressure have been received by a manufacturer.
Month 1: The manufacturer received two reports from different markets. No harm to the patient is reported, and the affected products are older items. Investigations and normal monitoring may suffice.
Month 2: The manufacturer receives four more complaints. Three of them involve batteries from the same vendor, while one happened during critical clinical procedures. It may be time for the company to investigate whether these complaints involve the same failure mode, whether there is any change in exposure, and if severity is increasing.
Month 3: Eight more complaints arrive. The complaint rate exceeds the historical level, failures cluster around one supplier lot, and a previous supplier correction did not prevent recurrence.
At this point, escalation may include supplier corrective action, CAPA, risk-management review, engineering assessment, maintenance review and cross-market regulatory assessment.
The trend is therefore not identified by complaint count alone. It emerges from recurrence, rate, supplier concentration, severity, failed previous action and potential patient consequence.
Common Medical Device Complaint Trending Mistakes
Reviewing Each Complaint Independently
Individual complaint closure can hide aggregate patterns.
Using Broad or Overly Narrow Categories
Broad categorization may mask failure modes, whereas overly narrow categorization may separate related events.
Reviewing Complaint Counts Without Exposure Data
A higher count may simply reflect increased distribution or use.
Ignoring Severity Because Complaint Volume Is Low
A small number of serious events may require more attention than a large number of minor complaints.
Failing to Compare Lot, Supplier, Market, or Device Version
Concentration can reveal important clues about the underlying issue.
Relying Only on Confirmed Root Cause
An emerging pattern may need attention before every investigation has established a definitive cause.
Ignoring Service, Supplier and PMS Information
Complaint records may not contain the complete picture.
Monitoring Without Defined Escalation Criteria
A trend process should establish what happens when a meaningful signal is identified.
Failing to Document the Decision
The manufacturer should document why a pattern was monitored, escalated or determined not to require further action.
What Should Be Documented in a Medical Device Complaint Trend Assessment?
A defensible assessment should identify:
⦿ Complaint group or failure mode
⦿ Products and variants
⦿ Review period
⦿ Complaint count
⦿ Exposure denominator, where applicable
⦿ Complaint rate
⦿ Historical comparison
⦿ Severity
⦿ Lot or supplier distribution
⦿ Related CAPA or service data
⦿ Data limitations
⦿ Trend conclusion
⦿ Escalation decision
⦿ Decision rationale
⦿ Responsible owner
⦿ Follow-up date
⦿ Reviewer and approval
The record should allow another reviewer to understand what was assessed, what the data showed and why the organization chose to monitor or escalate the issue.
The Key Question Is Not How Many Complaints You Received.
Repeated medical device complaints do not automatically establish a trend. A meaningful pattern may become visible through frequency, rate, severity, recurrence or concentration.
The strongest trend reviews combine complaint data with exposure, investigation findings, service information, supplier data, CAPA, risk management and other post-market information.
The key question is not simply, “How many complaints did we receive?” It is:
“What is the collective pattern telling us about device performance, risk and the actions we may need to take?”
For manufacturers needing support connecting complaint trends with CAPA, risk management, PMS and regulatory decisions, see Medical Device Complaint Handling and Vigilance Review Services.
Need Help Reviewing Your Medical Device Complaint Trends? Are recurring complaints giving you enough information to identify emerging risks?
FAQs
When do repeated medical device complaints become a trend?
There is no universal complaint number. A potential trend may emerge when frequency, rate, severity, recurrence or shared characteristics differ meaningfully from expected or historical performance.
Is complaint count alone enough to identify a trend?
No. Count should be considered alongside exposure, rate, severity, recurrence, historical performance and concentration.
When should a complaint trend trigger CAPA?
CAPA may be appropriate when evidence indicates a systemic or recurring issue that cannot be adequately addressed through individual complaint correction or routine monitoring.
Should supplier and service data be included?
Where relevant, yes. Supplier defects, service records, returns, manufacturing information and PMS data may reveal patterns not visible in complaint records alone.
