When it comes to vigilance, trend reporting is essential to determine if non-serious incidents can lead to a much bigger issue over time. These non-serious incidents could be device deficiencies, packaging issues, or an increase in the number of known side effects. Trend reporting allows for informed decisions to be made before non-serious incidents get out of control. If a manufacturer detects an increase in reports related to expected side effects that might lead to death or a serious deterioration of patient health, this might signal a deterioration in the performance of the device or an off-label use of the device. While many manufacturers may not be in the habit submitting a trend report to the National Competent Authority (NCA) for non-serious incidents, it is required by Article 88 of the Medical Device Regulation (MDR) 2017/745.
A trend report must be submitted to the National Competent Authority (NCA) where the manufacturer or its authorized representative has his registered place of business. The aforementioned has been the expectation from medical device manufacturers per the MEDDEV 2.12-1 Rev. 8: Guidelines on a medical devices vigilance system, January 2013 with the motive to detect device misuse or off label use, or an unexpected device performance due to unvalidated manufacturing or design changes. This motive is suggested by the GHTF SG2 N54R8 - 2006: Post Market Surveillance: Global Guidance for Adverse Event Reporting for Medical Devices, that a significant increase in the rate of events exempt from reporting may indicate an underlying change in the performance of the manufacturer's product or in its use by clinicians, patients or other customers.
With the same motive, Article 88: Trend Reporting, found in the MDR, Chapter VII: Post-market surveillance, vigilance, and market surveillance, Section 2: Vigilance, transforms the expectation from MEDDEV 2.12-1 into regulatory requirements. Taking into account the requirement of Post Market Surveillance (PMS), manufacturers do not only need to perform trend reporting per Article 88, but they must also include in the PMS Plan the methods and protocols to manage the events subject to trend reporting. This includes the methods and protocols to be used to establish any statistically significant increase in the frequency or severity of incidents as well as the observation period (per Annex III, section 1.1 b: Technical Documentation on Post-Market Surveillance).
Manufacturers who have followed the MEDDEV 2.12-1 for trend reporting will likely find Article 88 easier to implement than manufacturers who never followed this part of the MEDDEV. In this blog, Rania Gerges, a Qserve Consultant, will provide guidance on how to address Article 88 requirements by interpreting parts of this Article (Part A of this blog). Then, Rania will provide practical solutions on how to incorporate Article 88 aspects into the PMS Plan (Part B of this blog).
- Interpretation of Article 88 requirements:
Article 88 states “Manufacturers shall report, by means of the electronic system referred to in Article 92, any statistically significant increase in the frequency or severity of incidents that are not serious incidents or that are expected undesirable side-effects that could have a significant impact on the benefit-risk analysis referred to in Sections 1 and 5 of Annex I and which have led or may lead to risks to the health or safety of patients, users or other persons that are unacceptable when weighed against the intended benefits. The significant increase shall be established in comparison to the foreseeable frequency or severity of such incidents in respect of the device, or category or group of devices, in question during a specific period as specified in the technical documentation and product information.”
First, let’s examine what the “non-serious incidents” and “expected undesirable side-effects” that are subject to trend reporting. Non-serious incidents are normally received complaints that are not reported as they do not meet the definition of “serious incident”. These types of events should normally be predicted, estimated and documented during the design stages and captured as part of the device risk management documentation, in particular, in a Design, Use, or Application FMEA (dFMEA, uFMEA, aFMEA) and evaluated, mitigated and documented in the individual and the overall risk-benefit analysis. The same is true for the expected undesirable side effects. These events are part of the device labeling and are usually documented as “common side effects or complications”, if applicable. Similar to the non-serious incidents, the expected undesirable side effect incidents should also be documented in the risk management documentation with pre-defined occurrence and severity rates based on literature data, historical data from similar devices, etc. These are the typical events subject to trend reporting, but one should note that events constituting non-serious incidents or expected undesirable side-effects are not often reported back to manufacturers as complaints, therefore the baselines of these events will likely not be reflective of the actual occurrence of these events in the clinical field, unless a more thorough review of historical data and literature, testing, or Post Market Follow-up (PMCF) activities are conducted.
Secondly, article 88 states that if these events occur with a “statistically significant increase in the frequency or severity”, they will have to be evaluated against the benefit-risk analysis to determine if reporting is required. Before defining what a statistically significant increase in the frequency means, it is important to note that the use of any statistical techniques must be appropriate and justified. No statistically-based conclusion should be drawn using a very small number of samples, unless appropriate statistical tools such as non-parametric statistical techniques, have been used. If a manufacturer is dealing with a low overall complaint rate with only 2-5 events of interest in an observation period, evaluation on an individual basis should be undertaken, with a justification provided for not using statistical methods due to a small sample size. On the other hand, if the incidents a manufacturer is trending are frequent in nature, then the use of statistical techniques is appropriate and required to drive decisions. There are many statistical methods that can be explored, but whatever statistical method is used for detecting a significant increase in frequency, it should be combined with root cause analysis for the increase, and a report to regulatory authorities if the trend has an impact on the benefit-risk analysis. The One Sample t Test can be used to determine whether the sample mean of an event frequency is statistically different from a known or hypothesized frequency mean. The One Sample t Test is a parametric test; therefore, it should not be used unless the sample is confirmed to follow a normal distribution. In addition to the use of hypothesis testing, several types of control charts can be used to plot the occurrence rate over an observation period and detect a “trend” as defined by Shewhart control chart rules.
Thirdly, the article also states that if these events occur with a “statistically significant increase in the frequency or severity”, they will have to be evaluated against the benefit-risk analysis to determine if reporting is required. An event might have a harm associated it with it that is estimated to have a mild severity ( for example: severity of 2 on a scale 1-5) if it were to occur, but when the event occurred, it caused a harm with a higher severity than what was estimated in the risk management documents. Use of statistics might not be appropriate for detecting a statistically significant increase in severity since a number increase on a scale would not constitute any statistical significance. Therefore, the event occurring with more severe harm must be evaluated on individual basis and a decision should be made about reporting if the event has a significant impact on the risk benefit analysis, i.e., making the individual and overall risk become unacceptable.
It is worth mentioning that while assessing the increase in the events frequency and severity, it is implied that the risk management system must be based on quantitative categorization of probability of occurrence of harm and severity of harm versus the option to use either qualitative or quantitative categorization as required by ISO 14971: Medical devices - Application of risk management to medical devices.
Finally, the article requires that the significant increase shall be established in comparison to the foreseeable frequency or severity of such incidents in respect of the device, or category or group of devices, ….”. This statement implies that the manufacturer must ensure that the risk management documentation, a living documentation, to be as complete as possible reflecting all potential non-serious incidents and expected undesirable side effects with precise and defendable severity and occurrence rates estimation. In cases where new events not previously identified, they must be evaluated individually and integrated into the risk management documentation including the impact of the risk benefit analysis. Therefore, Article 88 is seen as one of the articles in the MDR that will challenge and validate the adequacy of a device risk management documentation.
- Article 88 and the PMS Plan linkage
How is trend reporting linked to the PMS plan? Annex III, section 1.1 b: Technical Documentation on Post-Market Surveillance requires that “the post-market surveillance plan shall cover at least: methods and protocols to manage the events subject to the trend report as provided for in Article 88, including the methods and protocols to be used to establish any statistically significant increase in the frequency or severity of incidents as well as the observation period;”
Appendix C of GHTF SG2 document N54: Global Guidance for Adverse Event Reporting for Medical Devices provides useful guidance on the trending procedure. The reader is encouraged to review this guidance document and take notes from Appendix C. Below is a summary of the key aspects including the trending parameters a manufacturer must define or refer to them the PMS Plan:
- Event subject to trend reporting: Define the nature of the non-serious incidents or expected undesirable side-effects that will be monitored as part of the PMS plan. Some manufacturers have what is called a “Hazard List” where they keep this list up-to-date as knowledge about the device progresses, and also real-world experience of using of the device provides more realistic occurrence and severity rates of incidents. If such list exits, it can be used or referenced in the PMS plan to define what incidents are subject to trend reporting. If such list does not exist or has not been complied by the manufacturer, the manufacturer is expected to list the particular events that raise a concern with the device safety and performance in the PMS plan itself. The manufacturer should also consider referring to the complaint handling and data analysis procedures where complaint investigation and determination of reportability criteria are defined.
- Baseline: For establishing a realistic (e.g. to avoid under-reporting) baseline to start with for trending these events, several sources of information should be considered such as risk analysis, reliability testing, historical data or scientific publications.
- Observation period: the observation period will vary depending on the volume of sales of the device on the market. It is suggested that for higher volume products a typical observation period of 1 month be used. It is also recommended that the observation period is short enough to facilitate timely corrective action, especially in case of high-risk products.
- Threshold value for reporting: The threshold value needs to be defined in a way to avoid over or under-reporting. This can also be pre-defined by a company policy with proper justification, such as a specified % increase of these events or severity rating increase to a specific harm level, as part of the risk management acceptability criteria.
- Trending statistical techniques and rationale: The trending technique selected must be indicated and justified for its appropriateness, and if use of statistical technique is considered to be inappropriate, a justification must be provided as well.
Most, if not all aspects discussed above are sourced from the device labeling, risk management documentation, outputs from the data analysis procedure (ISO 13485: 2016, Clause 8.4: Analysis of data), statistical techniques procedure (21 CFR 820.250 Statistical techniques), company policies and procedures. Therefore, it is recommended that these sources are referenced in the PMS Plan rather than including all the details in it. This approach should reduce the need to frequently revise the PMS plan if any of the trending parameters or type of events were to change.
We hope with this blog answers some of your questions on Article 88. If you still have questions, please feel free to contact Qserve Group.