This is a new requirement so, in theory, you need a new SOP. This is not the same as trend reporting in the current vigilance MedDev or in the GHTF guidance document that it refers to. However, if you do have an SOP that implements the MedDev then you could revise it for the Article 88 requirement.
First, you need to read Annex III, section 1.1(b) fifth indent. The details of how you satisfy Article 88 must be in the PMS plan.
You need to determine the data to analyze. You are analyzing incidents, not complaints and not customer feedback. Determine if the information you receive is a complaint under ISO 13485:2016. If so, then determine if the complaint is an incident using the definition in Article 2(64). If it is an incident, determine if it is a significant incident using the definition in Article 2(65). If it is a significant incident, determine if it is reportable using the criteria in Article 87(1)(a). If yes, report it. If not, add it to the incidents for trend analysis.
I'm going to describe how to do the analysis for frequency of occurrence. The other analyses are similar.
Because an incident includes any "malfunction or deterioration in the characteristics or performance of a device", you should have covered them in the EN ISO 1497:2012 hazard analysis, which means you have established an expected frequency of occurrence. This is the foreseeable point from Article 88.
Select an observation period as required by Annex III. Assume it is twelve months (but it could be anything convenient). Enter the data (month and occurrence rate) for each of the twelve months and ask Excel to create a line graph. The data is a time series. On the graph ask Excel for a linear trend line and to display the equation. The slope of the line should be zero and the intercept should be in the range of frequency of occurrence from the risk matrix.
Now use the Analysis Tool Pack's regression tool. Provide the ranges of the data, click Confidence Level, and select a number (use the default of 95%).
Examine the output. It will give you the slope and the upper and lower confidence values. If zero is between them, then there is no statistically significant change. If zero is not between them there is a statistically significant change. If it is an increase, it is reportable.
------------------------------
Dan O'Leary
Swanzey NH
United States
------------------------------
Original Message:
Sent: 11-Sep-2018 17:35
From: Anonymous Member
Subject: EU MDR Artlicle 88
This message was posted by a user wishing to remain anonymous
Hello Colleagues,
We are working on creating a PMS system for medical devices with emphasis on articles 83-89 from the EU MDR. Instead of creating new SOPs we are reviewing about 200 SOPs that will apply to PMS and comply with EU MDR. Once the SOPs have been identified they want to place them within each Article to make them compliant with EU MDR.
Right now we struggling with the interpretation of Article 88 and which SOPs would apply or do we have create a new one. I think I understand what the article is requesting;trending and analysis of complaints that could or have become reportable events due to the frequency and/or severity increasing or has increased within a period of time. The trending should explain the analysis method used to get to that conclusion. This is my interpretation and I could be totally wrong.
Some of my coworkers are arguing that they are referring to feedback received or gathered that could become complaints. Also they think that we need an SOP explaining analysis methods for this trending. I think that we could just add data analysis details to any reports for PMS, or make a reference to a current SOP that explains the various methods used for data analysis.
Any thoughts?
Any feedback will be appreciated!
MEL