Summative Evaluation with Post-Market Data?

Hello everyone, and thank you in advance for your ideas and assessments.
I have an in vitro diagnostic device that has been approved under IVDD and has been on the market for several years. The manufacturer would now like to certify it under IVDR.
The IVDR audit by the Notified Body gave the following feedback:

"Usability testing was only conducted with one laboratory manager and one physician. Summative usability testing of the XYZ software in its ready for release version should however be carried out on a representative number of intended users at different study sites (multi center evaluation). Please provide the results of usability testing considering these aspects."

I am now wondering whether it is possible to create or write a corresponding summative evaluation report retrospectively from PMS data?


My argument is: the device has been on the market for several years and has therefore been used by several intended users. If suitable (!) PMS data are available it should be, in my opinion, possible to "perform" "retrospectivaly" a summative evaluation.

Waht do you think?

Thanks,

Martin

I would expect that arguing your way through this is not going to be effective.

IEC 62366 has a provision (Annex C) for addressing a "user interface of unknown provenance" (UOUP). The annex starts out:

This annex was created in recognition of the fact that many MANUFACTURERS will be interested in applying the tools defined in this standard to USER INTERFACES or parts of USER INTERFACES that have already been commercialized prior to the publication of this edition of this standard. Such USER INTERFACES or parts of USER INTERFACES were not developed using the PROCESSES of IEC 62366-1 and as a result are of unknown provenance with respect to these PROCESSES.

It sounds like this might be your best approach. In essence, you establish a Use Specification per the standard, review your post-production data, identify hazards and hazardous situations related to usability, determine if existing risk controls (around use) are adequate or if additional ones are necessary, and do a residual risk evaluation. With all that in hand, you may be able to justify that a summative study is not necessary. (Or you may find that there are cases you hadn't considered and it WOULD be beneficial to do a study!)

Hi yodon,
Thank you for your response. To be honest I wasn't expecting your (precious!) answer and it's quite interesting. In fact, I know Appendix C (in theory ), but I didn't really think of that one.
Thanks, I think your suggestion will work!
Many greetings