Keeping the Usability File up to date

Dear community, I am very grateful for all the helpful advice so far.
I have a question regarding the Usability File, especially the conducted Usability Tests. For a stand-alone medical software, there were once conducted a series of usability tests (which are considered very valid) and this was done a year ago the last time. However, it is very difficult to organise such a large study and I am asking myself the following:
1) We develop our software with features that are more less "self-explanatory" and there are additional extensive user trainings. Is there a need to update the Usability File at all or can you justify not updating the file with the argument that there is enough valid feedback from the user trainings?
2) Could it lead it to problems in an Audit if your Usability File has not been updated since more than a year?
3) Are there any ways to keep the Usability File up to date without organisation-heavy Usability Tests because the intended users are really busy themselves? I came across other techniques such as cognitive walkthrough, thinking aloud which seem less organisation-heavy. Does anyone have experience with that?
4) Is there any fixed "schedule" to update the Usability File or what is the exact rationale for updating it?
5) Is there a way we can use protocols from user trainings (where there are new features) as an input/supplementary material for our Usability Tests?

Thanks everyone!

Auxilium said:

1) We develop our software with features that are more less "self-explanatory" and there are additional extensive user trainings. Is there a need to update the Usability File at all or can you justify not updating the file with the argument that there is enough valid feedback from the user trainings?
2) Could it lead it to problems in an Audit if your Usability File has not been updated since more than a year?
3) Are there any ways to keep the Usability File up to date without organisation-heavy Usability Tests because the intended users are really busy themselves? I came across other techniques such as cognitive walkthrough, thinking aloud which seem less organisation-heavy. Does anyone have experience with that?

Click to expand...

Interestingly, the standard is pretty silent regarding post-market changes. Several mentions of justifying that you do have a UI that promotes safe and effective use using post-market data. Personally, I always include a UI review as part of the periodic risk management review anyway. (So my UE 'file' includes the risk report with this information.)

While not prescribed, I also assess changes for impact on the UI and document rationale why additional studies are not needed (if that's the case). I wouldn't claim that they are "self-explanatory" since that might be considered a biased opinion. Since you're doing user training, you're really addressing the change so that plus the feedback should probably be a good argument against having another usability study.

Auxilium said:

4) Is there any fixed "schedule" to update the Usability File or what is the exact rationale for updating it?

Click to expand...

As far as I know, no, there is no prescribed schedule. As mentioned above, I piggy-back on the periodic risk review. The frequency should be a risk-based call.

Auxilium said:

5) Is there a way we can use protocols from user trainings (where there are new features) as an input/supplementary material for our Usability Tests?

Click to expand...

I would think you could treat these as "mini" summative studies if, in fact, you're getting feedback. And so you could add that to whatever documentation you maintain (update the report?).

I think @yodon suggestion to fold these into (periodic, ad hoc) risk reviews is the only practicable approach for an existing, on-the-market device. Presumably there exists "circumstances of use" which factor into elements of the risk management file and can factor into not just Use FME&A but also P1, P2 in a 'classic' 14971 assessment.

Try to imagine a world where the general populace and manufacturers of medical (electrical) devices were largely ignorant of the hazards present by "electricity". As reports of harms due to this (hypothetically) initially poorly understood/appreciated hazard, we (following 14971) would be obligated to assess our design choices and make an updated assessment of the risk profile for the device(s). If the electrical situation is too fantastical to imagine, instead try to imagine newly recognized hazards from implantable materials, media that allow the growth of infectious agents etc.

I want to make this point explicit: There is no fundamental hazard relating to usability, the (spectrum of) use of medical devices only alter the likelihood of physical hazards leading to recognized harms.

Caveat: It is of course possible for regulatory authorities to mandate certain actions even if such actions are not well-motivated by a risk analysis, so keep in mind that "you cannot fight city hall."