Usability testing required for FDA IDE (investigational device exemption)?

[noruskt]

What is the general thinking about the requirement of usability testing prior to the submission of an FDA IDE? Is FDA requiring it? It seems to me that we should be able to justify using pre-clinical data that the clinical trial will be used to support usability, as long as usability is evaluated during the trial. Right?

 

[yodon]

I don't think I've seen any specific requirements imposed. You may need to provide assurance that potential use / usability errors won't lead to harm. Consider doing a task analysis or PFMEA and consider missteps / errors and results. Maybe training is warranted or oversight?

 

[Tidge]

I'm sidestepping the direct question about any requirements for usability testing prior to IDE.

It seems to me that we should be able to justify using pre-clinical data that the clinical trial will be used to support usability, as long as usability is evaluated during the trial. Right?

As far as I know, this is still applicable : Applying Human Factors and Usability Engineering to Medical Devices

Relying on the IDE is probably not the same as considering 'simulated use', but the underlying concerns are roughly the same. Without knowing any details about the device, the IDE or specific usability concerns I will write this: I have a general expectation that the IDE will be able to inform elements of the usability assessment of the final product. A priori I don't expect the IDE work to completely replace the efforts during validation.

 

[indubioush]

Usability testing should be performed early in device development. Formative evaluations can result in design changes to reduce use error. Much easier to change the design prior to the validation stage. Regulating bodies have been a bit lenient when it comes to usability, but that may change.

 

[powersthatbe]

Responding here since I have the same question but I'm not totally sure it was answered - I am in a similar situation where I am trying to figure out the requirements prior to IDE. I certainly intend to do risk analysis, defining users and use environment and perhaps some preliminary formative work, but I am not sure what (if anything) we would need to submit or have ready on file before an IDE submission. But in reading this:

You may need to provide assurance that potential use / usability errors won't lead to harm.

I'm curious how you would do that without a full summative study, which in my experience is not done before early clinical work (we won't have a truly final design from a commercial perspective during the first trial).

Any thoughts? Not looking to wiggle out of HF work but I just don't see anything to go from in the HF or IDE guidances.

 

[yodon]

I'm curious how you would do that without a full summative study

Your risk analysis needs to provide guidance here.

Also, for electro-mechanical systems, we generally have some degree of 60601-1 testing done prior to the study to give some assurance of basic safety.

 

[powersthatbe]

Also, for electro-mechanical systems, we generally have some degree of 60601-1 testing done prior to the study to give some assurance of basic safety.

This has definitely come up in my conversations as well, and questions like "can we list compliance to 62366" but that would require a full summative study on file, no?

The risk analysis is in progress and will definitely be helpful as well, but I can already assure there will be some critical tasks. I'm hoping to evaluate them in a formative soon but with whatever issues come up I'm not sure to what degree we need to resolve them before the IDE.

 

[Tidge]

... and questions like "can we list compliance to 62366" ...

I wouldn't say that a product is 'compliant with 62366 (or 62366-X)' any more than I would say a piece of ME software is 'compliant with 62304'. You can self-advertise that you applied the principles of 62366, but 'compliance' is going to demonstrated by the applicable standard (for ME devices, it is 60601-1).

I don't want to speak for @yodon but I think the point he is making immediately above is that for a somewhat new device there are well-known hazards to which you wouldn't want to expose patients/users before even planning any sort of clinical study. Things like shocks, sharps, burns, pinches were included in the scope of the second edition of 60601-1 and were recognized as 'needing to be controlled/eliminated' independent of a 'modern' 60601-1 compliance assessment that includes risk analysis of a ME device taking Essential Performance (including usability) into account.

 

[yodon]

I don't want to speak for @yodon

@Tidge - I've seen enough of your posts where I'm comfortable having you speaak for me Nicely stated, too!