Intended Use vs Actual Use and Scope of Risk Management

[mdreg]

Hi,

The intended use is what we say (labelling, statements in the clinical evaluation) our device is intended to be used for.

There is potentially a difference between the intended use and what the device is actually used for by customers. An illustrative example: A (car) manufacturer could state that they sell benches/devices to sit on. There happens to be a (steering) wheel in front of one bench, and an engine below it, and customers use the device to sit and to drive from A to B. Potentially, the (car) manufacturer even tested the driving feature, however still markets their device as "bench".

I have a few questions regarding this:
Q1: Is the extent of this difference somehow regulated/limited?
Q2: What are the implications of such a difference?
Q3: How does this affect the scope of the risk management?

Regarding Q3 with reference to the car example: Strictly considering the intended use (a device to sit on), probably the worst risk would be a buttock lump due to a collapse of the bench. Once we know that there is also an engine below one bench, we would consider the risk of death/severe injury due to extensive heat or explosion of the engine. However, we would not consider risks associated with driving/speed/traffic. Nevertheless, not documenting risks associated with driving/speed/traffic seems a bit odd because when assessing risks probably those risks would come up.

Thanks

 

[mihzago]

The intended use is not only conveyed through the things you say, but also the things you do or the things the device does.
If you design a spoon, you can't say it's a new type of fork without prongs and tell users it shouldn't be used to eat soup. Similar to building a car and saying it's a bench for sitting (maybe if you remove the wheels and don't provide ignition key).

From the risk management standpoint, you have to consider normal use and foreseeable misuse. Off-label use could be considered a foreseeable misuse and can get tricky. You can't prevent all off-label or foreseeable misuse and many times have to rely on labeling, but if you specifically design something to be used certain way, then you should evaluate all the risks related to that use and potentially consider it a normal use.
Foreseeable misuse also has to be assessed for risks (according to ISO14971 and IEC 62366).

 

[Raisin picker]

Check also MDR Annex XIV 6.1.e

 

[mdreg]

OK, thanks, my gut feeling tells me the same thing. However, I wonder if there are legal means to limit this discrepancy. MDR Annex XIV 6.1.e is actually doing that (thanks), however only after marketing approval of the device. Are there legal means preventing this for the initial market approval?

Also, in reality the discrepancy between intended use and actual use is more subtle than in the car/spoon example.

A related question: Assume a device that can fix two body parts together and has a ruler engraved which could be used to position the body parts relative to each other. The manufacturer could state that the device is intended merely for fixation of body parts, or that the device is intended for fixation and assistance in the relative positioning of the body parts. Is the scope of the risk management different for the different intended uses? Exemplarily, is the risk that the relative positioning of the body parts is wrong in the scope of the risk management for one or both of the intended uses?

 

[indubioush]

A related question: Assume a device that can fix two body parts together and has a ruler engraved which could be used to position the body parts relative to each other. The manufacturer could state that the device is intended merely for fixation of body parts, or that the device is intended for fixation and assistance in the relative positioning of the body parts. Is the scope of the risk management different for the different intended uses? Exemplarily, is the risk that the relative positioning of the body parts is wrong in the scope of the risk management for one or both of the intended uses?

This is confusing. Are you saying that fixation and positioning are two different intended uses? I would seem to me that you can't have fixation without positioning. This appears to be one intended use.

 

[mdreg]

I am saying, that the intended use does not always describe all potential uses of a device.
A device can be used to fix two body parts without assisting in their positioning, i.e., the user positions the body parts, and then uses the fixation device to make sure the body parts stay where they are.
Thus the question remains: In the case where the device is capable of doing more (assist in positioning) than the intended use states (fixation), what is the scope of the risk management?

 

[indubioush]

Are you saying you don't intend the device to be used for positioning? If there is a ruler on it, it would seem to me that it is intended to be used for positioning.

You need to account for reasonably foreseeable misuse in risk management.

 

[Ronen E]

Check also MDR Annex XIV 6.1.e

To me this represents a slippery slope.
Not everything users come up with should become part of the intended use.
I do agree, however, that manufacturers have an ongoing obligation to seek and identify "systematic" off label uses (it's unrealistic to expect them to catch all off label uses regardless of prevalence), and address them in the labeling. Not necessarily by endorsing, it could be via new contraindications. Their obligation is to KNOW and ADDRESS, no to accommodate every whim.
Part of the problem is when users use the device off label knowingly (i.e. against an explicit contraindication or instruction), or simply by neglecting to RTFM.
The former would be Abnormal Use, which sadly is to date NOT properly addressed in regulations and guidance (the situation is as clear as mud). IMO it should not be the manufacturer's responsibility to account for Abnormal Use (after all, by definition Abnormal Use happens after the manufacturer already addressed the issue "reasonably" - but what constitutes "reasonable" is hardly defined in that context); but it seems the MDR (or how it's pursued by NBs) completely ignores that.
In my opinion the latter (not RTFM) should also be considered Abnormal Use (especially for devices with moderate or higher complexity o risk), but no one really asks me.

 

[Ronen E]

I am saying, that the intended use does not always describe all potential uses of a device.

This sounds to me like an attempt to eat the cake while keeping it whole.
Please don't take it personally - I've seen this kind of approach in dozens of companies.
As another example (which I also saw in many places), using the wording "to assist in..." is clearly a regulatory evasive maneuver. It's like saying "Well, Sirs, it's not REALLY intended to do that, not really... it's just assisting... other devices/sources will surely complete the picture/task!... So you can't really expect us to take responsibility when it doesn't work out. Our device was merely assisting, more or less... the thing failed for other reasons outside our control..."

Ideally, the intended use SHOULD cover all potential beneficial uses. Okay, it might not be a realistic expectation, but this is the ideal we need to try and approach. How to go about it practically? Try to anticipate as much as you can, document all that, and account for the foreseeable risks. Then launch and keep a close watch. There will - most likely - be off label uses you didn't anticipate. That's quite unavoidable (especially when the device is complex). Now, your job is to spot the common ones among those off label uses, and rule whether they are acceptable to YOU - the responsible manufacturer. If they are, update all your Technical Documentation (i.e. naturalize the off label usese as part of the new intended use). If not - analyze the risks thoroughly, do everything you reasonably can to prevent such use (preferably by design), and clearly contraindicate / instruct against it in the labeling.