Unacceptable Risks associated with Essential Function

[d_addams]

not sure if this would have been better to ask in the 60601 section, but does one consider the risks associated with the loss of Essential Performance unacceptable because a key benefit has been lost or because of the potential harm(s) induced by the loss of performance?

I assumed it is because of the loss of benefit because there isn't always a harm when a product fails, particularly for some interventional products. With implants the failure may not cause direct harm when the patient 'simply' reverts to their baseline health status. They would need surgery to remediate the failure (which re-exposes them to risks of surgery). Also while not causing harm, having a non-functioning implant doesn't mean there aren't potential harms in the future. In either case there is reduced or no benefit, thus the risk-benefit is thrown off and therefore it is to be recognized as unacceptable.

Is this the rationale behind defining loss of Essential Performance as an unacceptable risk or is there more to it?

 

[Tidge]

Because the shift from 60601-1 from 2nd to 3rd editions(*1), I prefer to think of Essential Performance as distinct from Safety (= Freedom from Unacceptable Risks)... but they are not unrelated. The manufacturer gets to specify the (Essential Performance) limits wherein the device does not expose people to unacceptable risks. The daylight between EP and Safety (as I see it) is that no one should have the option to purchase/use unsafe devices (basic safety) but that the market/clinical practice should know the EP limits of the devices intended to be bought. Simplistic perhaps, but this is how I think of them.

(*1) Perhaps apocryphal, but I was told (by NRTL instructors) that it was possible for ME devices to be certified as safe under the 60601-1 2nd edition.. as long as the devices passed the 2nd edition safety tests... even if those devices stopped working. The example given was the ventilator that stopped breathing when knocked over, but didn't catch on fire or shock people.

 

[d_addams]

so there can be items noted as Essential Performance that don't create unacceptable risks? But isn't the definition of essential performance those items that if they fail, create an unacceptable risk? I don't see how you can call something Essential Performance without it resulting in unacceptable risk.

In the ventilator example if the ventilator stops ventilating the patient, its an unacceptable risk since the patient is no longer receiving any benefit. Therefore the risks associated with being intubated are no longer acceptable. I had not considered there does needed to be a specific potential harm caused by the failure (i.e. the fire/shock). But perhaps I'm applying how one thinks about 14971 benefit/risk to better interpret the language of 60601.

It seems the NRTL instructors felt harms that arose from the failures were what created Essential Performance. This sounds like needing to specify performance in fault conditions (if the device tips over, it shall not expose the user to hazardous line voltage).

 

[Tidge]

so there can be items noted as Essential Performance that don't create unacceptable risks?

What I think is different about Essential Performance is that the EP is what is supposed to be providing the Benefit on the side of the Benefit-Risk Analysis. Not meeting Essential Performance is IMO nullifying the benefit rather than introducing risk.

There can subtleties, but I think to first order the prime importance is to show that the device meets its (well-understood) essential performance.

The ME devices that have commonly appreciated risks that are associated with Essential Performance tend to be those devices which have particular consensus standards in the "safety" (60601-2-xx) series. Some devices may have further regulatory expectations, again based (essentially) on EP.

 

[Al_Z1]

Essential perfomance in my opinion supports device possibility to fulfill its intended use, i.e. breakdowns some general requirements to particular and more narrow ones. I second @Tidge on this.

 

[d_addams]

What I think is different about Essential Performance is that the EP is what is supposed to be providing the Benefit on the side of the Benefit-Risk Analysis. Not meeting Essential Performance is IMO nullifying the benefit rather than introducing risk.

There can subtleties, but I think to first order the prime importance is to show that the device meets its (well-understood) essential performance.

The ME devices that have commonly appreciated risks that are associated with Essential Performance tend to be those devices which have particular consensus standards in the "safety" (60601-2-xx) series. Some devices may have further regulatory expectations, again based (essentially) on EP.

I think this is aligning with what I was thinking that the conclusion of unacceptable risk (within the definition of what is EP) is based on the B/R ratio being altered by the loss of benefit rather than a higher than anticipated risk of injury (occ or sev) upsetting the B/R ratio. Therefore, regardless of the resulting harms, it is deemed unacceptable because the benefit is lost.

Part of why I'm thinking about this is that regardless of any standards, why wouldn't one want to clarify or designate which of those aspects of the product can be considered EP (or whatever term one cared to use to avoid people getting uppity about using EP on a product where 60601 etc is not applicable). It just makes a lot of sense to identify those things that are critical to get the benefit regardless of any standards.

 

[Tidge]

Part of why I'm thinking about this is that regardless of any standards, why wouldn't one want to clarify or designate which of those aspects of the product can be considered EP (or whatever term one cared to use to avoid people getting uppity about using EP on a product where 60601 etc is not applicable). It just makes a lot of sense to identify those things that are critical to get the benefit regardless of any standards.

In the USA, it is essentially the classification code that effectively sets the boundaries on what any given device is for.

IMO "Essential Performance" has a close-to-tautological definition if we start trying to pin it down to functions (or features) because the 60601-1 definition of Essential Performance exists even when there are no particular standards for the device... where basic (in my words "no duh") consensus safety tests would otherwise be. My own opinion is that there needs to be separation from clinical use and marketing claims, and so I'd rather have the device classified in a way that makes sense to clinical users, and not try to be too clever and end up where each of two different devices that do the same thing stop working (under some condition), but only one of them fails to meet EP.

My experience with NRTLs is that they play a very weird game with manufacturers. I have been in conversation with them (hearsay warning!) where they will comfortably say "You (the mfg) need to tell me what EP is, for 60601-1 certification" and "I can tell from your device that there is no EP." The former is required by the standard, and the latter is a true-ish shortcut way for the NRTL to imply that they can't imagine there being any unacceptable risks as an outcome of the routine 60601-1 test they will perform.

I have some sympathy for the NRTLs when working with manufacturers seeking certification, but with limits. The NRTLs are very upfront about a MFG needing to provide a robust risk analysis, but I have yet to see a NRTL representative be able to understand a robust risk analysis. This allergy to reviewing existing risk analysis is problematic when the manufacturer and tester don't agree, which is why if a NRTL says something like "your device has no Essential Performance" you should ask them to put that assessment in writing and make sure that the RMF and test plan make that clear. I'm not suggesting to put the cart before the horse, more that a MFG wants to align with the brilliant, insightful, best-at-their-job, tester otherwise they get grumpy.