Electronic Fever Thermometer - Why not IEC 62304 Class C?

OccamMan

Involved In Discussions
I'm sure this is a dumb question, but why shouldn't the software in an over-the-counter electronic fever thermometer be 62304 Class C? If the thermometer's software displays a normal temperature when the user actually has a fever, it could delay treatment thus causing serious injury or death. (I'm assuming no hardware mitigation here.)

I normally work on things that are obviously Class C, but this class doesn't feel right for an OTC thermometer, particularly given that they're only Class I devices in the EU.

Thanks in advance!
 

Pads38

Moderator
Re: Electronic fever thermometer: Why not IEC 62304 Class C?

I think that a clinical thermometer is not used on it's own to form a diagnosis. You take a temperature to confirm the patient is running a fever - you already know they are sick.

But a reading of 105F might be enough to really set the alarm bells going, whilst 102F might be 'wait and see'. Perhaps the truly critical performance required of the thermometer is pure accuracy, which can be verified by simple testing. Incorrect operation of the 'software' aspects would likely result in a totally non-functioning device.

However, if the thermometer reading was used to automatically control something (heaters / coolers / drug infusion) I think that would require a higher classification.
 

Roland chung

Trusted Information Resource
Re: Electronic fever thermometer: Why not IEC 62304 Class C?

I am evaluating the software for clinical thermometer per EN 62304. I am wondering if it is suitable to assign a safety class A for the thermometer.

Thanks,
Roland
 

c.mitch

Quite Involved in Discussions
I Roland
The safety class depends on your risk analysis based on the device intended use.
What is a consequence of a software failure? Death? Minor injury? Or no injury at all?.
If the answer is no injury then the software is class A.
 

Pads38

Moderator
Assuming that we are considering an electronic thermometer for home use -

1) How much significance would any reading be given?

2) What "treatments" would be instigated or withheld depending upon the reading?

I would suggest that - in the home environment - the temperature is only really used to confirm what you already know. Your child is clearly unwell, you use the thermometer and, yes, they have a fever. But, as I said, you probably knew that already. My mother just used the back of her hand to "measure temperature".

What "treatment" would be instigated or withheld? Well, most likely the ill child would be put in bed with fluids and paracetamol (Calpol)/ aspirin. But that would happen whatever the thermometer did or did not record.

Let us now imagine that the condition of the patient deteriorates, so professional medical attention is sought. So as part of their initial enquiry they will ask for the "history". Significant facts here would be "started feeling unwell 12 hours ago, then 3 hours ago became semi-delirious and could not cope with the light being on". Maybe you can state that 12 hours ago temperature was 38.5C and 40C 3 hours ago.

From that there are clear signs that could cause grave concern, but the temperature readings are only a small part of that. The "trend" is clear, but the rapid deterioration of the patients condition is obvious. Again, what delay to treatment would be likely? The clear signs of rapid deterioration and the photophobia, particularly in a paediatric patient, would "set the alarm bells" ringing and demand immediate further investigation.

So, as I suggested earlier, in a home environment, a clinical thermometer has little, if any, clinical significance so you can be justified in claiming "Class A" designation.

In a hospital setting, where by careful use of timed readings taken using a standardised method, much greater significance can be given to the readings. It should be possible to detect an infection starting before the patient starts to feel unwell, or in an incapacitated patient (not communicating) the elevated temperature may be the only sign. So the temperature reading is much more significant in the choice of treatment, so elevating the risk level of any device involved.
 
G

Guillaume Prom

"particularly given that they're only Class I devices in the EU"

if you look at MedDev 2.4/1 rev 9, page 42, electronic thermometer can be class IIa (active device that allows direct diagnostic).

it all depends on its intended use.
 
M

mikezv

You need to question why did the software give a wrong reading. If this is due to hardware issues (sensor not calibrated, drifted, etc.), then the software is working correctly and doesn't cause the inocrrect reading. If this is a software bug, then you should ask why wasn't this caught during software V&V. As such, I don't think there is any way we can justify this software to be Class C.
 
Top Bottom