What to disclose for 60601-2-27? Type tests? Under performance? What the test house says?

[Loekje]

Hi,
60601-2-27 says in clause 201.7.9.2.9.101 Additional instructions for use
that disclosure shall be made for various performance parameters, like which maximum T-wave can be rejected.
We are a unsure how this disclosure is supposed to work:

• Is it something that rolls automagically out of test house testing?
• Do we have to specify it and does the test house test for compliance to our given figures?
• Is it common to be on the safe side, thus specifying underperformance?
• Does disclosure include specifying that figures are obtained from representative samples or is that not necessary?

Thanks for your attention,
Loek

PS. if someone alreday has the (direct link to) wave forms A1-4,B1-2 in a format that we can feed into a signal generator we would be very much onbliged.

 

[Peter Selvey]

• Is it something that rolls automagically out of test house testing?

No! OMG! Your task is to design a system that basically guarantees a result (within reason) so that when you submit a sample to the test lab, you already know what the result will be. This is fundamental to the concept of a TYPE TEST. Do not EVER, NEVER, EVER take the results from a test lab and use them as indicative of what your product can do in the real world. You should already know what a Tall T-wave test involves (discrimination between the high dV/dT of the T-wave and a normal QRS); you should know exactly how to deal with it and at what point your algorithm messes up a dV/dT of the T-wave as a new QRS.

• Do we have to specify it and does the test house test for compliance to our given figures?

Yes!!! That's what it's all about!

• Is it common to be on the safe side, thus specifying underperformance?]

Yes!!! That's what it's all about!

• Does disclosure include specifying that figures are obtained from representative samples or is that not necessary?

Disclosure is about confidence. Do not ever (ever) disclose specifications just because a third party test lab (that probably doesn't understand what they are doing) gave you a thumbs up in testing.

The waveforms are available for free from Physionet, at least from my memory. If you can't find them let us know.

 

[Loekje]

Hi Peter,

Your response is really appreciated.

A problem with specifying underperformance is that the device may show signals that your disclosure says it is beyond the capabilities
(like, you specify what the maximum T-wave is without interfering with heartrate computation). IMHO it may confuse a medical professional if your device shows signals that are not suppose to be shown, and vice-versa.
All other type tests are always about proving that your device is under or above a certain limit (and your quality system should assure that your production process is such that all products comply to the same type tests). But these disclosure things can refer to measured values that might be product sample dependent. It would be at least a bit easier if the standard asks to specify a value plus a tolerance (and requires a maximum on the allowed tolerance).

BTW: I have found the waveforms, but it is not straightforward:
The waveforms (A4-B2) in IEC-60601-2-27 are identical to the ones in ANSI/AAMI EC13, but in there they are called 3a-3d and 4a-4b.
Within Physionet the reference is to EC13: ANSI/AAMI EC13 Test Waveforms v1.0.0 , and that is why searching on references to -2-27 terms gives no results.

Ah well, I think that if this was a really important matter it would be better specified. So we will just measure, analyse and disclose

Cheers!
Loek

 

[Peter Selvey]

I get what you mean, there are some tests like that, where a disclosure is should be a range, not a single value. For Tall T-Wave, I don't think it's that kind of test. For my memory (which is getting old) is that the IFU declares a value like 1.2mV and the test lab checks it's all OK to that point. In reality the monitor might start double counting at 1.5mV, but I think it's wrong to use that figure.

From memory the Physionet waveforms are the same but the labels are different. Back in 2009 when I was developing the ECG testers (e.g. the SECG) I used the waveforms from Physionet, I guess I would have researched all this at the time. As far as I know, the IEC adopted the test from AAMI and the waveforms are identical.

At the time I got involved in ECG testing (around 2007) I had the feeling there was this small club of manufacturers that knew how to do the tests, and wrote the standard. The third party labs just blindly trusted the manufacturers. So the ECG standards were a mess. I was lucky UL Japan agreed to let me have a go at real independent testing. I found the ECG standards to be full of errors and missing info like where to get waveforms from. A lot of this has been fixed but there are still some errors or "in the know" stuff that remains.