Yes, know the IMDRF SaMD docs well, the point I was trying to make (badly) was that the two parts of Rule 11 that I quoted are very similar to that which defines a medical device in article 2, if a device is used in diagnosis, therapies or monitoring of disease, injury or disability, then it is a medical device. So how do you get to Class I and still be a medical device.
I guess this is a philosophical discussion, but in practical terms, I am telling my clients that if they have an SaMD, it is difficult to see how it can be classified as anything less than a Class IIa device, if Rule 11 is the classification rule used.
I'm still not sure why you are claiming that software can only be Class IIA.
If the software is NOT intended to provide information which is used to take decisions with diagnosis or therapeutic purposes and is NOT intended to monitor physiological processes, but is still a SaMD, it will be class I as per the rule.
I think the confusion here is exactly you trying to link to the medical device definition.
If you look closely to what Rule 11 says, is says "provide information which is used to take decision with diagnosis or therapeutic purposes". This is not in the medical device definition (the diagnosis and therapeutic purposes are). The same for monitoring in the second part of Rule 11.
If you look at the definition of a medical device (bold emphasis is from me)
—
diagnosis, prevention,
monitoring, prediction, prognosis,
treatment or alleviation of disease,
—
diagnosis,
monitoring,
treatment, alleviation of, or compensation for, an injury or disability,
— investigation, replacement or modification of the anatomy or of a physiological or pathological process or state,
— providing information by means of in vitro examination of specimens derived from the human body, including organ, blood and tissue donations
Not all aspects of the definition of medical devices matches what is in rule 11 (basically, only the ones in bold). If the SaMD, in principle, is not related to ones in bold, it will in principle be Class I. I mean in principle because it's not automatic, you would need to check some other aspects, but an initial analysis should have this conclusion.