Regulatory Open Forum

Dear Experts,

Should artificial intelligence software that provides a diagnosis or diagnostic aid, based on the analysis of responses to a questionnaire sent directly to the patient, on a web platform, be qualified as a DM or an IVD? If it is an IVD, should it be qualified as a self-test?
If the diagnosis is completed by the analysis of a blood sample (or other biological fluid), obtained by a health professional, does this change its qualification (DM or IVD)?

Thanks a lot for your answers.

------------------------------
PASCALE
France
------------------------------

Bonjour Pascale,

Unfortunately this probably can not be fully answered without completely understanding the purpose and use of the software.  This is definitely in the evolving area of regulatory classification understanding whether Software as a Medical Device (SaMD) could be a medical device or an IVD.  In my experience, if just taking health information is typically a medical device and if associated with analysis of human samples would be an IVD.  However, as mentioned there usually is many considerations made concerning the intended user, intended purpose, and clinical decision process from the results.

------------------------------
Richard Vincins RAC
Vice President Global Regulatory Affairs
Oriel STAT A MATRIX - ENTERPRISE
------------------------------

Original Message:
Sent: 27-May-2021 09:22
From: PASCALE BLANGARIN
Subject: European qualification of a software (MD or IVD)

Dear Experts,

Should artificial intelligence software that provides a diagnosis or diagnostic aid, based on the analysis of responses to a questionnaire sent directly to the patient, on a web platform, be qualified as a DM or an IVD? If it is an IVD, should it be qualified as a self-test?
If the diagnosis is completed by the analysis of a blood sample (or other biological fluid), obtained by a health professional, does this change its qualification (DM or IVD)?

Thanks a lot for your answers.

------------------------------
PASCALE
France
------------------------------

Dear Richard

Thank you very much. According your answer, I think that the software in question is partly a DM and partly an IVD and in this case I assume that each of them must meet the requirements of the corresponding regulation?

Kind Regards

------------------------------
PASCALE
France
------------------------------

Original Message:
Sent: 28-May-2021 02:27
From: Richard Vincins
Subject: European qualification of a software (MD or IVD)

Bonjour Pascale,

Unfortunately this probably can not be fully answered without completely understanding the purpose and use of the software.  This is definitely in the evolving area of regulatory classification understanding whether Software as a Medical Device (SaMD) could be a medical device or an IVD.  In my experience, if just taking health information is typically a medical device and if associated with analysis of human samples would be an IVD.  However, as mentioned there usually is many considerations made concerning the intended user, intended purpose, and clinical decision process from the results.

------------------------------
Richard Vincins RAC
Vice President Global Regulatory Affairs
Oriel STAT A MATRIX - ENTERPRISE

Original Message:
Sent: 27-May-2021 09:22
From: PASCALE BLANGARIN
Subject: European qualification of a software (MD or IVD)

Dear Experts,

Should artificial intelligence software that provides a diagnosis or diagnostic aid, based on the analysis of responses to a questionnaire sent directly to the patient, on a web platform, be qualified as a DM or an IVD? If it is an IVD, should it be qualified as a self-test?
If the diagnosis is completed by the analysis of a blood sample (or other biological fluid), obtained by a health professional, does this change its qualification (DM or IVD)?

Thanks a lot for your answers.

------------------------------
PASCALE
France
------------------------------