Hello Akshay,
Unfortunately your question is not so easy to answer in current context because this could be one of those or all of those methods depending on the type of software, level within the device, if SaMD, and/or risk involved. The important point is in a software verification testing protocol, this would be clearly defined. I would guess depending on some software the type of verification method might be implicit in the actual testing methodology, for example if a low level algorithm is being tested, then A and D (and maybe T) would be implicit. Then in other cases, there may be software controlling sensors where a physical I or visual inspection would need to be clearly stated. You are definitely on the right track to address IADT, apply this as appropriate to the software, the device, and the testing.
------------------------------
Richard Vincins RAC
Vice President Global Regulatory Affairs
------------------------------
Original Message:
Sent: 08-Dec-2021 15:34
From: Akshay Kulkarni
Subject: Software Verification
Should be more clear - by method of verification, I mean whether we should include which method of
- Inspection (I): control or visual verification
- Analysis (A): verification based upon analytical evidences
- Demonstration (D): verification of operational characteristics, without quantitative measurement
- Test (T): verification of quantitative characteristics with quantitative measurement
was used.
Thanks again!
------------------------------
Akshay Kulkarni
United States
Original Message:
Sent: 08-Dec-2021 12:06
From: Akshay Kulkarni
Subject: Software Verification
Hi All,
I had a question regarding SW Verification protocols - is it required to document the method of verification used as an entry of its own, or will it be covered either in the textual description of the test or implicit in the testing being carried out? What have your experiences been so far w.r.t. the FDA?
Thanks!
------------------------------
Akshay Kulkarni
United States
------------------------------