As I read 21 CFR 11, the regulation deals more with ensuring the security
of the electronic health record itself. Thus, it seemed to me that so long
as the software (SAS, R, Splus, etc.) could not alter the data base in any
way then you're fine (this may be naive, but that's how I understood it).
The 'software validation' referred to in the document seems to be concerned
with software that is directly related to the device (i.e. software that is
part of the functionality of the device). How on earth could you really
validate every process that a stat software package is capable of anyway?
It seems to me that we would be better off focusing our attention on the
quality of the programming/data manipulation rather than on the 'validation
of the software' (which in terms of verifying its complete functionality
isn't possible anyway). I'm also curious that the open source (i.e.
non-black-box) nature of R hasn't struck more of a chord with regulatory
bodies.
I tried to make the case above to our software quality people, but they
insisted on a full OQ/IQ/PQ plan (which I'm in the process of drafting -
possibly my most boring task ever). In addition, my boss insisted on using
Splus instead of R because I couldn't convince him that R would be
acceptable (even though the OQ/IQ/PQ requirements would have been the same
and R would have saved money). That said, I would have had to write new
OQ/IQ/PQ plans every time there was a new version of R available (ouch). I
have no complaints with regards to Splus, but R would have been free.
One last thought to add to Bert's - I think one other thing that is holding
up the spread of S to the pharma/device companies is the availability of S
programmers (SAS programmers are plentiful). I tried to get 'has experience
in S' added to our next job opening, but my boss insisted that we would
never find a person with experience in SAS and S. I countered by asking if
we could just ask for S experience (and drop the SAS requirement), but he
gave me a dirty look. :)
Cody Hamilton, PhD
Edwards Lifesciences
[[alternative HTML version deleted]]
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
