Does anyone know if for clinical studies the FDA would accept statistical analyses performed with R ? Delphine Fontaine
Delphine, Please see the following message posted a week ago: http://comments.gmane.org/gmane.comp.lang.r.general/80175. HTH, -Mat -----Original Message----- From: r-help-bounces at stat.math.ethz.ch [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Delphine Fontaine Sent: Friday, March 09, 2007 8:29 AM To: r-help at stat.math.ethz.ch Subject: [R] R and clinical studies Does anyone know if for clinical studies the FDA would accept statistical analyses performed with R ? Delphine Fontaine ______________________________________________ R-help at stat.math.ethz.ch 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.
Thanks for your answer which was very helpfull. I have another question: I have read in this document (http://cran.r-project.org/doc/manuals/R-intro.pdf) that most of the programs written in R are ephemeral and that new releases are not always compatible with previous releases. What I would like to know is if R functions are already validated and if not, what should we do to validate a R function ? -- Delphine Fontaine Quoting "Soukup, Mat" <Mat.Soukup at fda.hhs.gov>:> Delphine, > > Please see the following message posted a week ago: > http://comments.gmane.org/gmane.comp.lang.r.general/80175. > > HTH, > > -Mat > > -----Original Message----- > From: r-help-bounces at stat.math.ethz.ch > [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Delphine Fontaine > Sent: Friday, March 09, 2007 8:29 AM > To: r-help at stat.math.ethz.ch > Subject: [R] R and clinical studies > > Does anyone know if for clinical studies the FDA would accept > statistical analyses performed with R ? > > Delphine Fontaine > > ______________________________________________ > R-help at stat.math.ethz.ch 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. > > ______________________________________________ > R-help at stat.math.ethz.ch 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. >
A strength of R is that there is a wide variety of contribuitions to the package, giving it great breadth. A weakness of R is that there is a wide variety of contributers to the package, some of whom spend a lot of time on the task of function correctness, and some of whom spend little; some worry about backward compatability, some sneer at the idea; some spend a lot of time on maintainance, and some don't have the time to do so or move on to other things. The survival code, for instance, has a set of exact test cases. These are small data sets where the correct answer has been carefully worked out by hand. S (Splus or R) passes all the tests, SAS passes most of them. (Most of the tests are documented in an appendix of Therneau and Grambsch, Springer, 2000). These test cases has been a great help in creating and debugging the code, but overall represent a large amount of work. Most code that does not have a corporate sponsor will not have the resources to do this. I have them mostly because the survival library's genesis has been spread out over 20 years, and individual bits were important parts of clinical trials and so HAD to be right. (Aside. SAS has a deserved repuation for accuracy. It has an undeserved one for infallability --- one of my favorite bug reports for the S code started out "I've found a mistake in the coxph function, it gives a different answer than SAS". It turned out in that case that the S and SAS data sets in their example were not quite the same. As an earlier poster said, data management and manipulation is the root of most errors.) Our group uses SAS for data manipulation primarily, and a mix of SAS and S-Plus for the analysis. It would be difficult to become a pure S shop, but we've had no trouble with the mix. Terry Therneau Biostatistics, Mayo Clinic
Thank you to all those that responded to Delphine's original post on R and
clinical studies. They have provided much food for thought.
I had a couple of follow up questions/comments. Andrew is very correct in
pointing out that there are classes and workshops available for R. It's my
understanding that there are even commercial versions of R that now provide
formal commercial-style courses. And at any rate, the money saved by
potentially avoiding pricey software could certainly justify any training
expense in time or money - this assumes of course that the pricey software
could be dispensed with (I suspect that would take considerable time at my
current company as so many legacy projects have been done in proprietary
software). I still think that R provides less 'hand-holding' and
requires
more initiative (which may be more or less present on a per
programmer/statistician basis).
I guess one could always integrate R/Splus in with SAS, as Terry's group
has done at Mayo - I will probably do this at least as a start. I have a
few concerns with regards to this approach (these may be needless concerns,
but I will venture expressing them anyway). First, I'm worried about the
possibility of compatability concerns (will anyone be worried about a SAS
dataset read into R or vice-versa?). Second, I would prefer focusing all
my learning on one package if possible. I actually have more experience
with SAS (as do others in my group), and if the switch to R is to be made I
would like to make that switch as complete as possible. This would also
avoid requiring new hires to know both languages. Third, if SAS is to be
kept around, it defeats one of the main advantages of having open source
code in the first place (R is wonderfully free!). Like Mayo, Baylor Health
(my previous employer) used both Splus and SAS. I was warned that data
manipulation would be much more difficult in R/Splus than it was in SAS.
To be honest, and I say this humbly realizing that most posters to this
list have much more experience than I, I haven't found data manipulation to
be that much more difficult in R/Splus (at least as I have gained
experience in R/Splus). I can think of two exceptions (1) large datasets
and (2) SAS seems to play nicer with MS products (e.g. PROC IMPORT seemed
to read in messy Excel spreadsheets better than importData in Splus). Is
it possible (and I again say this with MUCH humility) that the perceived
advantages of SAS with regards to data manipulation may be due in part to
some users only using R/Splus for stat modeling and graphics (thus never
becoming familiar with the data manipulation capabilities of R/Splus) or to
the reluctance of SAS-trained individuals and companies to make the
complete switch?
Tony, the story about the "famous software" and the "certain
operating
system" at the "large company" was priceless.
In closing, I should mention that in all posts I am speaking for myself and
not for Edwards LifeSciences.
Regards,
-Cody
Thank you to all those that responded to Delphine's original post on R and
clinical studies. They have provided much food for thought.
I had a couple of follow up questions/comments. Andrew is very correct in
pointing out that there are classes and workshops available for R. It's my
understanding that there are even commercial versions of R that now provide
formal commercial-style courses. And at any rate, the money saved by
potentially avoiding pricey software could certainly justify any training
expense in time or money - this assumes of course that the pricey software
could be dispensed with (I suspect that would take considerable time at my
current company as so many legacy projects have been done in proprietary
software). I still think that R provides less 'hand-holding' and
requires
more initiative (which may be more or less present on a per
programmer/statistician basis).
I guess one could always integrate R/Splus in with SAS, as Terry's group
has done at Mayo - I will probably do this at least as a start. I have a
few concerns with regards to this approach (these may be needless concerns,
but I will venture expressing them anyway). First, I'm worried about the
possibility of compatability concerns (will anyone be worried about a SAS
dataset read into R or vice-versa?). Second, I would prefer focusing all
my learning on one package if possible. I actually have more experience
with SAS (as do others in my group), and if the switch to R is to be made I
would like to make that switch as complete as possible. This would also
avoid requiring new hires to know both languages. Third, if SAS is to be
kept around, it defeats one of the main advantages of having open source
code in the first place (R is wonderfully free!). Like Mayo, Baylor Health
(my previous employer) used both Splus and SAS. I was warned that data
manipulation would be much more difficult in R/Splus than it was in SAS.
To be honest, and I say this humbly realizing that most posters to this
list have much more experience than I, I haven't found data manipulation to
be that much more difficult in R/Splus (at least as I have gained
experience in R/Splus). I can think of two exceptions (1) large datasets
and (2) SAS seems to play nicer with MS products (e.g. PROC IMPORT seemed
to read in messy Excel spreadsheets better than importData in Splus). Is
it possible (and I again say this with MUCH humility) that the perceived
advantages of SAS with regards to data manipulation may be due in part to
some users only using R/Splus for stat modeling and graphics (thus never
becoming familiar with the data manipulation capabilities of R/Splus) or to
the reluctance of SAS-trained individuals and companies to make the
complete switch?
Tony, the story about the "famous software" and the "certain
operating
system" at the "large company" was priceless.
In closing, I should mention that in all posts I am speaking for myself and
not for Edwards LifeSciences.
Regards,
-Cody
Thanks for the tip. I will look forward to trying this package out soon!
Regards, -Cody
Hans-Peter
<gchappi at gmail.co
m> To
"Cody_Hamilton at edwards.com"
03/23/2007 09:31 <Cody_Hamilton at edwards.com>
AM cc
Subject
Re: [R] R and clinical studies
Hi,
2007/3/20, Cody_Hamilton at edwards.com <Cody_Hamilton at edwards.com>:
> and (2) SAS seems to play nicer with MS products (e.g. PROC IMPORT seemed
> to read in messy Excel spreadsheets better than importData in Splus).
(to pick one small detail from your post)
You can use my xlsReadWrite package which will (on windows) read and
write Excel data (-> see CRAN, a new version is pending). While there
is a pro version also, in lot of circumstances the free version is
perfectly fine.
--
Regards,
Hans-Peter