Does anybody know if R is FDA or ICH (or EMEA...) compliant? AFAIK S-Plus is but that means nothing... -- Trebate bolji pristup internetu? Nazovite IskonInternet na 0800 1000 ili pogledajte http://www.iskon.biz/individualni/usluge/dialup/
On Wed, 26 Nov 2003, Antonia Drugica wrote:> Does anybody know if R is FDA or ICH (or EMEA...) compliant? AFAIK S-Plus > is but that means nothing...Actually, "FDA compliant" means nothing. The FDA does not certify statistical software and responsibility for validation is left entirely up to the user. Insightful Co, the S-PLUS vendors provide some tools to help you with validation and some information on validation that pharmaceutical companies have done. The only validation tools provided with R are the tests used by the developers. These show which aspects of R have been tested and might help other people develop tests to examine specific areas more carefully. -thomas
"Antonia Drugica" <antoniamarija at net.hr> writes:> Does anybody know if R is FDA or ICH (or EMEA...) compliant? AFAIK S-Plus > is but that means nothing...As Thomas pointed out, that does mean nothing -- there was a group of folks discussing what might be done to help, earlier this year, but then everyone got busy... best, -tony -- rossini at u.washington.edu http://www.analytics.washington.edu/ Biomedical and Health Informatics University of Washington Biostatistics, SCHARP/HVTN Fred Hutchinson Cancer Research Center UW (Tu/Th/F): 206-616-7630 FAX=206-543-3461 | Voicemail is unreliable FHCRC (M/W): 206-667-7025 FAX=206-667-4812 | use Email CONFIDENTIALITY NOTICE: This e-mail message and any attachme...{{dropped}}
I'm quite new to this medical stuff. But my associates told me that we are not free in choice of Statistical Software because the FDA has high standards concerning this topic. But if they would prefer a specific package (like SAS) that could mean, that this package vendourer could lay back and hold it's hand open for licence money. Is there any part of the ICH document referring to software packages? I really would use R for some tasks but therefor I need arguments...>"Antonia Drugica" <antoniamarija at net.hr> writes: > >> Does anybody know if R is FDA or ICH (or EMEA...) compliant? AFAIKS-Plus>> is but that means nothing... > >As Thomas pointed out, that does mean nothing -- there was a group of >folks discussing what might be done to help, earlier this year, but >then everyone got busy... > >best, >-tony > >-- >rossini at u.washington.edu http://www.analytics.washington.edu/ >Biomedical and Health Informatics University of Washington >Biostatistics, SCHARP/HVTN Fred Hutchinson Cancer Research Center >UW (Tu/Th/F): 206-616-7630 FAX=206-543-3461 | Voicemail is unreliable >FHCRC (M/W): 206-667-7025 FAX=206-667-4812 | use Email > >CONFIDENTIALITY NOTICE: This e-mail message and any attachme...{{dropped}} > >______________________________________________ >R-help at stat.math.ethz.ch mailing list >https://www.stat.math.ethz.ch/mailman/listinfo/r-help >-- Trebate bolji pristup internetu? Nazovite IskonInternet na 0800 1000 ili pogledajte http://www.iskon.biz/individualni/usluge/dialup/
From: Frank E Harrell Jr <feh3k at spamcop.net>> per year in SAS licenses and have to hire armies of non-intellectually > challenged SAS programmers to do the work of significantly fewer > programmers that use modern statistical computing tools like R and S-Plus, > it is surprising that SAS is still the most commonly used tool in the > clinical side of drug development. I quit using SAS in 1991 because my > productivity jumped at least 20% within one month of using S-Plus.I have not used SAS for even longer than you but to give SAS its due: - its pretty easy to produce all the info you need for a complete analysis with a few SAS commands. It would be possible to create analogous R commands but as it stands you have to keep going back and forth with R rather than just get it all out at once like you can with SAS. - SAS has more functionality in missing values. You can have different types of SAS missing values but in R you can have only one type of missing value. - the BY phrase in SAS is incredibly powerful and handy. You can get the same effect in R but I think that specific functionality is easier with SAS. Obviously R is incredibly powerful and functional and I really am out of touch with the SAS world but I thought I would make whatever case I could. I am willing to be corrected by those more in the know with SAS if this wrong.
> "Antonia Drugica" <antoniamarija at net.hr> writes: > > > Does anybody know if R is FDA or ICH (or EMEA...) > compliant? AFAIK S-Plus > > is but that means nothing... > > As Thomas pointed out, that does mean nothing -- there was a group of > folks discussing what might be done to help, earlier this year, but > then everyone got busy...FDA has a guidance document for off-the-shelf software: http://www.fda.gov/cdrh/ode/guidance/585.html Note that if focuses on OTS used in medical devices. However you should read it. The document: http://www.fda.gov/cdrh/comp/guidance/938.html Has a section on applicability of the software guidance (which encompasses stuff outside the instrument itself. Since I am no lawyer, I can't say whether R falls within this scope. It is fair to say however that the FDA consider safety and effectiveness very important. If the effectiveness that you claim is based on statistics provided by software, or you rely in software for determining safe levels (eg of a drug) then I would say (as a layman) it is largely irrelevant whether the vendor claims some sort of "FDA badge" because that does not prevent someone from writing dodgy scripts. So what you can do (other than soliciting mail list opinions) includes: o Think. What are the implications for end users, patients etc. Would you take a pill based on your own stats? o Read what the FDA have to say. o Evaluate the risk and safety implications of the statistics you use. o Manage the risk. Eg can you indepently confirm the key results? o Your scripts are software - the FDA requires evidence of a credible process in the life cycle of software, whether they be spreadsheets, real time control systems or whatever. OTS software that is "validated" does not remove responsibility for reducing risk to acceptable levels. HTH paul
Concerning the debate about SAS vs. S/R I think that a multiple approach has to be taken. As far as I observed it many Pharmas are going to use more than one Spftwarepackage. We all now that S SAS SPSS and so on have their advantages and disatvantages. I think a modern statistician has to know at least three or four softwarepackages so that he can decide which problem can be solved by which software.>On Thu, 27 Nov 2003 07:48:02 +0100 >"Antonia Drugica" <antoniamarija at net.hr> wrote: > >> I'm quite new to this medical stuff. But my associates told me that we >> are not free in choice of Statistical Software because the FDA has high >> standards concerning this topic. But if they would prefer a specific >> package (like SAS) that could mean, that this package vendourer could >> lay back and hold it's hand open for licence money. > >Your associates are completely wrong. It is only sponsors that choose not >to be free in their choice, due in my humble opinion mainly to the fact >that SAS has been in use since 1966 and that "no one has ever been >criticized by the FDA for using SAS." FDA even receives submissions based >on Excel and we all know about the accuracy of Excel's statistical >calculations. High standards need to be held by statisticians doing the >analyses. Related to such standards open source systems such as R have >many advantages, and the reproducible reporting capabilities of R using >its Sweave package have major impacts on accuracy of reporting. > >I along with colleagues at another institution are working on an open >source R package for clinical trial analysis and reporting that should be >mature in about a year. I am currently using the package in two >pharmaceutical industry-sponsored randomized clinical trials to report to >data monitoring committees. I'm also working on a document addressing >validation of statistical calculations. Let me know if you'd like a copy >of the current version of that document. > >> >> Is there any part of the ICH document referring to software packages? I >> really would use R for some tasks but therefor I need arguments... > >Don't know of anything in ICH. > >In view of the fact that large pharma companies have to pay more than $10M >per year in SAS licenses and have to hire armies of non-intellectually >challenged SAS programmers to do the work of significantly fewer >programmers that use modern statistical computing tools like R and S-Plus, >it is surprising that SAS is still the most commonly used tool in the >clinical side of drug development. I quit using SAS in 1991 because my >productivity jumped at least 20% within one month of using S-Plus. >--- >Frank E Harrell Jr Professor and Chair School of Medicine > Department of Biostatistics Vanderbilt University >-- Trebate bolji pristup internetu? Nazovite IskonInternet na 0800 1000 ili pogledajte http://www.iskon.biz/individualni/usluge/dialup/