LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (August 2008, week 3)Back to main SAS-L pageJoin or leave SAS-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:   Wed, 20 Aug 2008 16:28:36 -0700
Reply-To:   Stephen McDaniel <stephen@STEPHENMCDANIEL.US>
Sender:   "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:   Stephen McDaniel <stephen@STEPHENMCDANIEL.US>
Subject:   Re: Validating SAS for FDA?
Comments:   To:
In-Reply-To:   <>
Content-Type:   text/plain; charset="UTF-8"

A very long list is used in submissions, from simple stats, t-tests, sample size calculations, power analysis, regression, ANOVA, non-parametric techniques (chi square, CMH, etc), monte carlo simulations, survival analysis, etc. This is why there are whole Masters and PhD programs just around Biostats... I left off design for early studies, pharmacoeconomic analysis, etc.

Best, Stephen

-----Original Message----- From: [] Sent: Wednesday, August 20, 2008 9:27 AM To: Stephen McDaniel Cc: Subject: Re: Validating SAS for FDA?

Just what specific statistical function does the clinical pharma industry use ( eg means ,mode, regression etc)

On 8/20/08, Stephen McDaniel <> wrote: > I remember way back in the late 90's when I worked for Oracle with the > Oracle Clinical consulting team... This is a product meant for managing > clinical trials data for pharma's. The main product manual prominently said > on page one something to the effect, "... not for use in nuclear, aerospace, > or medical applications..., any use in this realm is explicitly not under > warranty and Oracle can't be held accountable for the precise standards > needed for these types of applications..." or something to that effect. > > Clients loved it when I pointed out that their multi-million dollar software > package stated this in the main manual. Of course, many companies use it > for submissions. Ultimately, the FDA holds the pharma accountable, not the > software maker. That said, if they saw consistent problems with an > application, they might be much harder on any pharma using the app, > ultimately hurting the software company via lower sales and discount demands > by the customers. > > Regards, > Stephen > > On Wed, Aug 20, 2008 at 9:00 AM, Mary <> wrote: > >> They might want to change the name...I can't see the FDA as being wild >> about verifying that software with "BUGS" as part of the name is actually >> free of bugs! >> >> -Mary >> ----- Original Message ----- >> From: Kevin Viel >> To: SAS-L@LISTSERV.UGA.EDU >> Sent: Wednesday, August 20, 2008 10:47 AM >> Subject: Re: Validating SAS for FDA? >> >> >> >For what it's worth, my provisional plan is to demonstrate that >> >WinBUGS accurately replicates textbook examples, to document this, and >> >to place the document on file. >> >> I am a bit removed from that field, but why turn to textbook examples >> instead of simulating the data? >> >> -Kevin >> >

-- Regards,

Ajay Ohri

Back to: Top of message | Previous page | Main SAS-L page