Log in
Register|
AngusMcLean ★★ USA, 2013-05-10 23:36 (4364 d 21:37 ago) Posting: # 10568 Views: 4,411 |
|
|
Two programs commonly used for BE calculations include Phoenix from Pharsight and Proc. Mixed (SAS). Equiv test is used and is (I find) most useful, but is limited to a 2-way cross-0ver studies. Is there any information comparing the BE results obtained for Pharsight's Phoenix and Proc. Mixed for 2 way and 3-way cross over studies. I believe the FA does not endorse any one program. Angus Edit: Category changed. [Helmut] |
|
Helmut ★★★   Vienna, Austria, 2013-05-11 00:03 (4364 d 21:10 ago) @ AngusMcLean Posting: # 10569 Views: 3,651 |
|
|
Hi Angus! ❝ Is there any information comparing the BE results obtained for Pharsight's Phoenix and Proc. Mixed for 2 way and 3-way cross over studies. You’ll get the same results. Since you have to validate the software anyway simply compare results in Phoenix of data sets from Chow/Liu, Patterson/Jones, and Hauschke/Steinijans/Pigeot with their results – which were obtained in SAS. You only have to be cautious if you want to reproduce Proc GLM in Phoenix. Proc GLM drops incomplete data (i.e., a subject completes only one period), whereas Phoenix always runs REML. Therefore, in Phoenix you have to exclude subjects with incomplete data first. ❝ I believe the FA does not endorse any one program. Correct. See this thread and some linked references. — Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
AngusMcLean ★★ USA, 2013-05-11 06:51 (4364 d 14:22 ago) @ Helmut Posting: # 10570 Views: 3,764 |
|
|
Helmut: Thank you for your response. Yes; I do have one data set that I ran for BE in WinNonlin some years ago and also ran in a version of SAS (SAS LE). It is a 3-way X over design with 6 sequences (so it is balanced). I ran it in SAS GLM and WinNonlin. The BE results were exactly the same. No subjects dropped out. I am looking for such a data set and BE results of an example like above run in SAS Proc Mixed so I can compare with Phoenix Nonlin. I have Chow and Liu's book, but examples are limited. I do not see a Proc-Mixed example. Any idea of where I could find such an example. Angus Edit: Full quote removed. Please delete everything from the text of the original poster which is not necessary in understanding your answer; see also this post! [Helmut] |
|
Helmut ★★★ Vienna, Austria, 2013-05-12 03:58 (4363 d 17:15 ago) @ AngusMcLean Posting: # 10573 Views: 3,589 |
|
|
Hi Angus! ❝ I am looking for such a data set and BE results of an example like above run in SAS Proc Mixed so I can compare with Phoenix Nonlin. I have Chow and Liu's book, but examples are limited. I do not see a Proc-Mixed example. Any idea of where I could find such an example. Maybe this thread helps? Chow/Liu’s 6×3 example, evaluated in SAS Proc Mixed by Detlew and in PHX6.1 by myself. Another option will be to post your project-file at Pharsight’s Extranet; I’m pretty sure Ana or Linda will be happy to re-run it in SAS. — Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
Ing. Helmut Schütz