Bioequivalence and Bioavailability Forum • WinNonlin is lack of precision or?

WinNonlin is lack of precision or? [Software]

posted by yicaoting – NanKing, China, 2011-11-06 15:59 (5340 d 12:50 ago) – Posting: # 7622
Views: 13,486

Dear ElMaestro,

❝ All software that I know of with the exception of man-made Excel sheets get the LSMEans from fitting. They are the model effects extracted from the b vector after optimsation.

Yes, you are right at this point.

But, as mentioned in my first post, I used
WNL's 90% CI of LSMean(R-T) -0.066980983221424 0.124285167688542, but not the 90% CI of LSMean(R-T) calculated by other software, to calc 90% CI of GeoMean Ratio(%) (T/R), the result is different from that of WNL's.

LSMean is obtained from fitting or iteration, so some bias will be accepted.
But WNL dose a simple anti-log transform with a bias at level of 0.01(%), it is unacceptable, unless WNL use another "strange method" to do anti-log transformation and we have accept that "strange method".

Complete thread:

WinNonlin is lack of precision or? yicaoting 2011-11-06 13:03
- WinNonlin is lack of precision or? ElMaestro 2011-11-06 13:46
  - WinNonlin is lack of precision or? yicaoting 2011-11-06 14:20
    - WinNonlin is lack of precision or? ElMaestro 2011-11-06 14:32
      - WinNonlin is lack of precision or?yicaoting 2011-11-06 14:59
        
        EOD? Helmut 2011-11-06 15:39
    - WinNonlin is lack of precision or? Helmut 2011-11-06 14:46
      - use an ox cleaver to kill a chicken yicaoting 2011-11-06 15:25
        
        use an ox cleaver to kill a chicken Helmut 2011-11-06 15:47
- WinNonlin’s BE wizard is known to be buggy Helmut 2011-11-06 14:14
  - WinNonlin’s BE wizard is known to be buggy yicaoting 2011-11-06 15:38
    - WinNonlin’s BE wizard is known to be buggy Helmut 2011-11-06 16:53
      - WinNonlin’s BE wizard is known to be buggy yicaoting 2011-11-06 17:12
        
        WinNonlin’s power? Helmut 2011-11-06 18:29