Second opinion (PHX 6.3) [RSABE / ABEL]
❝ For AUCt
❝ Ratio: 94.7377; 90% CI: 87.4898-102.586
PHX tells me:
Ratio: 94.7377; 90% CI: 87.4898–102.586
Warning 11091: Newton's algorithm converged with modified Hessian. Output is suspect.
Model may be over-specified. A simpler model could be tried.
PHX after 6 iterations:
Final variance parameter estimates:
lambda(1,1)_11 0.44268399
lambda(1,2)_11 0.42362946
lambda(2,2)_11 0.16237525
Var(Period*Formulation*Subject)_21 0.053010265
Var(Period*Formulation*Subject)_22 0.061648310
❝ However, Proc Mixed failed to compute the above for ln AUCi
Contrary to SAS PHX ‘succeeded’ for AUCi as well …
Ratio: 95.2967; 90% CI: 88.4014–102.730
… but throws the same warning as above.
❝ The last output from Proc Mixed on ln AUCi was:
❝ Covariance Parameter Values At Last Iteration
❝ Cov Parm Subject Group Estimate
❝ FA(1,1) subject 0.3722
❝ FA(2,1) subject 0.3975
❝ FA(2,2) subject 0.2362
❝ Residual subject formulation Ref 0.06248
❝ Residual subject formulation Test 0.01460
PHX after 6 iterations:
Final variance parameter estimates:
lambda(1,1)_11 0.37216022
lambda(1,2)_11 0.39747315
lambda(2,2)_11 0.15492894
Var(Period*Formulation*Subject)_21 0.0624828
Var(Period*Formulation*Subject)_22 0.0463781
❝ Any idea?
Nope. Let’s wait for the SAS-guru Detlew.
IMHO, since a partial replicate according to FDA’s model is always (!) over-specified there is no guarantee that the LME-engine will converge. Don’t blame SAS and PHX; they warn us… Stupid design. If you want to have only three periods maybe it is better to run a fully replicated design (TRT|RTR) in the future.
P.S.: You are not alone. Last week a colleague posted at Pharsight’s Extranet an example where a replicate design failed to converge for Cmax (but not for AUCt and AUC∞). Pharsight suggested to change the variance structure to Heterogeneous Compound Symmetry (instead of FDA’s Banded No-Diagonal Factor Analytic [f=2]). In my experience this rarely helps…
BTW, does anybody know the rationale behind FDA’s partial replicate? Higher precision of the estimate of CVwR (see this post and followings)?
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
Complete thread:
- FDA's HVD SAS Code from Progesterone Guidance jag009 2013-04-22 16:39 [RSABE / ABEL]
- Proc MIXED vs. Proc GLM Helmut 2013-04-22 17:47
- Proc MIXED vs. Proc GLM jag009 2013-04-22 20:33
- SAS vs. PHX Helmut 2013-04-22 22:10
- SAS vs. PHX Shuanghe 2013-04-23 11:45
- SAS vs. SAS d_labes 2013-04-23 16:50
- SAS vs. SAS vs. Phoenix Helmut 2019-11-12 13:01
- SAS vs. SAS d_labes 2013-04-23 16:50
- SAS vs. PHX jag009 2013-04-23 15:56
- SAS Warning (Note) on Proc Mixed jag009 2013-04-23 17:07
- Nut Job.. jag009 2013-04-30 21:09
- Second opinion (PHX 6.3)Helmut 2013-05-01 16:21
- Second opinion (PHX 6.3) jag009 2013-05-01 17:16
- Second opinion (PHX 6.3) ElMaestro 2013-05-01 18:48
- In praise of a full replicate Helmut 2013-05-01 19:16
- Third opinion d_labes 2013-05-02 12:00
- Compound Symmetry Helmut 2013-05-02 14:41
- Compound Symmetry - SASian (1) d_labes 2013-05-02 16:26
- Variance=0 Helmut 2013-05-03 16:07
- Compound Symmetry - SASian (1) d_labes 2013-05-02 16:26
- Compound Symmetry Helmut 2013-05-02 14:41
- Second opinion (PHX 6.3) jag009 2013-05-01 17:16
- Second opinion (PHX 6.3)Helmut 2013-05-01 16:21
- Nut Job.. jag009 2013-04-30 21:09
- SAS Warning (Note) on Proc Mixed jag009 2013-04-23 17:07
- SAS vs. PHX Shuanghe 2013-04-23 11:45
- SAS vs. PHX Helmut 2013-04-22 22:10
- Proc MIXED vs. Proc GLM jag009 2013-04-22 20:33
- Proc MIXED vs. Proc GLM Helmut 2013-04-22 17:47