Lund's Test for Outlier Analysis

[ahmedazab88]

Dear all

Can you help me to make Lund Test for Outlier Analysis in Ratios to be used in bioequivalence studies, var AUC & Cmax

For Example:

Obs	Subject	Ratio for AUCT	Ratio for AUC00	Ratio for Cmax	Lund	Prescott	Outlier
1	1	1.40988	1.31543	0.91176	2.73085	3.27961	No
2	2	0.95385	0.94783	0.96886	1.00064	3.27961	No

Best Regards

Ahmed

[ahmedazab88]

Dear all

Can you help me to make Lund Test for Outlier Analysis in Ratios to be used in bioequivalence studies, var AUC & Cmax

For Example:

Obs	Subject	Ratio for AUCT	Ratio for AUC00	Ratio for Cmax	Lund	Prescott	Outlier
1	1	1.40988	1.31543	0.91176	2.73085	3.27961	No
2	2	0.95385	0.94783	0.96886	1.00064	3.27961	No

Best Regards

Ahmed

[Rick_SAS]

Try posting this to the Community for SAS Statistical Procedures

I am not familiar with Lund's Test. Can you post a reference?  I found this discussion

http://forum.bebac.at/mix_entry.php?id=2144

Many years ago james Handsfield at the CDC developed a SAS macro for detecting outliers. If you want to see what he did, his paper is at http://www2.sas.com/proceedings/sugi23/Posters/p197.pdf

[ahmedazab88]

Thank you Rick

[ahmedazab88]

I found this code:

Proc glm outstat=Lund;

class Subject Treat Seq Period;

MODEL Log_AUCT Log_AUCinf Log_Cmax=seq subject(seq) Treat period;

lsmeans Treat/stderr; Estimate 'test-ref' Treat -1 1;

Output out=Lund P=PRED R=RESID stdr=eresid STUDENT=STDRESID;

run;

Best Regards

Ahmed

[varinderk]

I have to use Lunds test for Oultlier detection in BE study. In some reference, I found SAS code CALCKE is used for this test. please share the script to run this test

[sbxkoenk]

Hello @varinderk ,

I think the CALCKE code is offered in a very voluminous PDF (45MB) from the FDA in the USA.
https://www.accessdata.fda.gov/drugsatfda_docs/nda/2020/203760Orig1s000.pdf

Search for CALCKE, there's SAS macro code beyond page 400!

Kind regards,

Koen