From: Paul Hutson <>
Subject: Multiple dose levels
Date: Fri, 28 Sep 2001 12:57:53 -0500

I am analyzing Phase I data from patients treated at multiple (4) dose
levels (LVL). Is there a convenient means of having analysis skip records
for all patients except those with a desired dose level so that THETAs for
a given dose level can be determined separately? Is it preferred to just
set up separate data files for each level?
Thanks, everyone. Have a great weekend.

Paul Hutson, Pharm.D.
Associate Professor (CHS)
UW School of Pharmacy
NOTE NEW ADDRESS effective 6/2001
777 Highland Avenue
Madison, WI 53705-2222
Tel: (608) 263-2496
FAX: (608) 265-5421
Pager: (608) 265-7000, #7856





From: (Joern Loetsch)
Subject: Re: Multiple dose levels
Date: Fri, 28 Sep 2001 21:45:10 +0200

Why do you want separate analyses for each dose level?
If this is needed, I would play with different MDV records, i.e., 4 MDV
records, each with MDV=0 for the dose level you want to analyse, and MDV=1
otherwise (of course, MDV=1 when it is a "real MDV").

Jorn Lotsch, MD
pharmazentrum frankfurt, Department of Clinical Pharmacology
Johann Wolfgang Geothe-University Hospital
Theodor-Stern-Kai 7
D-60590 Frankfurt
Phone: +49-69-6301-4589
Fax: +49-69-6301-7636





From: Lewis B Sheiner <>
Subject: Re: Multiple dose levels
Date: Mon, 01 Oct 2001 07:10:43 -0700

It is possible, but why would you want to do it?
Do you mean all parameters including OMEGA and SIGMA?
If you really mean that, then just analyze the 4 data sets separately
(although I cannot imagine the circumstances that would cause ALL
parameters to differ when you change only the dose ...)?
If you mean to keep some parameters in common among the data sets,
then you will need an indicator variable denoting the data set, and
appropriate coding of models to invoke different thetas when the data sets
(doses) differ.