From: "Tsai, Max"
Subject:  [NMusers] Random sampling times 
Date: Mon, March 14, 2005 8:30 am 

Hello all.  I am trying to design a sparse sampling scheme by
simulating random sampling times, which would subsequently be used
to generate concentration data.  This data would then be fitted
to a previously developed POPPK model and the PK parameters
would be compared to those modeled from the full time series.
I would like to implement a control stream that would simulate
a time series with the capability of generating random deviations
from the nominal time within certain time intervals (e.g. 1-4 or
8-12 hr) and possible missed sampling times as shown below.  How
would this be implemented using either CALL RANDOM statement or
verbatim code?  I can generate a time series easily outside
NONMEM (e.g. in SAS), but I would like to use NONMEM to produce
multiple replicate sets of random sampling times and utilize the
$SUPER statement to automate the process of simulation and
estimation within NONMEM, similar to the proposed control
stream submitted by Dr. Banken several years ago. 
(  Thanks.


1       0              0.02            .

1       4              2.6             .

1       8              .               .  <- ultimately, this line will need to be removed prior to simulating concentrations

1       24             23.8            .

2       0              0.53            .

2       4              1.2             .

2       8              8.5             .  

2       24             .               .  <- ultimately, this line will need to be removed prior to simulating concentrations


Max Tsai, Ph.D.
Senior Scientist (DM/PK)
Schering-Plough Corporation
2015 Galloping Hill Road
Kenilworth, NJ 07033-0530
(:  (908) 740-3911
:  (908) 740-2916

From:  "Bachman, William (MYD)"
Subject: RE: [NMusers] Random sampling times -> He Sun 
Date: Mon, March 14, 2005 8:55 am 

Dr. He Sun at the FDA did some work with simulation of sampling times several
years ago.  You might try contacting him or looking for his references.