___________________________________________________________________
| |
| YLO EXAMPLE |
|_________________________________________________________________|
In this example, blood concentrations below a known quantification
limit (BQL) have been eliminated from the analysis. A one-sided
interval with lower boundary equal to the quantification limit is
defined. The conditional likelihood for an observation is "condi-
tioned" on values of the population parameters and also on the condi-
tion that the observation is inside the one-sided interval.
(See pr_y, nmpr12, rocm44).
Here is an NM-TRAN control stream.
$PROB YLO example
$INPUT ID TIME DV AMT EVID MDV
$DATA ylow.dat IGNORE=C
$SUBROUTINES ADVAN=1 TRANS=2
$PK
CL=THETA(1)*EXP(ETA(1))
V=THETA(2)*EXP(ETA(2))
K=CL/V
S1=V
$ERROR
; Limit of Quantification (LOQ) is 1
; DV's below LOQ are given MDV=1 in dataset
; PR_Y (PRB) is probability that DV is > LOQ
; PR_Y (PRB) is provided even if MDV is 1
YLO=1
PRB=PR_Y
Y = F+ERR(1)
$THETA
(0,0.693); CL
(0,24); V
$OMEGA
0.09
0.09
$SIGMA
$ESTIMATION
MAXEVAL=9999
NSIG=4
METHOD=COND INTERACTION LAPLACIAN
PRINT=5
$TABLE ID TIME AMT MDV EVID PRB
NOPRINT
ONEHEADER
FILE=ylo.tab
Here is a segment of the dataset for a subject with last DV < LOQ
ID TIME DV AMT EVID MDV
. . . . . .
. . . . . .
. . . . . .
2 0 0 100 1 1
2 0.333 2.47 0 0 0
2 2 3.26 0 0 0
2 6 2.32 0 0 0
2 15.917 2.46 0 0 0
2 24 0 100 1 1
2 48 0 100 1 1
2 72 0 100 1 1
2 95.917 4.04 0 0 0
2 96 0 100 1 1
2 96.08 8.85 0 0 0
2 120 3.16 0 0 0
C Last observation record (below) has DV < LOQ (LOQ = 1) and MDV = 1
C This observation does not contribute to the -2LL estimate. However,
C probability that DV is > LOQ can be tabled using PRB=PR_Y.
2 168 0.163 0 0 1
REFERENCES: None.
Go to main index.
Created by nmhelp2html v. 1.0 written by Niclas Jonsson (Modified by AJB 5/2006,11/2007)