[NMusers] PRED for BLQ-like observations

From: Pavel Belo <nonmem_at_optonline.net>
Date: Fri, 20 Nov 2015 11:47:21 -0500 (EST)

Hello The NONMEM Users,
When we use M3-like approach, the outputs has PRED for non-missing
observations and something else for BLQ (is that PRED=CUMD?).  As in=
diagnostic figures PRED for BLQs looks like noise, I remove them.  It =
not always perfect, but OK in for most frequent cases.  
When we use count data such as a scale with few possible values (for=
example,  0, 1, 2, 3, 4, 5), it makes more sense to use PHI funct=
(home-made likelihood) for all observations rather than to treat the=
count as a continuous variable an apply M3-like approach to 1 and=
5 while only (as we know, they are like LLOQ and ULOQ).  In this =
all PRED values look like noise.  A hard way to replace the noise=
PRED value is to simulate PRED for each point and merge them with=
 the DV
and IPRED data. Is there an easy way? 
(The model runs well and better than when the count is treated as=
continuous variable.)

Received on Fri Nov 20 2015 - 11:47:21 EST

This archive was generated by hypermail 2.3.0 : Fri Sep 27 2019 - 16:47:16 EDT