2009
DOI: 10.1007/s10928-009-9115-y
|View full text |Cite
|
Sign up to set email alerts
|

Informative study designs to identify true parameter–covariate relationships

Abstract: This study explored how study design influences the probability of selecting a 'true' covariate from two competing covariate models. The probability of selecting the 'True Model' (lean body weight on clearance) over the 'False Model' (total body weight (WT) on clearance) was compared for designs where WT was either lognormally distributed (i.e. non-stratified), or stratified into 3 equal strata. The probability of selecting the 'True Model' increased as the WT inclusion criterion widened, and was always greate… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
11
0

Year Published

2011
2011
2016
2016

Publication Types

Select...
7
1

Relationship

2
6

Authors

Journals

citations
Cited by 17 publications
(12 citation statements)
references
References 23 publications
0
11
0
Order By: Relevance
“…This result is in contrast to Svensson et al, who found an improvement in model fit when allometric scaling of disposition parameters with fixed coefficients were included (0.75 for clearances, 1 for volumes) (7). Our finding was not surprising, however, given the weak relationship observed between the eta for CL/F and size covariates, and a nonstratified distribution of weight, which can result in a lower probability of identifying a true covariate effect (18). It is likely in this case that the underlying covariate effect of weight was instead partially described by the lower CL/F observed for sicker MDR-TB patients than that for DS-TB patients and healthy volunteers.…”
Section: Discussionmentioning
confidence: 53%
“…This result is in contrast to Svensson et al, who found an improvement in model fit when allometric scaling of disposition parameters with fixed coefficients were included (0.75 for clearances, 1 for volumes) (7). Our finding was not surprising, however, given the weak relationship observed between the eta for CL/F and size covariates, and a nonstratified distribution of weight, which can result in a lower probability of identifying a true covariate effect (18). It is likely in this case that the underlying covariate effect of weight was instead partially described by the lower CL/F observed for sicker MDR-TB patients than that for DS-TB patients and healthy volunteers.…”
Section: Discussionmentioning
confidence: 53%
“…Hence, a pharmacokinetic modeler may easily conclude that a simple linear relationship exists between weight and clearance when the weight distribution of the study population is narrow. More detailed simulations that corroborate this simplistic example are provided in an excellent article …”
Section: Allometry and Drug Dosingmentioning
confidence: 64%
“…The inclusion of this covariate in the present model is likely due to the wide, stratified distribution of body size descriptors in our study. Such a design has been suggested to increase the power to detect true covariate relationships (32).…”
Section: Discussionmentioning
confidence: 99%