Re: When to study BOV?

From:Re: [NMusers] When to study BOV? Subject:Nick Holford Date:Tue, 03 Dec 2002 08:02:21 +1300 Atul, IMHO this is an issue of principle rather than experience. The total Population Parameter Variability (PPV) can be partitioned into Between Subject Variability (BSV) and Within Subject Variability (WSV). Using occasion as a covariate allows one to distinguish between BSV and WSV. Note that WSV consists of BOV plus Within Occasion Variability (WOV). Using occasion to identify BOV means that the remaining estimate of variability is really BSV+WOV. In an analysis of aminoglycoside pk with a 2 cpt model the BOV in Clearance and V2 was able to be estimated. I am sure there must be BOV in V1 and Q but attempts to estimate these parameters failed to converge. The addition of covariates to the model reduced BOV a little but did not seem to have any important impact. While overall covariates predicted 56% of PPV in CL the BOV in these covariates (weight and renal function) was presumably not large. If you can estimate BOV then you should. I would put it in the model at the outset before adding other covariates which attempt to exlain the fixed (as opposed to random) components of PPV. Nick -- Nick Holford, Divn Pharmacology & Clinical Pharmacology University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand email:n.holford@auckland.ac.nz tel:+64(9)373-7599x6730 fax:373-7556 http://www.health.auckland.ac.nz/pharmacology/staff/nholford/ _________________________________________