RE: interoccasion variability

From: "Mats Karlsson" Subject: RE: [NMusers] interoccasion variability Date: Mon, 6 Feb 2006 20:52:14 +0100 Dear Ann and Serge, It is true that interoccasion variability is introduced to handle parameter variability within a subject over time. Thus the lower limit of what may constitute an occasion is the interval necessary in order to obtain an estimate of the parameter. Thus, every dosing interval may constitute an occasion. For example for a parameter like bioavailability, this is often the most reasonable definition. To allow separation between interoccasion and residual variability it is necessary to have at least two observations per occasion for at least some of the occasions. Thus, regarding Ann's original question, I think it is reasonable to treat each day as a separate occasion (at least is the half-life is below about 8 hours). However, I would be careful and inspect the result to make sure that you are not treating what are systematic changes with a random effects model. If some patient has deteriorating eliminating organ function over time and some other patient on the other hand improves consistently over time, interoccasion variability is not the right model (but it will lower the OFV compared to a model with no time-variation). Best regards, Mats -- Mats Karlsson, PhD Professor of Pharmacometrics Div. of Pharmacokinetics and Drug Therapy Dept. of Pharmaceutical Biosciences Faculty of Pharmacy Uppsala University Box 591 SE-751 24 Uppsala Sweden phone +46 18 471 4105 fax +46 18 471 4003 mats.karlsson@farmbio.uu.se