Re: questions about IOV

Hi Jia Two things: 1. Do you have any reason to think that the CL (or other parameters) are dose dependent e.g. saturable kinetics? You will need a column in the dataset (DS, for dose) that specifies the AMT in each record for an individual. In your parameter output table, you can include DS so that you can see whether the parameter values are correlated with the DS as well. If so, then you should put the dose in your model as a fixed effect, rather than a random effect. You can code for an effect of dose on parameters (PAR) as follows: TVPAR=THETA(1)*DSEFF ; Dose effect DSEFF=THETA(2)*DS ; This assumes a linear effect of the dose on the typical value of the parameter 2. IOV is a separate issue you can investigate regardless of your dose. I would try the above first and then add IOV to the model. You can code your IOV as follows: OCC1=0 OCC2=0 IF (TIME.LT.100) OCC1=1 IF(TIME.GE.100) OCC2=1 IIV=ETA(1) IOV=ETA(2)*OCC1 + ETA(3)*OCC2 TVPAR=THETA(1)*EXP(IIV+IOV) $OMEGA 0.1 ; IIV 0.1 ; IOV_OCC1 SAME ; IOV_OCC2 constrained to be the same as that for OCC1 ETA1 above will be your IIV whilst ETA2 and ETA3 will be your IOV.The parameters that should be estimated with IOV are the ones that you think can change between 2 different occasions (which may be most of your parameters). You need to investigate them one at a time. Further reading on IOV can be found in the paper below: Karlsson and Sheiner. The importance of modeling interoccasion variability in population pharmacokinetic analyses. Journal of pharmacokinetics and biopharmaceutics. 1993. vol 21, No. 6. Regards, Emmanuel Emmanuel Chigutsa (BPharm. Hons) Research Fellow, Pharmacometrics Group Division of Clinical Pharmacology, University of Cape Town K-45 Old Main Building, Groote Schuur Hospital Anzio Road, Observatory, 7925 Cape Town, South Africa Telephone: +27 214066758 Fax: +27 214066759 Mobile: +27 782826538 Email: [email protected] >>> Jia Ji <[email protected]> 11/11/2009 04:02 >>> Dear All, I am a new NONMEM user and now trying to model clnical PK data with a two-compartment model. More than half of patients in our trial had escalated dose in the second cycle. So I should have inter-occasion variability. But I got a couple of questions here. First, what parameters should be estimated with IOV? I have seen some models with IOV on CL, but not on others (maybe because I have seen too few models). Now I have IOV on CL, V1 and Q and it run successfully. But when I run with either two of them, I got minimization problem. But, am I having too many IOVs in the model? Second, I was wondering how to model IOV. Now I have the code as ; Define IOV, DESC = DOSE ESCALATION DESC=0 IF (TIME.GE.100) DESC=1 ETCL = ETA(1) + DESC*ETA(5) ETV1 = ETA(2) + DESC*ETA(6) ETQ = ETA(3) + DESC*ETA(7) ETV2 = ETA(4) So IOV is modeled as additive relationship to IIV. But what about multiplying relationship? Like ; Define IOV, DESC = DOSE ESCALATION DESC=0 IF (TIME.GE.100) DESC=1 ETCL = ETA(1)*(1+DESC*ETA(5)) ETV1 = ETA(2)*(1+DESC*ETA(6)) ETQ = ETA(3)*(1+DESC*ETA(7)) ETV2 = ETA(4) When I run with this multiplying relationship, I got increased OFV and minimization terminated due to rounding error. But I didn't understand why it is not working. Thank you so much for your patience and time! Jia ______________________________________________________________________________________________ UNIVERSITY OF CAPE TOWN This e-mail is subject to the UCT ICT policies and e-mail disclaimer published on our website at http://www.uct.ac.za/about/policies/emaildisclaimer/ or obtainable from +27 21 650 4500. This e-mail is intended only for the person(s) to whom it is addressed. If the e-mail has reached you in error, please notify the author. If you are not the intended recipient of the e-mail you may not use, disclose, copy, redirect or print the content. If this e-mail is not related to the business of UCT it is sent by the sender in the sender's individual capacity. _____________________________________________________________________________________________________