From: Rebecca Wrishko <wrishko@unixg.ubc.ca>
Subject: IOV and Physiological Variables
Date: Thu, 24 Aug 2000 22:33:43 -0700

Dear NONMEM Users,

Based upon Karlsson MO, Sheiner LB. The Importance of Modeling
Interoccasion Variability in Population Pharmacokinetic Analyses. J Pharmacokinet Biopharm 1993; 21: 735-751, explicit modelling of IOV is encouraged when variability between occasions is likely. However, when this variation may be linked to physiological processes, such as age, serum creatinine or urine output as in neonates, is it possible to treat all data records for the individual as being contiguous with employing EVID=4 to reset the time and dose scale instead? Alternatively, should one consider each occasion as a separate individual, thereby assigning each occasion a different patient ID? Is there a method to determine if this latter analysis should be considered, that is when IOV>IIV?

Also, we are presently modeling vancomycin from a large neonatal dataset (gestational age 23-42 weeks) and have collected demographic, therapeutic, laboratory and physiological information preceding, during, and following serum drug concentration determinations. Often physiological and laboratory data are measured at particular times, and not daily, must these be entered for each event record or is it possible to use a blank for those times for which a physiological value are not reported. That is, if a serum creatinine value is reported to be 50 on the first day of vancomycin therapy but no further measurements were obtained, is it possible to enter this value at the initiation of therapy and leave that field blank thereafter, or must enter 50 for each event record and infer that no change is occurring? Similarly, if an APGAR score will be used in the evaluations, is a separate entry required for each event record or can one make one entry at the initiation of therapy? Further, are these covariates ignored until EVID=0, or are there continually assessed, even when EVID=1 or 2?

Thank you, in advance, for your assistance and insight into to what is probably common knowledge within the group.

Rebecca Wrishko M.Sc.
Ph.D. Candidate, Clinical Pharmacy
Faculty of Pharmaceutical Sciences
University of British Columbia
2146 East Mall
Vancouver, British Columbia
Canada, V6T 1Z3
Email: wrishko@unixg.ubc.ca