Dear SoJeong Yi,
Sometime we cannot estimate the variances for IIV and IOV separately, although we have multiple dosing data within each subject. At that time, the alternative could be to combine the IIV and IOV to a single random effect parameter, which is the way you have done. In this case, the random effect parameter can be described as “random effect parameter reflecting both IIV and IOV”, and so on. Best regards, ============================================================================================= Hyeong-Seok Lim MD PhD Associate Professor Department of Clinical Pharmacology and Therapeutics, Asan Medical Center, University of Ulsan 88, Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, Republic of Korea Tel: +82-2-3010-4613 Fax: +82-2-3010-4623 LinkedIn: http://kr.linkedin.com/pub/hyeong-seok-lim/28/926/848 Email: [email protected] <mailto:[email protected]> , [email protected] <mailto:[email protected]> ============================================================================================= From: [email protected] [mailto:[email protected]] On Behalf Of 이소정 Sent: Friday, September 4, 2015 11:13 AM To: [email protected] Subject: [NMusers] Describing Omega that includes both BSV and IOV Dear all, Currently I’m summarizing the NONMEM estimates of population PK for writing a manuscript. However, I wonder how to describe the omega which includes both between-subject variability and inter-occasional variability. The code of ‘variability’ is followed below, $PK IF(OCC.EQ.1) IOV = ETA(8) IF(OCC.EQ.2) IOV = ETA(9) IF(OCC.EQ.3) IOV = ETA(10) …. KA = THETA(9) * EXP(IOV) (--> In final model, PK parameter was estimated like this) ; KA = THETA(9) * EXP(ETA(4)+IOV) (--> When I used this code, there were some problems (boundary error, large RSE (>80%), very small estimate of BSV and so on) ) ; KA = THETA(9) * EXP(ETA(4)) (--> when I used BSV only, the OFV is quite higher than upper two cases, so I thought that IOV should be considered. ) $OMEGA BLOCK(1) SAME $OMEGA BLOCK(1) SAME $OMEGA BLOCK(1) 0.3 In this case, the individual eta was re-calculated in one subject according to occasion, isn’t it? Then, how should I describe this ‘variability’ and estimate of omega in a manuscript? (i.e., between subject variability containing inter-occasional variability? Or any other appropriate term?) I will appreciate if someone give any advice. Thanks in advance. Best regards, SoJeong Yi SoJeong Yi, Ph.D Department of Clinical Pharmacology and Therapeutics, Seoul National University College of Medicine and Hospital 101 Daehak-ro, Jongno-gu, Seoul 110-744, Korea Tel: 82-2-740-8291 Fax: 82-2-742-9252 C.P: 82-10-3178-4133 E-mail: <mailto:[email protected]> [email protected]
