RE: VPC for non-uniform sampling

Dear Norman, Ayyappa and NMusers, The situation described by Ayyappa does not necessary call for any out of the ordinary adjustments of a VPC. Especially so since it is stated that observations are made at random. If the design in any way was conditioned on the previous observations on the other hand the situation would be different. In the case when the design is conditioned on the observations e.g. TDM studies we have recently demonstrated that prediction corrected VPCs (pcVPCs) are a better diagnostic than traditional VPCs [1]. The pcVPC and pvcVPC (prediction and variability corrected VPC) have also been demonstrated to have advantages in the presence of other influential independent variables than the one depicted on the x-axis of the VPC (typically time). It can be worthwhile to once again point out that a VPC does not necessarily have to have time as the x-variable and DV as the y-variable. For example in the Ayyappa example time-after-dose might be an informative alternative as an x-variable and in some cases change from baseline can be an informative y-variable. Regarding the questions from Norman Z: Prediction and variability correction are new functionalities that is included in released versions of PsN. PsN also includes a lot of options for binning and further development in this field is on its way. An atomized binning algorithm was recently published by Lavielle et.al. [2] and implemented in MONOLIX. Atomized binning procedures will also be included in future PsN releases. Other rather recent functionalities in PsN with regards to VPCs is the ability to do VPCs for Kaplan-Meier plots (time-to-event), VPCs functionality for categorical data and handling of censored observations (BQL, dropout etc). As previously discussed at PAGE and NMusers “Standardized Visual Predictive Check” is the same as NPDEs previously developed by France Mentré. [1] Bergstrand M, Hooker AC, Wallin JE, Karlsson MO. Prediction-Corrected Visual Predictive Checks for Diagnosing Nonlinear Mixed-Effects Models. AAPS J. 2011 Feb 8. [2] Lavielle M, Bleakley K. Automatic data binning for improved visual diagnosis of pharmacometric models. J Pharmacokinet Pharmacodyn. 2011;38(6):861-71. More about the PsN VPC functionalities can be read in this document: http://psn.sourceforge.net/pdfdocs/npc_vpc_userguide.pdf Kind regards, Martin Bergstrand, PhD Pharmacometrics Research Group Dept of Pharmaceutical Biosciences Uppsala University Sweden