Dr Alan Bye, Head of Clinical Pharmacokinetics Department
Clinical Pharmacokinetics Department, Glaxo Research and Development Limited, Greenford, UK
The general industry view of population analysis tends to be NONMEM and something which provides the means of extracting information from a collection of disjointed samples late in a development programme. Needless to say the results are often disappointing, because there is not enough thought given to the design, monitoring and bioanalysis of the samples before the pharmacokinetic analysis is carried out.
The logistic issues can be split into defining the need for population analysis (case by case), study design, monitoring, bioanalysis, datafeeds, database and analysis software. Because we are a new group in this area, we chose to build on existing systems and resources where possible. The most critical issue was getting to view the data in a “relational” way quickly and for this we initially chose SAS-lnsight. For modelling purposes, NONMEM was used as standard. Population modelling objectives are debated at the Clinical Development Plan stage. Currently there is potential for duplication, as population analysis is regarded as additional rather than primary analysis. The Regulatory guidelines in this area are still being established but we are working on the basis of what makes good science and participating in active debate within the regulatory groups of Europe.
Reference: PAGE 3 () Abstr 875 [www.page-meeting.org/?abstract=875]
Poster: oral presentation