Creating NONMEM datasets — how to escape the nightmare

Abstract

Pharmacokinetics (PK) is the impact of the body on the drug and pharmacodynamics (PD) is the impact of the drug on the body. The effect of the drug on the target population of patients is predicted using models during population PK/PD analysis. This type of analysis is becoming more and more common in clinical trials, but creating the dataset structure required by the widely used non-linear mixed effects modelling software called NONMEM H is often the nightmare part of the process. It usually takes months to prepare the NONMEM dataset before the pharmacokineticist feels that it is ready for them to use. It is also produced after unblinding, adding to the delay in finalising the dataset before analysis can be performed with it. This delay can lead to holding back decisions about future trials, or those decisions are then made without taking into account the population PK analysis report. This paper will look at the issues which cause problems when creating a NONMEM dataset, and what steps we can take to avoid these problems and minimize the time taken to create the final dataset.