2.4 Study Comparison
We developed virtual patients with different age: infants (10 kg, 1-year-old) taking 300 mg OXC every 12 h (q12h), children (30 kg, 10 years old) taking 300 mg OXC q12h, and adults (70 kg, 40 years old) taking 600 mg OXC q12h. The concentration–time profiles in virtual patients were drew on the basis of the restored PPK model and study population in each included study. They received multiple doses of OXC alone and achieve stable state. The sex of the virtual patient was selected as male.
The identified covariates effect on clearance (CL) in the included studies was analysed and showed by a forest map. The CL change less than between 80%-125% was not considered to have a significant clinical correlation [19]. For binary covariates such as co-administration with medications, we use 0 for monotherapy and 1 for co-administration with medications. For continuous covariates that were included in only one model, we used the same range values as the study. For continuous covariates included in multiple studies, we scaled them to the same range for comparison. The effect of each covariate on CL was shown as the ratio of CL in the range of the covariate divided by the typical CL value in each study.
The simulation was conducted using NONMEM (version 7.4; ICON Development Solutions, Ellicott City, MD, USA) software. R (version 3.5.1; http://www.r-project.org/) was used to generate the concentration–time plots and forest plots.