Abstract

THE EFFECT OF SPARSE SAMPLING ON THE PREDICTION ERROR IN PHARMACOKINETICS OF INTRAVENOUS PARACETAMOL IN CHILDREN

Baba Sulemana Mohammed* and Gary Cameron

ABSTRACT

Aims: The aim of this study was to assess the effect of sparse sampling on the accuracy and precision in the estimation of Pharmacokinetics (PK) of intravenous (IV) paracetamol in children less than 15 years. Method: The clearance (CL) of IV paracetamol in 20 children was estimated first by a reference model using dense blood samples. Different sampling frames were designed to contain sparse samples and used to estimate clearances. Agreements between the clearance value from the reference model and that from each sparse sampling model obtained from each child were assessed using Bland-Altman plots. Results: Mean error was lower for reference sampling design. When one sample was taken at any time up to 1 hr, the inclusion of another taken after 1.4 x t ½, and a third at 6 hrs post dose (design 6), the error was lowest (0.016 L/h). The error was highest (0.48 L/h) when the last sample was taken 2 hrs (less than 1.4 times the estimated t1/2) post dose. The error in design 6 was in a wider range (-2.3 to 2.4), hence less precise estimations. Conclusions: Sampling density did not affect the size of the error in estimated clearance. The inclusion of samples at 6 hours post dose leads to accurate estimation of the CL of IV paracetamol in children than samples taken at 2 h; sampling around 2 h may be ignored in PK studies of IV paracetamol in children.

Keywords: sparse, paracetamol, children, pharmacokinetics, clearance.