DETERMINATION OF PHENOBARBITONE POPULATION CLEARANCE VALUES FOR SOUTH-AFRICAN CHILDREN

Non-linear Mixed Effects Modelling (NONMEM) was used to estimate pheno barbitone population clearance values for South African children, usin g 52 serum levels gathered from 32 patients during their routine care. NONMEM was also used to evaluate the influence of fixed effects such as weight, age and concomitant medication. The final model describing phenobarbitone clearance was CL = [Exp(0.0288 Wt - 2.53)] M, where CL = clearance (1 . h(-1)), Exp the base of the natural logarithm, Wt = p atient weight (kg) and M = a scaling factor for concomitant medication with a value of I for patients on phenobarbitone monotherapy, 0.62 fo r those receiving concomitant valproate and 0.87 for those patients re ceiving concomitant carbamazepine or phenytoin. Mean (95% confidence i nterval) phenobarbitone clearance values were 7.6 ml . h(-1). kg(-1) 9 .0 ml . h(-1). kg(-1)) for the monotherapy 5.0 ml . h(-1). kg(-1) (4.0 , 6.0 ml . h(-1). kg(-1)) in the presence of concomitant valproate and 6.8 ml . h(-1). kg(-1) (5.6, 8.0 ml . h(-1). kg(-1) in the presence o f concomitant carbamazepine or phenytoin. These values are similar to those previously reported from both traditional and NONMEM pharmacokin etic studies.