Evaluation of a morphine maturation model for the prediction of morphine clearance in children: how accurate is the predictive performance of the model?

Abstract

The pharmacokinetics of morphine are well known in pre-term and term neonates, infants, younger and older children, as well as in adults. There are circumstances when a pharmacokinetic study may not be possible in children (especially neonates and infants), and as a result one would like to predict drug clearance in children. Several methods, such as allometric scaling and prediction based on incorporation of physiological parameters, have been suggested. Recently a morphine maturation model has been proposed to predict morphine clearance in the paediatric population. The current study evaluates the predictive performance of the morphine maturation model for the prediction of morphine clearance in pre-term, term, infants and children up to 5 years of age. The results of the study highlight the shortcomings of the model to predict morphine clearance in children in aforementioned age groups. Recently, a maturation model that incorporates a sigmoidal E(max) type model has been proposed for the estimation of morphine clearance in paediatric patients. The primary objective of this report is to evaluate the predictive performance of the morphine maturation model for the prediction of morphine clearance in children of different ages. The secondary objective of this report is to evaluate the predictive performance of exponent 0.75 on bodyweight in the absence of the sigmoidal part of the morphine maturation model. In order to evaluate the predictive performance of the morphine maturation model, the clearance values of morphine for individual children (preterm neonates to 5-year-old children; n= 147) were obtained from the literature. The predicted clearance of morphine in an individual child, obtained from the maturation model as well as from the fixed exponent 0.75 was compared with the observed clearance in that individual child. The morphine maturation model's predictive power in neonates, infants and younger children is poor and the inclusion of the sigmoidal part in the model only helps in reducing the substantial error introduced in the prediction due to the application of exponent 0.75 on bodyweight. Furthermore, the real benefit of the sigmoidal E(max) part of the model disappears by 1 year of age. The morphine maturation model has a poor predictive power of morphine clearance in preterm and term neonates, infants and very young children and may not be of any practical value for the prediction of morphine clearance in this age group.