D. Santos Buelga, G. Frutos, MJ Garcia, MJ Otero, A. Dominguez-Gil
Non linear Mixed Effects Modelling (NONMEM) was used to estimate Theophylline population pharmacokinetic parameters values in a group of adult patients, using 454 serum levels gathered from 192 patients during their routine care. The patients were treated with aminophylline (Eufilina®), administered by oral or IV infusion. Serum theophylline levels were measured by routine hospital laboratory using the fluorescence polarisation immunoassay. The concentration data were analized according to a one-compartment model with two parameters: clearance (CL) and distribution volume (Vd). The influence of body weight (total, lean and ideal), age, smoke habit, congestive failure hearth was investigate. In first step a population pharmakcokinetic analysis was performed without including any covariates in the model (basic model). Estimates of individual parameter values od CL and Vd were obtained as empirical bayesian estimates (POSTHOC option). In a second step the individual parmeters estimated were regressed on the covariates using an additive model. Two analysis of variance and covariance were performed inorder to increase the precision of comparison between treatments by reducing the within-group parameters in the response variable due to influence of covariates and by adjusting the effects of any imbalances betwenn groups. Finally NONMEM was used to optimize the population model.
Reference: PAGE 7 (1998) Abstr 284 [www.page-meeting.org/?abstract=284]
Poster: poster