VALIDATION OF A QUICK MODELING PROGRAM GENERATING CLEARANCE ESTIMATESAT STEADY-STATE FOR ROUTINE THERAPEUTIC DRUG-MONITORING

Therapeutic drug monitoring (TDM) of chronic treatments is justified f or several reasons, including relative over- or underdosage due to var iable individual elimination, pharmacokinetic interactions in drug com binations, and noncompliance. In all these circumstances, the prescrib ing physician is interested in having an estimation of the patient's c learance of the drug, even from one measurement. We compare a validate d bayesian program, USCPack of Jelliffe, found difficult to use in da ily routine, with a ''home-made'' program. The latter, which is capabl e of taking data from a clinical database, will generate a graphic sim ulation of daily plasma drug concentrations together with an estimatio n of steady-state clearance more rapidly than does USCPack. Both prog rams were run with only one measured plasma level. The patients were 8 3 children or young adults treated with phenobarbital (PB), carbamazep ine (CBZ), and/or Valproic acid (VPA) who were resistant to monotherap y and who were to be sampled two to four times between doses. Drugs we re routinely assayed by high-performance liquid chromatography (HPLC). Despite the rough character of Phacile (numeric integration and adjus tment of only two of three parameters, without an acknowledged minimiz ation algorithm), the results are comparable to those obtained with US CPack for estimating clearance and predicting plasma drug concentrati ons. Phaciie algorithm, although simple, has proven of interest in rou tine TDM and as an introduction for medical students to the bayesian a pproach of population pharmacokinetics.