Diagnostic accuracy of a hyperbolic model in predicting digoxin concentrations based on glomerular filtration rates.

ABSTRACT

BACKGROUND: Inappropriate doses and high serum concentrations of digoxin are highly prevalent in patients with renal impairment, and the drug dosage adjustment according to the glomerular filtration rate (GFR) is recommended. The aim of our study was to evaluate the dependence degree of digoxin total clearance (CL) on GFR, and the diagnostic efficiency of a predictive model for serum digoxin steady-state concentrations (Css) from estimated GFR by Cockcroft-Gault formula. METHODS: In 400 outpatients treated orally with digoxin, serum Css were determined (fluorescence polarization immunoassay from Abbott Laboratories), and total CL was calculated. The prediction of Css was carried out using a hyperbolic model developed by Konishi et al in Japan (J Clin Pharm Ther 2002;27257), and the constants of the equation were modified for a Caucasian population. RESULTS: Only 26% of the digoxin CL interindividual variability may be explained by differences in GFR, and this fact is a serious limitation for the derived predictive models. A 65% diagnostic efficiency was obtained for original and modified hyperbolic models in the correct classification of predicted Css as subtherapeutic, therapeutic or supratherapeutic with respect to obtained Css concentrations. CONCLUSIONS: The diagnostic efficiency obtained in the prediction of serum digoxin concentrations from estimated GFR values is unacceptable for the drug dosage adjustment in clinical practice.