Re-examination of digoxin dosage regimen: comparison of the proposed nomograms or formulae in elderly patients.

1 Comparison of measured trough serum digoxin concentrations (SDC) with those predicted at steady-state by forty sets of previously described pharmacokinetic equations in conjunction with either measured or estimated creatinine clearance (Clcr) was made in an elderly group of patients with varying Clcr (25-89 ml/min) who received maintenance digoxin therapy. 2 Mean values of day-to-day variation (% coefficient of variation) for serum creatine and measured Clcr examined three times within a week in 7 of 40 elderly patients were 6.42 and 8.36% respectively, while this value for SDC was 12.46%. 3 There was a statistically significant correlation (r = 0.72 to 0.79, P less than 0.001, n = 40) between measured Clcr and that estimated from three different methods. Siersbaek-Nielsen's nomogram approach gave the best correlation. 4 The calculated correlation coefficient values between measured and predicted SDC lay within a relatively narrow range but were weak (r = 0.443-0.628). Prediction error analysis, however, allowed us to determine the best set of pharmacokinetic equations for a first approximation to appropriate digoxin maintenance dosage requirements, namely Gault-estimated elimination rate constant, measured Clcr, and Jusko-estimated volume of distribution.