|
The ability to precisely predict serum digoxin concentration using 7 published methods in a group of 50 patients was undertaken. Two methods of estimating creatinine clearance and two estimates of lean body weight were employed as input variables using the 7 dosing methods. TDX was used to determine the nadir SDCs(serum digoxin concentrations) in 50 in patients meeting predetermined study criteria. All patients, whose ages ranged 19-71 years, had steady-state digoxin levels, were in oral digoxin, and were free from liver dysfunction, thyroid dysfunction and renal failure. The correlation coefficients(r) of predicted versus observed SDCs were determined,. and mean error(ME) was determined for each method to reflect bias, respectively. No substantial differance in predictive reliabliity was evident among the methods studied in total group. Poor correlations existed between predicted and observed SDCs(r<0.4) and these correlations were not significantly affected by age and gender. But relatively higher correlation and lower ME was founded for the CHF group in Jelliffe method(r=0.5, p<0.05).
|