Can digoxin dose requirements be predicted?

A search for patient variables relevant to digoxin dose requirements was made in fourty‐three patients with a wide range of renal and hepatic function. The daily dose of digoxin to achieve a mean serum concentration of 1.5 ng/ml, the standardized dose, was calculated for each patient. The standardized dose correlated significantly with the following variables, in descending order of correlation coefficient; creatinine clearance, serum creatinine concentration, body weight and serum albumin concentration. An equation containing the two independent variables, creatinine clearance and serum albumin concentration, had a significantly stronger correlation with standardized dose than creatinine clearance alone. Attempts were made in each patient to predict the standardized dose using both empirical prescribing methods and the published nomograms. Although a maximum of 70% of the variance of the standardized dose was explained, this corresponded approximately to one patient in three having a predicted dose outside the 95% confidnece limits for the standardized dose. There remain important sources of individual variation in digoxin dose requirements yet to be identified. Future application of empirical prescribing methods, such as multiple linear regression and Bayes' theorem, to prescription for large, defined patient groups may improve dose prediction for individual patients.