A review of lithium dosing methods

Lithium's narrow therapeutic range, concentration-effect relationship, and large differences in interindividual disposition make it an attractive candidate for predictive dosing methods. Lithium disposition is influenced by a number of factors, including patient characteristics, renal function, and concomitant drug therapies. These factors are accounted for, in varied degrees, in the different dosing methods available. One a priori method takes many of these factors into account and appears to have better reliability than others that ignore nonrenal influences on lithium disposition. The lack of data on pharmacokinetic values in different populations limits the usefulness of many of the single-point methods. Better data on the disposition of lithium in differing populations would greatly enhance the utility of both the a priori and single-point methods. Several multipoint methods, which estimate an individual's pharmacokinetic values, have also been described. These appear to be the most reliable, although they are time consuming and may delay the initiation of therapy. A well-designed Bayesian approach to the dosing of lithium, although not currently available, would be the most useful.