Prediction of lithium dose: a mathematical alternative to the test-dose method

A method of estimating the optimal dose of lithium is presented. The charts of 548 patients were reviewed to obtain data regarding the factors thought to affect the lithium dose, and an equation to estimate the dose was derived by stepwise multiple linear regression. The equation was also applied to 390 patients to determine the difference between the estimated and the actual dose; the mean difference was only 19 mg/day and the standard deviation was 325 mg/day. Lithium level, presence of a cyclic antidepressant, age, sex, and weight were found to be important variables for estimation of lithium dose.