Multiple treatments are available for nearly all the mood disorders. This r
ange of treatment options adds a new dimension of choice to clinical decisi
on making. In addition to prescribing the best initial treatment, clinician
s should have an algorithm for deciding if and when to make subsequent chan
ges in treatment to take advantage of second-line treatment options when ne
cessary. This article aims to 1) show that a wide variety of clinical decis
ions can be framed as choices among adaptive (within-patient) threshold-bas
ed strategies or algorithms, illustrating the generality of the concept; 2)
illustrate two ways to design randomized clinical trials to compare treatm
ent strategies with each other to decide which strategy is best; and 3) dis
cuss some of the advantages offered by these designs, in terms of both pati
ent acceptability and adherence to experimental protocols. (C) 2000 Society
of Biological Psychiatry.