Psychotropic Medications for Patients With Bipolar Disorder in the United States: Polytherapy and Adherence

OBJECTIVE: Because treatments for bipolar disorder include a growing number of psychotropic agents, the authors evaluated psychotropic polytherapy and adherence to treatment among U.S. patients with bipolar disorder. METHODS: National health plan claims data (2000-2004) were used to identify patients diagnosed as having bipolar disorder who had continuous benefits and had not been prescribed medication for bipolar disorder for six months or more. The study compared drugs dispensed to these patients initially and at one year and characterized patients who were adherent to mood-stabilizers. RESULTS: A total of 7,406 patients had a bipolar disorder: bipolar I (55%), bipolar II (15%), or bipolar disorder not otherwise specified (30%). Women represented 57% of the sample; mean+/-SD age was 35.4+/-12.4 years. Initial prescription fills involved one psychotropic agent in 67% of patients, and two or more psychotropics (polytherapy) in 33%. Initial prescription drug selections involved: antidepressants > anticonvulsants >or= antipsychotics > sedatives > lithium; initial mood stabilizer use ranked: valproate > lithium > carbamazepine or oxcarbazepine > lamotrigine; antipsychotics ranked: olanzapine > quetiapine >or= risperidone > ziprasidone > aripiprazole > clozapine. Rankings were similar at one year, when only 31% of patients received monotherapy (a 2.2-fold decline), 32% received polytherapy, and 37% received no psychotropics. Initially patients received 1.42 psychotropic drugs per person; at one year, patients received 175, and at both times polytherapy was less likely with lithium than with anticonvulsants. In multivariate modeling, one-year mood stabilizer use was greater with the following: older age, type of mood stabilizer (lamotrigine > valproate > carbamazepine or oxcarbazepine > lithium) and was associated with more psychiatric office and emergency visits, clinician type (more common with psychiatrists than with primary care physicians), and nonuse of off-label anticonvulsants. CONCLUSIONS: Polytherapy was used by one-third of patients initially and at one year, antidepressant use was highly prevalent initially and later, but lack of treatment was prevalent at one year. Plausible clinical and treatment factors were associated with sustained mood stabilizer adherence.