Objective: Since sustained treatment-adherence is often problematic and may limit clinical outcomes among bipolar disorder (BPD) patients, we sought risk factors to guide clinical prediction of nonadherence.
Methods: Data were from a 2005 US national sample providing questionnaire responses by 131 randomly selected prescribing psychiatrists and their adult BPD patients. We contrasted demographic and clinical factors in treatment-adherent versus nonadherent patients (strictly defined as missing > or =1 dose within 10 days) in univariate analyses followed by multivariate logistic-regression modeling.
Results: Of 429 DSM-IV BPD patients (79% type-I; 62% women; 17% minorities), 34% reported missing > or = 1 dose of psychotropic medication within 10 days, 20% missed entire daily doses at least once, and only 2.5% missed all doses for 10 days. However, their prescribing psychiatrists considered only 6% as treatment-nonadherent. Factors significantly associated with nonadherence in multivariate modeling ranked: alcohol-dependence > youth > greater affective morbidity > various side effects > or = comorbid obsessive-compulsive disorder > or = recovering from mania-hypomania. Unrelated were sex, diagnostic subtype, and other comorbidities. Since most patients received > or = 2 psychotropics, potential relationships between treatment-complexity and adherence were obscured.
Conclusions: Prevalent treatment-nonadherence among American BPD patients, and striking underestimation of the problem by prescribing clinicians may encourage increasingly complex treatment-regimens of untested value, but added expense, risk of adverse effects, and uncertain impact on treatment-adherence itself.
(c) 2007 John Wiley & Sons, Ltd.