1038 — Medication Adherence, Ethnicity, and the Influence of Multiple Psychosocial and Financial Barriers in Veterans with Bipolar Disorder

Zeber JE (VERDICT) , Miller AL (UTHSCSA ), Valenstein M (SMITREC), McCarthy JF (SMITREC), Zivin K (SMITREC), Teh CF (UPMC / Pittsburgh VAMC), Cruz M (UPMC / Pittsburgh VAMC), Kilbourne AM (SMITREC)

Objectives:
Patients with bipolar disorder are often poorly medication-adherent, resulting in deteriorating symptomology, higher admission rates, and diminished quality of life. Many factors are strongly associated with adherence, including financial burdens and a variety of psychosocial factors. However, analyses typically consider potential barriers independently rather than conjointly from the patient’s perspective. Such approaches neglect the complex interplay of risk factors, many of which are amenable to health policy or clinical interventions. This study uses self-reported data to evaluate the differential and cumulative impact of nine barriers upon medication adherence.

Methods:
We recruited 435 patients from the Continuous Improvement for Veterans in Care - Mood Disorders (CIVIC-MD, FY04-06) study which examined quality of care provided to veterans with bipolar disorder. Surveys collected information on multiple financial and psychosocial adherence barriers: medication copayments, foregoing treatment due to cost, binge drinking, access difficulty, social support problems, poor therapeutic alliance, and low medication insight. Medication adherence was measured by the validated Morisky scale. Multivariable logistic regression modeled adherence as a function of perceived barriers upon adherence, controlling for demographics, homelessness, and affective symptomology.

Results:
Nearly half of the respondents reported adherence difficulty. Patients experienced an average of 2.8 barriers, with 41 percent perceiving at least three. Minority veterans reported poorer adherence than white patients (56 percent versus 40 percent, p=.01), while claiming more overall barriers, particularly financial burden, binge drinking, and difficulty obtaining psychiatric care when needed. Multivariable models revealed that the total number of barriers was significantly associated with poor adherence (OR=1.24 per barrier). The most significant were low medication insight, binge drinking, and difficulty accessing psychiatric care (ORs of 2.41, 1.95 and 1.73, respectively).

Implications:
Veterans with bipolar disorder experience multiple barriers to medication adherence, a scenario possibly exacerbated by recent copayment increases. Besides the total number of barriers, certain psychosocial and financial obstacles proved especially pernicious in connection with worse adherence.

Impacts:
Recognizing multiple barriers can assist in developing tailored clinical interventions to improve poor adherence by tailoring efforts towards reducing psychosocial risk factors. Furthermore, the interaction with VA health benefit policies potentially contributes to burdens faced by vulnerable veterans already experiencing adherence problems.