Quality of care in a Medicaid population with bipolar I disorder.

OBJECTIVE: This study examined whether presenting diagnosis and treatment in intensive settings (hospitalization, partial hospitalization, or residential programs) are correlated with the subsequent treatment of bipolar I disorder.
METHODS: Claims data were studied retrospectively (fiscal years 1994-2000) for 2,644 patients with bipolar I disorder who had been enrolled in Medicaid at least six months before their first observed bipolar diagnosis. Logistic regression models estimated the association between the presenting diagnosis and initial treatment setting and the subsequent treatment up to one year after the first observed bipolar diagnosis. Measures included receipt of guideline-recommended care (antimanic agent plus psychotherapy) or care discouraged by guidelines (an antidepressant without an antimanic agent).
RESULTS: Only one-third of enrollees received both guideline-recommended treatments after the first observed bipolar diagnosis. Patients were less likely to receive both recommended treatments if the first observed mental health service occurred in an intensive setting. Enrollees presenting with a bipolar diagnosis were less likely to receive psychotherapy, whereas rates of antimanic medication use were similar to those with other presenting diagnoses. Presenting with depression or anxiety or other, nonbipolar diagnoses was associated with a higher likelihood of receiving pharmacotherapy discouraged by guidelines.
CONCLUSIONS: This study raises general concerns for the treatment quality of bipolar I disorder in this medically complicated, largely disabled Medicaid population. Also, how bipolar I patients enter treatment can be associated with subsequent differences in treatment quality--information that can be useful to clinicians and policy makers when planning quality improvements to treatment programs.