This study will determine if a version of the bipolar care model adapted for a nonresearch audience can improve patient health and correct use of the model in community-based mental health care sites.

For the approximately 5.7 million Americans suffering from bipolar disorder, recurring manic and depressive symptoms, psychosis, and a high suicide risk prevent normal functioning and normal interpersonal relationships. Bipolar disorder has a high economic cost, as well, both because care for those with the disorder is expensive and because those suffering from the disorder have difficulty holding jobs. The total cost of bipolar disorder in the United States is estimated to be $45 billion. Part of this economic cost comes from poor quality of care. Only half of bipolar patients receive adequate outpatient care, and only a third receive adequate drug levels and safety monitoring. This can lead to preventable hospitalizations, emergency room use, and deaths.

Often, new results from research can take years or decades to be translated into practice in community mental health care. Even then, new methods can be less effective in practice because there is inadequate understanding of or technical support for these methods. This occurs because research is not presented in enough detail or with enough flexibility to be adapted for community settings.

The Centers for Disease Control and Prevention's (CDC's) Replicating Effective Programs (REP) project is a systematic method of replicating effective behavioral interventions for reducing HIV risk. The REP has three components:

• Packaging, or adapting, a behavioral treatment so its scientific protocols are in nontechnical, user-friendly language
• Providing formal training for health care professionals in implementing the package
• Supporting the use of the package through technical assistance available to individual treatment sites

The REP framework has been effective in extending the availability of HIV behavioral interventions and ensuring proper implementation, but it has never been used for non-HIV-related interventions, nor has it been proven that wider availability of these new interventions improves patient health and quality of life. This study will examine both of these untested areas of the REP framework by packaging a new intervention for bipolar disorder, called the Bipolar Care Model (BCM). Research shows the BCM to be effective in real-world settings and across different ethnic groups. This study will determine if a packaged version of the BCM will be effective across multiple, community-based, mental health settings. Success with the packaged version of the BCM will not only make this particular treatment more available to mental health care providers, but it will create a model for transitioning all research on new behavioral interventions for mental health into effective practice.

This study will last 2 years from its first implementation to the final collection of patient data at the participating mental health care sites. The care sites that will participate in this study will be randomly assigned to receive either the BCM package, training, and technical assistance or the BCM package alone. Adherence to the BCM model and patient outcomes at each care site will be measured at baseline and after 6, 12, and 24 months. Also at these intervals, patients at each care site will be contacted for a 30-minute survey on their treatment. Care sites will be assessed on the following: adherence to the BCM model, as assessed through site records; patient clinical outcomes, as assessed through questionnaires on bipolar symptoms, functioning, and quality of life; and costs, as assessed through records of training hours, technical assistance hours, and hours spent by REP implementers. The total cost of developing the intervention according to the REP framework will also be measured.

Inclusion Criteria:

• Currently being seen as a patient at the participating site
• Active diagnosis or treatment plan for Bipolar I, Bipolar II, or Bipolar not otherwise specified (NOS)

Exclusion Criteria:

• Serious illness that would prevent participation in the BCM components as indicated by the provider
• Living in a nursing home or other institution