The January 26 newsletter focused on integrated care and identified some of the barriers associated with it.  This issue highlights an example of how Duke University clinicians are implementing the collaborative care IMPACT model, in which primary care physicians are screening for depression and referring consumers for more specialized care.

Transforming Depression Care Management in Duke Health: The Collaborative Care Approach
Primary care providers increasingly serve as the de facto mental health system in the US. Patient preference, stigma, insurance discrimination, and the shortage of psychiatrists have all contributed to this shift in the locus of care. As a result, the majority of prescriptions for psychotropic medications are written in primary care settings, yet resource constraints, lack of time, and gaps in training all limit the effectiveness of behavioral health care in primary care. Efforts to integrate mental health treatment in primary care, while garnering significant attention, have been less than successful and can be linked to inadequate use of evidence-based models of behavioral health-primary care integration. Several decades of implementation research, including seminal work on the IMPACT model at Duke and elsewhere, demonstrates that well-implemented Collaborative Care (CC) models, such as the IMPACT model, do achieve treatment targets, improve patient outcomes, enhance patient satisfaction, and reduce total health care costs.

For several decades, researchers have documented that up to 12% of primary care patients meet criteria for major depression, have high rates of co-morbid medical illness, and up to twofold higher health care utilization and costs. Indeed, in studies of Medicaid populations, approximately 2/3 of patients with a chronic medical condition also have a co-morbid behavioral health condition, which accounts for a doubling or more of annual health care costs per patient. Studies have also documented that fewer than half of patients with depression are accurately diagnosed, receive adequate dosage and duration of pharmacotherapy, or effective psychotherapy. Depression CC trials were developed in the 1990s to incorporate many elements of the Chronic Care Model, addressing the elements of the organization of care needed to improve outcomes for patients with chronic illness. Duke was a study site for the landmark IMPACT (Improving Mood-Promoting Access to Collaborative Treatment) study addressing depression in the elderly. The IMPACT intervention demonstrated significantly higher effectiveness than usual care in reducing depressive symptoms, improving functioning and quality of life, and reducing total health care cost by approximately $3000 per patient over a 4-year period. Subsequent adaptations of CC models have been applied to a variety of psychiatric conditions, varied care settings, payers, and populations. Recent meta-analyses and systematic reviews of over 70 CC RCTs found strong evidence of the effectiveness of CC in reducing depression symptoms and achieving remission; improved quality of life and functional status; and increased satisfaction with care. Similar results are emerging for multi-condition management of psychiatric conditions co-morbid with other chronic illnesses, such as depression with diabetes and/or cardiovascular disease. A 24-month cost analysis of multi-condition CC found that compared to usual care patients, intervention patients had 114 additional depression-free days and lower mean outpatient health costs of $594 per patient.

These successful CC interventions prototypically utilize 3 main elements: 1) systematic psychiatric assessment employing well-validated measurement-based assessment tools such as the Patient Health Questionnaire-2/9 (PHQ-2/9) for depression measurement; 2) use of non-physician care managers to provide monitoring, care coordination and protocol-based counseling and; 3) stepped-care recommendations to primary care providers from a CC team psychiatrist. The psychiatrist routinely reviews serial, measurement-based assessments with care managers, makes treatment recommendations per a structured treatment algorithm, and focuses on treating to established targets. Care managers follow up positively-screened patients and, if requested by the provider, enroll them in the CC program. In addition, they also provide structured assessments over time and coordinate treatment recommendations with primary care providers. Care managers also coordinate psychiatric referrals as needed for more specialized care.

Beginning with an institutional award from the Duke Institute for Health Innovation, collaborators at Duke’s Department of Psychiatry and Behavioral Sciences and Duke Primary Care (DPC) have made steady progress in implementing this model of depression care management across DPC, primary care practices operated by Duke Health. The process included developing an effective workflow in these practices, integrating depression care managers and consulting psychiatrists, and creating an efficient documentation and consultation process within Duke’s electronic health system. The program is soon to reach 19 DPC clinics with a plan to reach all of these clinics in the coming year. The psychiatrists leading the project include Drs. Virginia O’Brien and Marvin Swartz in collaboration with DPC leadership including Drs. John Anderson, Edward Cooner, and Pat Johnson, from DPC Nursing.


Commentary by Marvin Swartz , MD