skip navigation

OffLine Winter 2000Spring 2000

Volume 12 Number 3

The adverse impact of depression on personal well-being, work productivity and health care costs has been well established. Center for Health Studies (CHS) and University of Washington (UW) researchers have collaborated for the last decade on improving the care for depression, specifically in the primary care setting. Why the emphasis on primary care? Six to ten percent of primary care patients suffer from depression, making it one of the most common conditions seen by primary care providers (PCPs). Since most patients turn to PCPs when seeking help for a mental health problem, these providers are in a unique position to affect the course and progression of depressive illness.

On the positive side, PCPs can recommend treatments that have been shown to be effective in treating depression, such as antidepressant medications and psychotherapy. On the not so positive side, the management of mental health problems in primary care is often disorganized, with many patients coming in for initial treatment but never returning. A CHS study found that, contrary to guideline recommendations, one-half of Group Health patients seen by a PCP for a new episode of depression did not make a follow-up visit within 2 months; 40 percent who were prescribed antidepressants did not refill their medications in a timely manner.

These types of problems are common to the treatment of many chronic conditions, of which depression is just one example. However, disorganized care may represent a more significant barrier to patients with depression because of the sense of hopelessness that often characterizes this condition. Recognizing this situation, Group Health has designated depression as one of its eight "Clinical Roadmaps." The Clinical Roadmap Framework, designed to facilitate systematic care for chronic conditions, embraces key disease management strategies such as: 1) population registries; 2) evidence-based guidelines; 3) practice reorganization; 4) patient self-management support; and 5) monitoring process of care and outcomes.

"Collaborative Care," a depression care model pioneered by CHS/UW scientists and adopted by health care organizations and researchers across the country (see "Dissemination") embodies these principles. At the core of the "Collaborative Care" model is a practice redesign featuring the co-management of patients by the PCP and on-site mental health specialists. The multi-faceted program also includes physician training, patient education and monitoring processes of care, such as medication adherence. Two randomized trials conducted in the early 1990s found that this Collaborative Care model, featuring co-management of depressed patients by PCPs and psychiatrists (Trial 1) or psychologists (Trial 2), improved depressive outcomes, increased satisfaction with care and increased medication adherence.

A recently completed randomized trial extended these earlier efforts. Recognizing that a portion of patients with depression will improve with Usual Care, researchers recruited patients who were still experiencing significant depressive symptoms 6 to 8 weeks after initiation of antidepressant treatment. This scenario meant that the added resources of the Collaborative Care model were directed at those patients with the greatest need, that is, those patients who were not better after 2 months of usual depression care. Would the Collaborative Care model be successful among more "treatment resistant" cases? The Collaborative Care model in this trial consisted of Usual Care supplemented by 2 to 5 visits with a study psychiatrist over an initial three-month period. Other intervention components included physician training, systematic patient education and prospective monitoring of follow-up visit frequency and medication adherence. Similar to the earlier trials, scientists found that among patients with persistent depression, the Collaborative Care model succeeded in enhancing patient satisfaction, increasing medication adherence and improving depressive outcomes relative to Usual Care.

The researchers also found that the Collaborative Care intervention reduced depression-related disability compared to Usual Care. This is significant because people with depression often limit their work, family and social activities as much as, or more than, people who suffer from chronic physical diseases.

Reduction of disability was also one of the goals of a randomized trial conducted among high utilizers of health care who suffered from depression. This trial took place at three health care systems nationwide, including Group Health under the direction of CHS investigator Greg Simon. Previous CHS research has found that health care costs for patients with depression were twice those of age-gender matched patients who were not depressed. Most of the "excess" cost associated with depression was due to generalized increases in health care use, not to outpatient or inpatient mental health treatment. Thus, the thinking behind the multi-site trial of depressed high utilizers was that effective treatment of depressive illness would not only lead to improved depressive outcomes, including a reduction in depression-related disability, but also to a decrease in overall health care utilization.

The trial screened patients in the top 15 percent of outpatient utilization to identify those most likely to benefit from depression treatment. The intervention consisted of physician and patient education and antidepressant treatment initiated by the PCP. Systematic phone monitoring of patients’ treatment adherence and outcomes by a depression case manager was another integral part of treatment. The intervention differed from the Collaborative Care Model described above in that depression care resided with the PCP and case manager—mental health specialists were consulted only as needed (approximately 10 percent of the cases). After one year, intervention patients experienced less disability, fewer depressive symptoms, and better medication adherence compared to Usual Care controls. Health care utilization was not reduced; but the incremental costs of improving depression care were modest, about $550 per patient per year. The researchers conclude that "real world" PCPs can deliver effective depression treatment if they are supported by systems incorporating patient and physician education, proactive psychiatric consults for treatment non-responders and patient monitoring.

Simon focused on the last component—patient monitoring—in a recently completed randomized trial. He tested whether depression management could be improved with relatively simple, low-cost interventions designed to increase treatment adherence and systematic follow-up care. More than 600 Group Health enrollees initiating antidepressant treatment for depression were randomized to Usual Care or to one of two intervention programs. The Feedback Only program used computerized pharmacy and visit data to provide feedback and algorithm-based recommendations to physicians. The Feedback plus Care Management program also included systematic telephone follow-up of patients, more sophisticated treatment recommendations and practice support by a dedicated case manager. The case managers assessed medication adherence and symptom levels and assisted in scheduling follow-up appointments. No counseling was provided. While the computerized feedback alone did not yield significant benefits, the feedback in combination with systematic telephone follow-up and case management improved depressive outcomes and increased medication adherence, but did not impact follow-up visit frequency. Incremental cost was approximately $80 per patient. While cautioning that this approach is not a panacea, Simon concludes that the results support the implementation of organized monitoring and care management programs to improve management of depression.

Relapse Prevention: Given the high health care costs associated with depression, and the costs associated with treatment, ample incentive exists to keep people well once they have recovered. Unfortunately, this does not always happen. Sixty percent of primary care patients diagnosed with depression have had two or more prior lifetime episodes. A CHS study found that 36 percent of depressed patients in primary care whose symptoms had resolved experienced a recurrence of major depression within one year. Depression treatment has historically focused on the crucial first few weeks. However, improvements during this acute phase are often not sustained over the long run, leading experts to conclude that a sub-set of patients need longer-term support, or "maintenance" therapy. To test this theory, CHS and UW researchers undertook a novel study designed to prevent relapse into depression among patients recently recovered from a depressive episode.

Study subjects were patients who were diagnosed with depression and started on antidepressants, but who had essentially recovered after 2 months. Although their symptoms had basically resolved, they were at high risk for relapse due to the presence of risk factors such as multiple prior lifetime depressive episodes. Patients randomized to the intervention had a sustained relationship over a one-year period with a Depression Prevention Specialist (DPS) working in tandem with the PCP, under the supervision of a consulting psychiatrist. Intervention patients developed their own relapse prevention plans which identified high-risk situations and early signs of relapse, stress management techniques and coping strategies. A key part of the intervention focused on shared-decision making around the use of antidepressants. While maintenance use of medications has been shown to help prevent relapse, some patients resist this strategy. To help patients reach decisions about antidepressant use, the DPS used "motivational interviewing", a facilitative, non-confrontational, counseling style designed to help patients explore and resolve ambivalence.

Motivational interviewing and other intervention components led to increased adherence to medications among intervention patients compared to Usual Care patients up to one-year post-randomization. According to automated data, 63 percent of intervention patients (vs. 49 percent of controls) were taking medications one-year post-randomization. While intervention patients reported lower overall levels of depressive symptoms on one of two clinical outcome measures, they were not less likely to experience a recurrence of depression in the one year following entry into the study.

Dissemination: With its work cited in the U.S. Surgeon General’s Report on Mental Health, the CHS/UW depression team’s work is beginning to have an impact nationally and internationally. There are several examples of this "research being translated into practice." Group Health’s Depression Roadmap has incorporated several components of interventions tested over the years: on-site behavioral health specialists, proactive monitoring and feedback. Henry Chung, medical director of a community health center in Chinatown, New York, serving low income, medically underserved Asian Americans, calls the work of the Seattle team an "inspiration." Chung has implemented core ideas from the Collaborative Care model including on-site liaison teams, assertive follow-up and outreach.

Another "real world" application of this research comes through the Breakthrough Series to Improve Primary Care for Depression, part of the Robert Wood Johnson National Program "Improving Chronic Illness Care." Sponsored by the MacColl Institute, CHS depression researchers Evette Ludman and Michael Von Korff, along with several other faculty members, are conducting workshops for representatives from 26 health care organizations. These organizations run the gamut from community health centers serving primarily uninsured populations to the Mayo Clinic. After being exposed to strategies for improving depression care, many of which stem from the CHS/UW depression research program, the organizations immediately begin testing innovations in their practice settings.

In research contexts, the Collaborative Care model has been modified to serve different populations. In a clinical trial conducted at the Seattle Veterans Administration, Susan Hedrick and colleagues added a Social Work Assessment to the basic Collaborative Care model to increase its relevance to the VA’s population. Project IMPACT, a seven-site nationwide study, is currently using versions of the CHS educational materials and Collaborative Care model that have been adapted for a large population of depressed adults over age 60. The Collaborative Care model has also been adapted for conditions besides depression. Bruce Rollman, Assistant Professor at the University of Pittsburgh Department of Psychiatry, says: "Our NIMH-funded clinical trial is extending the collaborative care approach pioneered by our colleagues in Seattle with depressed patients towards the treatment of primary care patients with panic and generalized anxiety disorder."

The methods and materials used in the Relapse Prevention Study described above played a large role in an NIMH-funded pharmacist intervention designed to increase patients’ compliance with antidepressants. Kathy Bungay, a pharmacist from the Tufts/New England Medical Center, credits Evette Ludman for her invaluable contributions to the pharmacist intervention manual and for introducing the pharmacists to Motivational Interviewing. Describing her research over the last 2 years, Bungay says, "We would not be this far without the support from [the Seattle group]."

Jurgen Unutzer, Kenneth Wells and Lisa Rubenstein of the Rand Corporation and Kathryn Rost of the University of Arkansas are collaborators on a major quality improvement initiative testing approaches to improving depression outcomes in primary care. This initiative includes four studies, one of which is the "Partners in Care" study (PIC) which was recently the subject of an article in the Journal of the American Medical Association. According to Rubenstein, the PIC "owes a tremendous debt to the outstanding work of the [Seattle group] over the past decade." UW psychiatrist Wayne Katon worked with the PIC researchers to revise the educational materials to increase their relevance to many diverse managed care organizations. Katon also consulted on how to make the Collaborative Care model nurse-assisted. This modified Collaborative Care model was implemented as part of the PIC study in 46 primary care clinics in 6 managed care organizations nationwide. The intervention improved process measures (for example, medication adherence), mental health outcomes and retention of employment over a one-year period (JAMA 2000;283:212–220).

Scientists on the CHS/UW depression team are: Terry Bush, Wayne Katon, Elizabeth Lin, Evette Ludman, Joan Russo, Carolyn Rutter, Greg Simon, Michael Von Korff, and Ed Walker. Summarizing this group’s contribution, Bedirhan Ustun of the World Health Organizations says, "Investigators from Group Health Center for Health Studies have made important contributions to the World Health Organization’s global program of mental health and disability research in the last 10 years. Their work has helped improve health services for mental disorders worldwide."

Top