Collaboration Key to Integrating Depression Treatment in Primary Care

“We now have 12-month data from IMPACT that are providing us with some preliminary answers to our questions,” said Jurgen Unutzer, M.D., M.P.H., an associate professor in residence of psychiatry and behavioral sciences at the University of California at Los Angeles and the study’s primary investigator.

The IMPACT trial is looking at the effectiveness of a collaborative treatment program aimed at the treatment of depression in the primary care setting. The study is operating at seven sites with 18 primary care clinics in five states with more than 400 primary care providers. Clinic settings include both urban and rural.

A total of 1,801 patients over the age of 60 (average age of 71.2 years) were randomly assigned to either the IMPACT intervention or to “usual care” of depression within the primary care setting. For those assigned to the intervention, the patient and their primary care provider chose either a brief psychotherapy (six to eight sessions) for depression (Problem Solving Treatment in Primary Care, or PST) or a stepped model of intervention.

All patients had either major depression or dysthymia identified by the Structured Clinical Interview for Depression. Patients who were already receiving psychiatric treatment, had suicidal ideation, or significant cognitive impairment were excluded from the trial; however, no medical comorbidity was cause for exclusion.

The stepped-intervention patients were provided access for up to 12 months to a depression care manager—most commonly a registered nurse or psychologist—who was supervised and supported by a psychiatrist. Patients were offered an educational video and written brochures on depression and its treatment, close care management including follow-up monitoring of symptoms and any adverse effects through a Web-based tracking system, and antidepressant medications. All care was administered in the patients’ primary care settings; however, a psychiatric consult was offered to those who did not improve.

At the 12-month point, Unutzer reported, 45 percent of patients receiving the intervention had a 50 percent or greater reduction in their Hamilton Depression Rating Scale score, compared with 19 percent of those patients receiving usual care. At the start of the intervention, 40 percent of patients were prescribed an antidepressant. At the 12-month point, that had risen to 60 percent.

Unutzer indicated that patients receiving either the brief psychotherapy or the stepped intervention were experiencing similar improvement; however, he said, the data are preliminary and more time and analysis will allow the team to analyze fully the differences between the psychotherapy group and the group receiving the stepped therapy.

“Clearly IMPACT is successful,” Unutzer said. “Especially with severe depressive symptoms, there was robust improvement with significant decreases [in symptoms] seen at three, six, and 12 months.”

The most severe outcome in geriatric depression is, of course, suicide, said Ira R. Katz, M.D., Ph.D., a professor of psychiatry at the University of Pennsylvania and primary investigator for PROSPECT there. Men over the age of 65 have the highest rate of suicide of any demographic group.

The NIMH-sponsored PROSPECT study, like IMPACT, is also testing the effectiveness of depression care managers within the primary care setting. PROSPECT is measuring the effects of a structured intervention versus usual care on depressive symptoms, suicidal ideation, hopelessness, and functional and quality-of-life outcomes.

Approximately 1,200 subjects have been enrolled at 18 primary care centers in the northeastern United States, with an overrepresentation of individuals over the age of 75. Patients are randomly assigned to receive either usual care (potentially augmented by depression screening) or the structured intervention that includes the antidepressant citalopram (Celexa) and interpersonal therapy.

“We asked whether or not you can impact suicide by treating depression and/or dysthymia in high-risk elderly,” Katz said. “And clearly the answer is yes.”

Katz reported that results at four months indicate that 41 percent of patients receiving the intervention achieved a 50 percent reduction on their Hamilton Depression Rating Scale scores, compared with 27 percent of those patients assigned to the usual care group.

“But what we are finding,” he emphasized, “is that this works for major depression alone. The results are not statistically significant in any other category [with less severe symptoms].”

Katz and his group found that for those patients who expressed suicidal ideation at baseline, there was a statistically significant decrease in scores on the Scale for Suicidal Ideation.

“The punchline is,” Katz said, “both PROSPECT and IMPACT work. But we don’t know what works for whom and how we sustain the gains.”

The PRISMe study, overseen by Bartels at Dartmouth, is comparing two collaborative models: concurrent, co-located mental health and primary care services, and an enhanced referral system.

“We are asking here,” said Bartels, “whether we need an integrated model to enhance a patient’s engagement in mental health services or can we achieve that enhancement through a rigorous referral process from primary care out to specialty mental health services?”

With more than 2,000 patients being followed by PRISMe, the range of diagnoses involved is a bit wider. Approximately 70 percent have depression, 20 percent have at-risk alcohol use, 3 percent have an anxiety disorder, and about 7 percent have a mixed or dual diagnosis.

Overall patient “engagement”—defined as the patient attending his or her first visit with mental health services and accepting treatment—was clearly better in the integrated-service model, where mental health and primary care services are offered in the same office: 71 percent of patients in the integrated model engaged in treatment, compared with 49 percent in the enhanced referral model.

Interestingly, Bartels noted, those with at-risk drinking had high rates of engagement. Severity of symptoms was once again noted as a predictor of success, with 83 percent of those actively having suicidal ideation engaging in the integrated model.

“It really makes you think about what the definition of a functional integrated model of services is,” Bartels said. “But an even more key question is, How do we accomplish integrated care, and perhaps even more importantly, How do we pay for it?” Bartels asked.

All three primary investigators noted that many more questions remain, although these preliminary data are encouraging. All three noted the difficulty of integrating mental health services into the “prevailing culture of primary care.”

Moreover, each knows that the cost of integrating care is justified by the reduced morbidity as well as reduced mortality inherent in successful treatment of geriatric mood disorders. Yet further research, they agreed, is going to have to address the costs and show a substantial offset for systems such as Medicare and Medicare-Plus HMOs to see a benefit.