Integrated Care: High-Impact Mental Health–Primary Care Research for Patients With Multiple Comorbidities

Although antidepressants are an effective treatment for depression, adherence is generally poor, which greatly diminishes their effectiveness (8). PCPs, where the majority of antidepressants are prescribed, need practical interventions to promote antidepressant adherence. Clarke (9) is leading a research team that is testing a very low-cost, direct-to-patient adherence intervention to prompt antidepressant refills and thereby boost adherence. The randomized controlled trial (RCT) involves 3,100 PCP patients newly dispensed an antidepressant for a mood or anxiety disorder, as identified by the health care system EMR. The intervention, which uses automated telephone interactive voice recognition (IVR), phones patients with reminders timed to their antidepressant refill dates. The IVR program also offers patients brief antidepressant psychoeducation, or they can request to be transferred to a live pharmacist or the mail refill pharmacy. Recruitment of patients for the study is by mailed brochures, which is inexpensive and similar to how patients might access such an intervention in real-world practice. Study exclusion criteria are minimal, resulting in a highly generalizable sample, and the intent-to-treat design assesses both intervention drop-outs and completers to better understand who participates. To inform future implementation, investigators will qualitatively assess patients’ reasons for not taking prescribed medications and health care administrators and PCP staff’s perceptions of the intervention’s benefits and costs and their intention to continue the program. A cost-effectiveness analysis is also planned.

Collaborative care for depression is effective across a variety of patient populations (10). The collaborative care model includes a primary care physician, depression care manager, and consulting psychiatrist; employs a treat-to-target approach; and uses stepped care, reserving specialty consultation for more complex patients. Despite the repeated success of this model in dozens of clinical trials, the uptake of collaborative care by PCPs has been modest, due in part to health care system concerns over relative costs and benefits. Accordingly, Simon and colleagues (11) assessed the incremental long-term cost and effectiveness of this model when delivered via telephone at two intensity levels: lower-cost care management delivered by a bachelor’s-level clinician versus higher-cost care management plus cognitive-behavioral therapy (CBT) delivered by a master’s-level clinician. Patients starting antidepressants but not receiving psychotherapy were identified through the EMR of the large provider system, which serves a diverse population. Both interventions resulted in more depression-free days compared with usual primary care. However, patients who received telephone care management plus CBT gained more depression-free days than those who received telephone care management only, and their net outpatient health care costs were lower over the two-year period.

Homeless mothers with depression are a vulnerable group who need basic services and mental health care. When they seek psychiatric services, it is mainly through PCPs. The chronic care model, which emphasizes self-management and leveraging community resources, is effective for diverse PCP patients. However, the stressors accompanying homelessness necessitate adaptations for these women around problem solving, illness self-management, and access to essential resources. To develop a practical depression care program for this hard-to-reach population, Weinreb (12) partnered with national Health Care for the Homeless primary care clinics, which offer case management and other supports to meet this population’s complex needs. The quasi-experimental pilot study of ethnically diverse homeless mothers with depression aims to reduce depression symptoms and improve functioning and housing status. Exclusion criteria are few, and Weinreb will assess individual and program factors that may moderate outcomes. This study will also inform development of integrated care interventions for the growing number of low-income and previously uninsured individuals soon to enter PCPs under health care reforms.

Attempts to close the research-practice gap by speeding research evidence into practice are one-sided and inadequate (5,6). Generating practice-ready mental health–primary care interventions also entails practice-based evidence, where the experience of providers, patients, and health care systems directly informs the research at all stages. This “health care system as lab approach” is exemplified by practice-based research networks (13), such as the Mental Health Research Network (MHRN). The MHRN is a consortium of 11 public-domain research centers based in not-for-profit health care systems (www.mhresearchnetwork.org). With more than ten million patients in 12 states, the MHRN is the largest U.S. mental health research network.

In this learning health care system, critical practice problems drive the research questions and data needed to answer them, research methods leverage the existing research infrastructure for rapid and efficient knowledge production, the health care system informs the interpretation of findings by providing critical context, and decision makers obtain practice-relevant evidence on which they can act. The MHRN has created standard definitions of mental health events and standard mental health assessments and has built a virtual data warehouse that links health information databases. This research infrastructure dramatically increases the efficiency of research while protecting patient privacy. For example, the MHRN is conducting a surveillance study of depression treatments and evaluating their effectiveness using routinely collected scores on the nine-item Patient Health Questionnaire (PHQ-9). The MHRN is also developing capacity for a practical trial of population-based suicide prevention programs with 15,000 outpatients who are at risk of suicide attempts on the basis of elevated PHQ-9 scores.

Patients in PCPs frequently have co-occurring depression and diabetes, with each condition complicating the other and requiring self-management. Self-efficacy is a key mediator in patients’ ability to perform critical self-care activities, but current self-efficacy interventions are seldom available in PCPs. Jerant (14) is conducting an RCT pilot study of self-efficacy–enhancing interviewing techniques (SEE IT), a brief provider training intervention to increase patient self-efficacy for managing these conditions in PCPs. In three 15-minute sessions, primary care providers from 14 offices are trained to use SEE IT with actual patients during routine office visits. Patients with co-occurring diabetes and depression are identified from the EMR and claims database. Outcomes include patients’ self-care for depression and diabetes, medication adherence, and symptom severity. Jerant held focus groups for PCP physicians, staff, and patients to inform the intervention development and will meet again with PCP physicians and staff after the RCT analyses to capture lessons learned. This highly flexible, low-intensity, low-cost intervention has potential applicability to PCP patients with a variety of mental and general medical conditions requiring self-management.