Wishing Upon a STAR*D: The Promise of Ideal Depression Care by Primary Care Providers

Depression continues to be a leading cause of morbidity in the United States, but it remains undertreated. Many depressed patients seek treatment from primary care providers, which makes improvement in the quality of depression care in primary care settings vital. Studies of collaborative, or enhanced, care approaches show that depression can be effectively treated in primary care settings, with improved short- and long-term outcomes, when appropriate systems of care are put into place ( 1 ). These approaches have also been shown to be cost-effective ( 2 ).

The accumulation of evidence from these and other studies led the U.S. Preventive Services Task Force to revise its clinical practice guideline for depression so that it now recommends screening and treatment of depression in primary care practices when appropriate follow-up systems are available ( 3 ). However, studies of enhanced primary care for depression were not designed to compare results between primary care and specialty mental health care.

A lingering question has been whether better outcomes can be achieved by directly referring depressed individuals to mental health specialists. For example, in studies of collaborative care approaches in primary care that show improved depression outcomes compared with usual care, the difference might result from a lack of access to mental health specialty care for participants in the comparison groups. If so, perhaps fast-tracking all depressed patients seeking treatment in primary care into mental health specialty care would produce still better outcomes than those achieved by patients in supported primary care models, such as collaborative care. The Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial can shed light on this issue, because its design allowed for direct comparison of outcomes of depressed patients after treatment in either a primary care setting or a mental health specialty setting ( 4 ). This commentary examines the implications of STAR*D for primary care providers.

STAR*D and primary care

The STAR*D trial was a broadly inclusive effectiveness trial that enrolled depressed outpatients from primary and specialty care settings using identical enrollment criteria. Enrolled patients in both primary care and mental health specialty settings were similar with regard to presenting depression symptoms and severity ( 5 ). The overall aim of the STAR*D trial was to define prospectively which of several treatments are most effective for depressed outpatients who have unsatisfactory initial and, if necessary, subsequent clinical treatment outcomes.

The initial treatment in the STAR*D trial was 12 weeks of citalopram. There were four "levels" of care through which patients could proceed. Patients whose depression did not sufficiently improve at a given level were encouraged to move on to the next level. Protocols helped primary care or mental health specialist clinicians move patients through other levels of treatment. The protocol called for the clinician to switch antidepressants or augment treatment with another antidepressant or psychotherapy, depending on the clinical response to the designated therapies.

Treatment support did not involve use of a collaborative care model, although STAR*D provided some of the usual components of collaborative care, including requiring frequent visits and providing the treating clinician with standardized clinical assessment information and treatment algorithms. In each treatment level, clinical treatment visits after the initial visit occurred at two, four, six, nine, and 12 weeks. Collaborative care typically involves a similar visit pattern; however, most visits in collaborative care models occur with a nurse, social worker, or pharmacist care manager. STAR*D clinicians used the clinical information collected at these visits to raise the dosage if needed on the basis of the treatment protocol ( 6 ). After the first level of treatment (citalopram only), patients in primary care settings had the same or slightly better rates of depression remission and level of depression severity than those treated in specialty care settings ( 4 ).

What have we learned as a result of STAR*D? If the only difference in care is the treatment setting—primary care or mental health specialty setting—patients with depression have equal outcomes after initial treatment with a 12-week course of citalopram. This is an important finding, because it suggests that treatment of depressed individuals in primary care settings will not, all things being equal, result in worse outcomes. Mental health specialist training and practice styles are apparently not necessary to enable treating clinicians to achieve the best treatment outcomes currently possible for depression, as long as they provide the types of support provided by STAR*D.

Unfortunately, this finding does not mean that depressed patients will fare so well in usual primary care practices. Such practices often lack adequate appointment availability, do not provide proactive care for detecting problems with or monitoring adherence to treatment, and lack access to specialty mental health care for patients who have complex conditions or who do not respond to treatment. In addition, most primary care clinicians are not paid for providing mental health care. Finally, STAR*D did not address the potential benefits of primary care-based psychotherapy as an equivalent nonpharmaceutical option for depression treatment.

Lack of appointments

Most primary care practices do not have delivery systems in place that can mimic STAR*D care patterns. The STAR*D sequence of care included visits at two-week intervals over the three months after initial diagnosis. The lack of timely appointments with primary care physicians is known to be a problem ( 7 ), and thus it is highly unlikely that a depressed patient could receive this level of consistent follow-up. What happens during a primary care appointment is also an issue. A recent study found that the median time spent on a mental health topic in a primary care setting is two minutes, and mental health is addressed with only 25% of depressed patients ( 8 ).

Lack of care management

Care management could provide the frequent contacts that occurred in the STAR*D trial or help address the competing demands that prevent primary care physicians from having the time to address mental health topics during visits ( 9 ). However, most primary care practices consist of groups of three or fewer physicians ( 10 ), and sustaining a full-time care manager would be difficult for small practices. Telephone care management ( 11 ) could overcome this barrier if pooled across clinics. However, this requires systemic organization, such as a practice network ( 12 ), which is usually not in place in the current health care environment. Care management could be organized at a health plan level, such as through a health maintenance organization or a managed behavioral health organization ( 13 ). Current arrangements for providing mental health care, in which a primary care practice may interact with multiple managed behavioral health organizations, may make it difficult for most practices to finance and coordinate care management ( 14 ).

Lack of payment

Primary care physicians cannot bill for depression care services when these services have been carved out to managed behavioral health organizations that do not include primary care physicians in their network of providers ( 15 ). Primary care clinicians are also unpaid for the extra work required to meet standard protocols or guidelines for providing mental health care. Furthermore, care management for depression in primary care is not reimbursed by major insurers at this time.

Lack of access to specialty care

In the STAR*D trial, responses to the first level of treatment were similar for patients who were treated in primary care or mental health specialty care. It will be important to determine whether outcomes differ by setting for patients whose depression failed to remit after the initial treatment. These individuals may need specialty mental health care, and unfortunately, access under current systems of care still needs improvement. Collaborative care models that incorporate improved access to mental health specialty care are more effective than those without such access ( 1 ).

Lack of access to psychotherapy

Some patients prefer psychotherapy over pharmacotherapy ( 16 , 17 ) and should be able to receive psychotherapy initially. Making short-term psychotherapy available to depressed patients in primary care was found to have a positive impact on health outcomes for more than two years ( 18 ). These services are not easily accessible from primary care clinics, and primary care clinics are often unwilling to set up the services needed to conduct psychotherapy within their clinics, such as setting aside an office for a mental health specialist or introducing problem-solving therapy by depression care managers.

Conclusions

Despite the difficulties in improving primary care for depression, progress is critical. Studies show that around 15% of primary care patients screen positive for depression on brief instruments ( 19 , 20 ). Even with the new resources that may be created by recent mental health parity legislation ( 21 ), on the basis of numbers alone it is unrealistic to expect to refer all patients who screen positive to mental health specialists for a full evaluation and for treatment of major depression or related mood disorders, such as dysthymia. Furthermore, patient preferences need to be taken into account. Many of these patients do not see themselves as needing mental health specialty care and will refuse to attend ( 22 ). Yet without treatment, significantly more patients suffer the downward spiral of deteriorating clinical status and job loss ( 23 ), negative life events ( 24 ), and worsening functional status ( 25 ).

STAR*D has often been described as a "real-world" trial. Unfortunately, even this version of the real world is still not a part of reality for most primary care clinics—or even for specialty mental health. Substantial reengineering of payment and delivery systems is needed to make the STAR*D approach viable in primary care clinics. This reengineering must take place if patients are to receive adequate care. We look forward to the day when the STAR*D approach, possibly supplemented by enhanced access to brief, evidence-based psychotherapy, can be implemented in all primary care settings.

Acknowledgments and disclosures

This work was partly supported by the UCLA-RAND NIMH Partnered Research Center for Quality Care (National Institute of Mental Health grant P30-MH082760) and by the Center for the Study of Healthcare Provider Behavior, which is a partnership of U.S. Department of Veterans Affairs Health Services Research and Development Service; the University of California, Los Angeles; and RAND.

The authors report no competing interests.