Can Collaborative Care Cure the Mediocrity of Usual Care for Common Mental Disorders?

Despite a 3- to 4-fold increase in the use of antidepressant medications, the prevalence of depression and anxiety disorders in Australia, Canada, the United Kingdom, and the United States has remained unchanged over the past 20 years. ¹ In the absence of compelling evidence that the incidence of these disorders is on the rise, a natural conclusion is that depressed or anxious patients who could benefit from treatment are still not identified and treated, or that the duration of illness has remained unchanged in those who are treated. This is a striking and troubling finding, considering the known efficacy of antidepressants and psychotherapies. It emphasizes both a well-delineated treatment gap, whereby many patients with depression or anxiety do not receive treatment, and a quality gap whereby those who are treated either do not need to be treated or do not receive effective treatment. ^{2 –7}

Several factors contribute to these gaps. Even knowledgeable and well-intentioned physicians face competing demands, including the need to concurrently address medical comorbidities and social determinants of health, making identification and management of common mental disorders challenging. ‘Usual care’ for these disorders requires both a patient-initiated encounter and the clinician’s subjective impression of the presence of a treatment-responsive condition, leading to delays in the initiation and titration of treatments, ^8,9 high treatment dropout rates, and low recovery rates. ^5,10,11

A better integration of the care provided by primary care providers and specialists has been advanced as one solution to improve access to care and the quality of the care delivered. ^{12 –14} Integrated care encompasses models ranging from colocation (mental health providers delivering care within the primary care setting), to shared care (with increased coordination of care through provider-to-provider communication and shared health records), to the collaborative care model (CCM) (also known as ‘chronic care model’). ^{12,15 –18} The CCM is a well-established and effective approach to both mental and physical disease management that includes several key elements: 1) team-based care that includes patients as active member of the treatment team and establishes patient-centred goals and care plans; 2) measurement-based care where patient-reported outcomes are monitored using rating scales and treatments are regularly adjusted to reach predefined targets (e.g., a specific depression score corresponding to remission); 3) treatment selection and adjustments based on evidence (encapsulated in algorithms or care pathways) or, in the absence of evidence, expert opinion; and 4) population-based care whereby patients with a target condition are identified via systematic screening, tracked in a clinical registry, and reached proactively if they are disengaging from treatment. ^{19 –21}

In primary care, where most common mental disorders are treated, ^14,22,23 more than 80 randomized controlled trials (RCTs) have shown that collaborative care is more effective than usual care to improve quality of care processes, clinical outcomes, patient and provider experience, health disparities, and cost-effectiveness. ^{24 –27} A recent commentary on collaborative care suggested that further RCTs on collaborative care were unnecessary and possibly unethical given the existing body of evidence supporting its implementation. ²⁸

In this issue, 2 systematic reviews assess the evidence for integrated care in pediatric and geriatric patients. ^29,30 Campo and colleagues ²⁹ highlight the early onset of many mental illnesses and the substantial quality gap in care for children and adolescents, which may portend poor outcomes, including suicide. They describe a wide array of pediatric mental health integrated care initiatives, including telephone consultations, education, colocation, and the CCM. They conclude that, in the pediatric population, integrated care approaches are more effective than usual care for depression, anxiety, and disruptive behaviour disorders but not substance use disorders. These approaches improve key clinical processes (such as treatment initiation or completion), clinical outcomes (such as response and remission), and patient and provider experience. Among the various integration initiatives reviewed, the CCM yields more robust and positive outcomes. The advent of the Patient’s Medical Home, a primary care reform initiative that provides comprehensive and accessible team-based primary care, amplifies opportunities for integration. Fortunately, the CCM ‘has powerfully influenced the conceptualization of the medical home’, ²⁹ emphasizing proactive and longitudinal care and the ‘incremental steps that produce sustained progress’. ³¹

At the other end of the life span, Bruce and Sirey ³⁰ review integrated care interventions for older depressed patients. They point out the magnitude of the disease burden, as well as its impacts on caregivers, comorbid medical conditions, and mortality. The high prevalence of geriatric depression in primary care, older patients’ preference for being treated by their primary care provider, and the shortage of geriatric psychiatrists are critical drivers for integration of depression care for older adults.

The authors methodically lay out the varying evidence for different types of integration. The PRISM-E study of colocated care included reduced wait times and structured communication between specialists and primary care providers, as well as showed improved acceptance of treatment but no difference in clinical outcomes. By contrast, the IMPACT trial was a seminal collaborative care intervention, operationalized as a care manager offering brief psychological interventions, pharmacotherapy, and proactive patient monitoring using measurement-based care and a clinical registry, as well as on-site psychiatric expert input to guide stepwise depression care to target remission. In IMPACT, intervention patients were more likely than usual-care patients to receive and continue treatment, achieve response and remission, and be satisfied with their care. Similarly, the PROSPECT collaborative care trial showed improved depression care and outcomes, including reduced suicidal ideation and long-term mortality. These and other studies of collaborative care have shown its impact not only on depressive symptoms but also on other outcomes important to patients, including physical health functioning, self-efficacy, and quality of life. Subsequent studies have also demonstrated the value of CCM for older patients with anxiety or Alzheimer disease or for those who are homebound. Greater effect sizes are observed with interventions in patients with more severe depression or with comorbid chronic physical illness, as well as in populations that commonly experience health disparities such as those with low socioeconomic status.

Colocation of specialists in primary care is becoming more common in Canada, but colocation by itself is not enough to achieve improved outcomes. Despite 20 years of evidence supporting the effectiveness of the CCM, it has rarely been adopted in clinical settings. Both reviews discuss how this problem is perpetuated by knowledge gaps regarding what adaptations of the model may be less resource intensive and more scalable, as well as what supports and structures are needed to successfully implement collaborative care in the ‘real world’. Answering these questions will require study designs beyond RCTs: realist reviews, pragmatic trials, quasi-experimental or mixed-methods program evaluations, and rigorously evaluated quality improvement interventions.

There are other barriers and challenges to implementation in Canada.

Access and the Role of Health Care Reform

Primary care reform toward team-based models of care (e.g., the Patient’s Medical Home) is in variable but overall limited stages of progress. In Ontario, where primary care reform is relatively more advanced, Family Health Teams (FHTs) and Community Health Centres could provide expanded access, interdisciplinary care, and quality improvement. However, they care for less than a third of the provincial population, and their fidelity to the CCM components is highly variable. ³² Moreover, owing at least in part to financial disincentives, people with mental illness and addictions are underrepresented in FHTs, although they are among those who could benefit the most. ^33,34 To better support CCM implementation, primary care reform will need to include appropriate adjustment of capitation funding models for case complexity and ensure that mental health is on the primary care quality improvement agenda. Finally, across the country, access to empirically supported psychotherapies remains poor. The promised federal investments in mental health care should consider the strategies that have been adopted in other countries such as England or Australia. ³⁵

Culture and the Need for Champions

These news ways of practicing require changes to the culture of health care delivery, roles, workflows, and competencies. ^{36 –39} The DIAMOND study (Depression Improvement Across Minnesota—Offering New Directions) was the largest effort to scale collaborative care across an entire population. ⁴⁰ Overall, DIAMOND failed to improve depression outcomes, but we can learn from the analysis of its implementation and of those who benefitted from it. The high referral rate to the study was attributable to several factors: strong clinical and managerial leadership support; perceptions of care manager abilities and ‘fit’; well-defined care manager role with adequate time, space, and support; care manager visibility and accessibility on site; and a strong primary care provider champion. ⁴¹ These findings resonate with our own experience with the implementation of the PARTNERs trial of telephone-based collaborative care for depression, anxiety, and alcohol use disorders. ⁴² They are echoed in Atul Gawande’s New Yorker article, ‘Slow Ideas’, illustrating that changing health care practice norms is fundamentally a social process. ⁴³ In DIAMOND, 6-month remission rates were influenced by in-person referral from primary care providers to the care manager (‘warm handoffs’), an engaged and responsive psychiatrist, and adequate funds for clinic operating costs. ⁴¹ Leadership and infrastructure may be the most overlooked core element of the CCM; they will be crucial for its successful uptake in ‘real-world’ settings. DIAMOND findings underscore the complexity of CCM implementation and the potential contributions of implementation science and quality improvement science to advance this field.

The Psychiatric Guild and the Role of Training

Psychiatric practice remains siloed, inefficient, and inaccessible, ^35,44 a problem that is perpetuated by funding models and the isolation of many community-based psychiatrists. Campo et al ²⁹ and other authors have noted that education is a necessary component to promote implementation. Although collaborative care competencies and a training requirement for psychiatry residents in Canada have been established, ^38,45 there are challenges with the quality of training. ⁴⁶ A systematic review of the literature suggests a lack of 1) continuing professional and faculty development, 2) team-based training of all relevant disciplines together, and 3) quality improvement of the clinical training environment. ⁴⁷

Innovation, evaluation and the role for practice-based evidence

Bruce and Sirey ³⁰ describe creative efforts to develop scalable integrated care interventions, including the use of telephone-based care and lay providers. We agree that testing and refinement of these types of interventions are needed. There are ongoing examples of such innovation in Canada: the Champlain adult and pediatric psychiatry e-consult service ^48,49 ; the Rapid Access to Consultative Expertise (RACE) telephone-based specialist consultation service in British Columbia ⁵⁰ ; the PARTNERs study in which lay providers supervised by a psychiatrist provide care management and decision support by telephone for Ontarian adults with depression, anxiety, or at-risk drinking; and the Ottawa Depression Algorithm online tool supporting measurement-based evidence-based stepped care, self-management, and community resources. ⁵¹ While promoting the uptake of evidence-based collaborative care practice, we also wish to encourage the creation of new ‘practice-based evidence’ ⁵² (i.e., rigorous implementation and outcomes evaluation of innovations in clinical practice settings), ideally through academic-community-patient partnerships that address the diversity of stakeholder perspectives necessary for spreading and scaling collaborative care. ^53,54

In conclusion, the CCM is a critical strategy to closing the current access and quality gaps in mental health care across the life span. However, the overwhelming evidence supporting its efficacy is not sufficient to ensure its success. We need to understand how to successfully implement it across Canada. Health care policy and financing are key enablers to the spread of CCM, but so too are clinician innovators.