A blueprint for integrated mental health care

The long‐standing debate over how best to organize services and payment policies for mental health has hinged on how we choose to simplify the multiple complex connections of mental health conditions with other human ailments. Because mental health problems can be legitimately viewed as highly distinct from, or, alternatively, necessarily intertwined with, medical conditions, policies have vacillated between encouraging or discouraging integration of mental health services with other medical services. Both approaches have advantages and disadvantages, and both are hobbled by structural issues in mental health care financing and the complexity of treatment options. For example, whether due to stigma or the challenges demonstrating the clinical impact of mental health services, caps on benefits and lower reimbursement rates have led to chronic under‐investment in mental health services. This under‐investment has led to inadequate capacity for both inpatient care and outpatient care, particularly for those with serious mental illness (SMI). In addition, the number of different professions providing mental health care results in a confusing array of services, impeding access and coordination. The numerous medications, each with their own profile of effectiveness and side effects, can be complicated to manage in general medical settings, which provide half of all the mental health care in this country. ¹ Short‐term psychotherapies such as cognitive behavior therapy may be very effective, but numerous barriers to providing these services at meaningful scale remain. And these challenges do not even begin to touch on the complex interplay between mental health and social determinants of health.

This is the context within which we should view the recent trends from both inside and outside mental health that are driving the current movement toward increased integration of mental health services with medical care. Key trends include the increased complexity of safely administering and monitoring therapeutic advances; the near‐universal adoption of information technology; and the movement toward comprehensive primary care, including integrated mental health services. This latter trend has itself been driven by payment policies (alternative payment models or APMs) encouraging the management of total medical expenses (TME), evoking the now well‐established role of mental health conditions as a major driver of medical costs.

What do we know about integrated mental health care models? Viewed from the perspective of just a decade ago, the literature on integrated models of care delivery appears fairly extensive. ² , ³ , ⁴ We now know that it is possible to deliver mental health services that are integrated with primary care at relatively low cost, of high quality, and result in improved access. Most mild and some moderate mental health conditions can be handled with the well‐managed deployment of resources in primary care. Some key elements include the following: periodic population screening to catch undiagnosed illness; warm hand‐offs to reduce barriers to transitioning into mental health care; guidance from behavioral health specialists acting as consultants rather than direct service providers; assessment and triage to short‐term therapy or coaching, which provides individual benefits while enabling access by freeing up professionals from long‐term engagements; and coaching for substance use disorder. All of this requires design input and clinical backup from psychiatric specialists, but for the most part may be delivered by non‐physician professionals. Clearly the more specialized services required for moderate to severe mental illness go well beyond what the typical integrated primary setting can offer. Some have gone so far as to build behavioral health homes for SMI patients, a form of reverse integration where components of medical care are embedded into a unit focused on serious mental health needs. ⁵ , ⁶ , ⁷ Despite these advances in our conceptual knowledge, given the numerous design options and management decisions necessary to create functioning integrated mental health services, the implementation handbook remains comparatively thin. ⁸ , ⁹

Understanding this broader context and current trends toward integrated mental health care delivery may be the easiest part. For those on the front‐line trying to improve access to mental and behavioral health services, the challenge is figuring out how to get there. This is the context in which we must consider the contribution of Kai Yeung, Julie Richards, Eric Goemer, Paula Lozano, Gwen Lapham, Emily Williams, Joe Glass, Amy Lee, Carol Achtmeyer, Ryan Caldeiro, Rebecca Parrish, and Katharine Bradley in this edition of Health Services Research. This paper provides valuable guidance, a blueprint of sorts, to those working on health system transformation. Both of us have worked together within a large integrated delivery system on a 10‐year continuing journey attempting to implement an integrated mental health model. Having such a budget in mind enables conversations with finance about what the expected costs and revenue will be, and conversations with clinicians about the required hires, reporting relationships, and available patient services. It also provides a scaffolding for building detailed plans, which are a required component of most change processes—the more detailed the plan, the greater the confidence of the various stakeholders that the transformation will be successful. Additionally, by acknowledging the important distinction between start‐up costs and ongoing (operating) costs, the authors have recognized that allocating capital is a separate process from allocating operating expenses. And the inclusion of ongoing quality improvement costs is an explicit (and welcome) recognition that there are still many unknowns in organizing optimal integrated mental health services.

Importantly, the program costs accounted for in this paper reflect only one program design option. Needless to say, there are many options for deploying integrated mental health models. For example, the authors have described the types of professionals on the particular team they built, but would a different mix produce similar outcomes at lower cost? What metrics determined their individual or practice level performance? What are the costs of communications and metrics required to gain acceptance and commitment to adopt this model? And beyond the questions that might have become apparent with a more detailed forensic accounting approach, what was the impact of this model on primary care revenue? Was the delivery system in this paper relatively homogenous (ie, all employed physicians) or did it include multiple types of practices and settings? The authors note that they did not consider the “downstream” costs of addressing the behavioral health issues identified by periodic screening. These costs could be substantial. On the other hand, if this model was deployed in a system that accepts financial risk for TME, a return on the required investment is also possible. Also, the amount any particular delivery system headed down this path needs to budget will differ depending on the resources already deployed. For example, unlike the authors, we did not need to budget for some of the information systems and medical home and screening infrastructure accounted for in this paper because these were already deployed and/or included in other budgets.

Reading through this paper, we saw numerous elements that would have helped us on our own journey in behavioral health integration at our health system had we known them at the time. We commend the authors for taking the time to assemble this information in a useful format so that others might benefit from what these authors have. Better still would be an analysis of design trade‐offs, relative effectiveness, and return on investment under different contracting models. While such analytics are well beyond our current reach, this paper is nonetheless an important blueprint for those interested in building a better health care delivery system. We remain hopeful that the integration of mental health services into primary care will address some of the cost and access issues that people with mild and moderate mental illness encounter so that we are better able to strengthen our system's capacity to care for those with serious mental illness.

Supporting information

Supplementary Material