Skip To Navigation Skip To Content Skip To Footer
    Insight Article
    Home > Articles > Article
    Blake Griffin Edwards
    Blake Griffin Edwards, MSMFT, LMFT

    In 2020, a Milliman study of 2017 commercial healthcare claims data from 21 million people ages 2 to 64 found that while only 27% had a behavioral health condition, those patients accounted for 56.5% of total healthcare expenditures. Most of those costs were for medical and surgical services. Behavioral health conditions were highly correlated with total spending, yet little was invested in proportion: Only 4.4% of healthcare costs were attributed to behavioral health services.1

    Among those with comorbid behavioral health and physical health conditions, a disproportionately high percentage was spent on physical health-related treatment (e.g., ED, surgery, labs, office visits), and a disproportionately low amount was spent on behavioral healthcare. Prescription drugs represented 40% of those total behavioral health-specific costs in 2017.

    Half of those studied with at least one behavioral health condition — 5.93 million individuals — incurred less than $68 of total annual costs for behavioral health treatment. Another 2.97 million patients’ expenses with at least one behavioral health condition ranged from $68 to $502. The remaining 2.97 million incurred costs for behavioral health-specific services averaging more than $502 per year.

    This is not necessarily surprising. Different types of services and procedures have very different costs — a physician’s education costs and pay extends in the upward realms when compared to a master’s-level mental health therapist, for example, and the types of laboratory tests, medications, and emergency procedures carried out in medicine add to the cost of supplies and of specialists and critical medical infrastructure, including high-cost clinics, sterilization services, nurses, and much more. Whereas outpatient mental health and addiction services involve lower cost staff, facilities and fewer supplies.

    Yet Henry Harbin, MD, former CEO of Magellan Health, called the Milliman results “astonishing,” explaining, “This is despite having been diagnosed or treated by a healthcare professional for a behavioral illness.”2 Understanding this disconnect requires further analysis of the gaps between behavioral healthcare needs and availability, as well as where money is spent across the care continuum.


    There is a common misconception that patients with more severe mental illness such as bipolar and other mood disorders or schizophrenia and other psychotic disorders drive most healthcare costs among patients with comorbid conditions; in fact, they represent a very small percentage. Individuals studied with physical and behavioral health conditions may be relatively easy to treat since the data indicate that most of them have less severe mental health and substance use related conditions; yet according to that 2017 data, many simply did not receive any or much treatment at all.

    Identifying patients with comorbid conditions sooner and ensuring behavioral care access is offered for many conditions may eventually reduce overall healthcare costs — some have estimated that effective medical behavioral health integration could save $38 to $68 billion.3 The Patient Protection and Affordable Care Act4 encourages development of integrated approaches, especially driven by primary care, as one means of improving quality and lowering overall costs. Still, it’s not yet clear whether expanding integrated care offerings will lead to significantly improved health outcomes.

    What is clear is that current system capacity is insufficient — a recent Department of Health & Human Services (HHS) data review5 found several indicators suggesting the mental health treatment system does not have the capacity to address current rates of treatment:

    • There are rising numbers of young adults with perceived unmet needs for mental health treatment.
    • Inpatient and residential beds designated for mental health treatment have high utilization rates.
    • There are low rates of 30-day follow-up after hospitalization for mental illness. 

    Even before the pandemic, the trend was that while the use of mental health treatment increased steadily, availability did not keep pace with need.

    The HHS study concluded:
    The mental health treatment system does not have the capacity to address current rates of treatment need. Treatment system enhancements are needed to expand access for those with treatment needs who do not receive any treatment and to improve the continuity and quality of care among those currently receiving treatment. Expanding capacity will likely need to include increasing the number of mental health professionals in the workforce but will also likely require innovative approaches to extend the behavioral health workforce capacity, such as telehealth services and mobile applications.6

    The pandemic greatly accelerated telehealth utilization in the years since,7 yet many signs indicate that need and demand for behavioral health services still far exceed capacity.8

    Transformation value(s)

    In 2018, Sue Birch, director of Washington State’s Health Care Authority, cast this vision: “We want to move from encounter-based, volume-driven work to value-based work.”9 This has been the standard talking point of healthcare transformation initiatives. At its worst it has justified ramrodding insufficiently proven transformation schemes aimed primarily at cost savings rather than improving access to or quality of care; in practice, in many cases across the United States, it has served, in effect, as a smokescreen for reductions in adequate and predictable funding for services that people need, and sometimes in reductionism about the ways in which mental health and addiction services are delivered.

    Further, transformation initiatives give lip service to the social determinants of health (SDoH) yet too often capitulate to models of transformation which are, in practice, little more than another kind of siloing, with another set of problems, with negligibly better research outcomes once you adjust for observer effects, confirmation bias, design failures and conflicts of interest. The jury is still out on what “integrated care” entails and whether particular manifestations of it help.

    When they help, is it the model that makes the difference or the access to care?

    Regardless, we chase value in a vacuum. Whole-person care innovation efforts are more clinically focused than anyone cares to admit. Birch hit the nail on the head when she proclaimed, “When we overmedicalize, we take dollars away from addressing environmental and social needs.” Investments in non-clinical, community-based systems which directly impact SDoH remain largely inadequate; a great deal of social services infrastructure remains woefully neglected.

    Many transformation efforts push for systemic changes in behavioral healthcare that amount to medicalization of a sector of healthcare that evades much medicalization — measurement-based treat to target, and “value-based” pay arrangements tied, in part, to process and “outcome”-based metrics that require higher overhead with cost savings so indirect they cannot be calculated with certainty, leaps of faith in inefficient, self-serving managed care schemes, and little to no improvement.

    In one workshop at the Pediatric Population Health Forum where Director Birch provided that keynote, I asked a workshop presenter, “In your capitated system where provider compensation is tied to positive patient outcomes to drive organizational achievement of value-based contract measures, how do you ensure you don’t have a provider who only wants to see easy, non-complex patients?” The presenter, representing a large urban hospital system, assured, “Oh, we don’t have a single doc like that.” I responded in disbelief, “How could you possibly know that?”

    It is true that some integrated behavioral health models do better with providing access to care with less focus on measurable quality and others are well suited to measurement yet limit access. And primary care behavioral health normalizes and expands engagement with behavioral health professionals for a large population who would otherwise not receive care from qualified behavioral health professionals. However, there are many active, thriving examples of integrated behavioral health that do not fit neatly into a brief primary care behavioral health or collaborative care model. We need a diverse toolbox to meet our patients’ needs.

    The complexity of the real healthcare terrain requires flexibility to blend and sometimes ignore these models. In fact, many rural health centers find themselves practicing a hybrid of integrated and specialty service delivery models along their continuum of care as they respond fluidly to the unique needs of their patient populations and communities. These integrated models involve a lot of collaboration and continual, strategic planning and process improvement.10

    Different settings need different structures and processes to succeed, which are not always provided by purist fidelity to a model treated as the new standard on the basis of being designed in a way that is more research-friendly and less often on the basis of relative efficacy.

    What does the patient need? What does the population need? What are the workforce strengths and constraints? Answering these questions locally forges paths forward.

    “Co-location” should not be demonized as the opposite of “integration.” It is a commonly valued element of an integrated continuum of care, frequently offering a fulcrum in the perpetual balancing of needs between access to integrated brief intervention and treatment of complex conditions that involve everything from physical and behavioral health comorbidities, unresolved traumatic anxieties, a trust in the embedded behavioral health provider on the basis of the warm handoff they received from their medical provider down the hall, and a host of severe and persistent mental illnesses that eschew light tweaks and quick fixes, especially in more rural areas that have a less-than-robust specialty behavioral health system of care.

    We become too focused on standardization when judgment, flexibility and access are what are most needed. While it is tempting and perhaps easier to impose standard clinical protocols, the most effective treatment we can provide is patient centered. When we find ourselves transforming, too often it involves new and more protocols aimed at efficiencies rather than efficacy.

    Donald Berwick, a Harvard-based quality-improvement expert noted for employing and promoting transformation efforts in the evidence-based medicine movement, wrote in 2005 that we had “overshot the mark” and turned evidence-based practice into an “intellectual hegemony that can cost us dearly if we do not take stock and modify it.”11 In 2009, advocating for “patient-centered care,” he declared, “evidence-based medicine sometimes must take a back seat.”12

    In the end, it is important to acknowledge a few final considerations relevant to the question at hand. A small minority of those studied by Milliman drove a significant majority of total costs, and most of that small minority had comorbid behavioral health conditions, with no or minimal spending on behavioral health services for that group. While the methodology did not allow researchers to attribute causality between behavioral health conditions and very high medical spending, it is sufficiently clear through the methodology of common discernment that access to services for behavioral health conditions prevalent among the population is an important strategic priority in managing total healthcare costs and, more importantly, maximizing positive outcomes for patients.

    While it may be disputed whether this conclusion is “value-based,” it is indisputably values-based. Expanding access to care should be our No. 1 priority. Anything that hinders that is a distraction.


    1. Stoddard D, Gray TJ, Melek SP. “How do individuals with behavioral health conditions contribute to physical and total healthcare spending?” Milliman Research Report. August 2020. Available from:
    2. National Alliance of Healthcare Purchaser Coalitions. “Study reveals individuals with behavioral health conditions in addition to physical conditions drive high total healthcare costs; Small portion spent on behavioral health treatment, vast majority spent on physical treatment.” Aug. 13, 2020. Available from:
    3. Melek SP, Norris DT, Paulus J, Matthews K, Weaver A, Davenport S. “Potential economic impact of integrated medical-behavioral healthcare: Updated projections for 2017.” Milliman Research Report. January 2018. Available from:
    4. Patient Protection and Affordable Care Act of 2010, Pub L No. 111-148. 2010. Available from:
    5. Bouchery E. “Mental health treatment need and treatment system capacity.” ASPE Issue Brief. March 2021. Available from:
    6. Ibid.
    7. Samson LW, Tarazi W, Turrini G, Sheingold S. “Medicare beneficiaries’ use of telehealth in 2020: Trends by beneficiary characteristics and location.” ASPE Issue Brief. December 2021. Available from:
    8. American Psychological Association. “Worsening mental health crisis pressures psychologist workforce: 2021 COVID-19 practitioner survey.” Oct. 19, 2021. Available from:
    9. Birch S. “Transforming the health care delivery system.” Pediatric Population Health Forum 2018, Yakima, Wash.: Washington Chapter of the American Academy of Pediatrics. September 22, 2018.
    10. Edwards BG. “Bringing balance to primary care behavioral health and specialty behavioral health.” MGMA Connection. March 12, 2020. Available from:
    11. Berwick D. “Broadening the view of evidence-based medicine.” Quality and Safety in Health Care, 14. 2005, 315-316.
    12. Berwick D. “What ‘patient-centered’ should mean: Confessions of an extremist.” Health Affairs, 28(4), 555-565.

    Blake Griffin Edwards

    Written By

    Blake Griffin Edwards, MSMFT, LMFT


    Explore Related Content

    More Insight Articles

    Ask MGMA
    Reload 🗙