Payment Reform in the Patient-Centered Medical Home: Enabling and Sustaining Integrated Behavioral Health Care

Abstract

The patient-centered medical home (PCMH) is a promising framework for the redesign of primary care and more recently specialty care. As defined by the Agency for Healthcare Research and Quality, the PCMH framework has 5 attributes: comprehensive care, patient-centered care, coordinated care, accessible services, and quality and safety. Evidence increasingly demonstrates that for the PCMH to best achieve the Triple Aim (improved outcomes, decreased cost, and enhanced patient experience), treatment for behavioral health (including mental health, substance use, and life stressors) must be integrated as a central tenet. However, challenges to implementing the PCMH framework are compounded for real-world practitioners because payment reform rarely happens concurrently. Nowhere is this more evident than in attempts to integrate behavioral health clinicians into primary care. As behavioral health clinicians find opportunities to work in integrated settings, a comprehensive understanding of payment models is integral to the dialogue. This article describes alternatives to the traditional fee for service (FFS) model, including modified FFS, pay for performance, bundled payments, and global payments (i.e., capitation). We suggest that global payment structures provide the best fit to enable and sustain integrated behavioral health clinicians in ways that align with the Triple Aim. Finally, we present recommendations that offer specific, actionable steps to achieve payment reform, complement PCMH, and support integration efforts through policy.

Editor’s note.

This article is one of five articles in a special issue of American Psychologist titled “Patient-Centered Medical Homes: The Role and Value of Psychology” (January, 2017). Anne E. Kazak, Kimberly Hiroto, and Nadine J. Kaslow provided scholarly lead for the special issue. Susan H. McDaniel served as action editor for the articles in the issue.

Never before in our nation’s history has there been such a largescale attempt to change clinical health care delivery, while placing the patient in the forefront of redesign efforts (Berwick, Nolan, & Whittington, 2008; Moses et al., 2013). The dominant health policy of the land, the “Triple Aim,” challenges clinicians, practices/health system leaders, and state officials to consider how to reorganize health care delivery to improve outcomes, decrease cost, and enhance the patient experience (Berwick et al., 2008). This challenge could not come at a more appropriate time in the United States. U.S. health care costs are between $2.8–$2.9 trillion annually and account for nearly 18% of the Gross Domestic Product (GDP; Moses et al., 2013); yet, this level of spending is not associated with higher performance outcomes or quality. According to the World Health Organization, the United States ranks 37th in the world in health care quality (Bach & DeLisa, 2014). While some of these issues may improve subsequent to the implementation of the Patient Protection and Affordable Care Act (ACA; Goodson, 2010; Young & DeVoe, 2012), true transformation requires that health care stakeholders in the United States address these issues of cost and quality.

Creating a more efficient and effective health care delivery system, with primary care at the base, is central to helping contain costs and improve outcomes (Shi, 2012; Starfield, 2001; Starfield & Shi, 2004; Starfield, Shi, & Macinko, 2005). Over 200 domestic and international studies document the critical role primary care plays in controlling health care costs and ensuring quality (Shi, 2012). In the United States, primary care is currently undergoing a substantial redesign through implementation of the patient-centered medical home (PCMH). As defined by the Agency for Healthcare Research and Quality (AHRQ), the PCMH framework has five functions and attributes: comprehensive care, patient-centered care, coordinated care, accessible services, and quality and safety (AHRQ, n.d.). In 2007, four major medical societies outlined joint principles of the PCMH, which are: (a) improve the health of populations, communities, families, and individuals, and (b) increase the value of health care through enhancing access, comprehen-siveness, integration/coordination and relationships involving sustained partnership (American Academy of Family Physicians (AAFP), 2008; American Academy of Family Physicians (AAFP), American Academy of Pediatrics (AAP), American College of Physicians (ACP), & American Osteopathic Association (AOA), 2007; Kazak, Nash, Hiroto, & Kaslow, 2015; Stange et al., 2010). Traditional primary care delivery was built around the relationship (continuity) with a medical doctor and comprehensiveness (Shi, 2012; Starfield & Shi, 2004); however, care in a PCMH is intended to enhance these functions by focusing on better coordination of services for the patient across medical team members.

Until recently, behavioral health has not been a focus within the PCMH (Ader et al., 2015; Brown Levey, Miller, & deGruy, 2012; Kessler et al., 2014; McDaniel & Fogarty, 2009). Although the Institute of Medicine declared physical and behavioral health to be inseparable from primary care more than 20 years ago (Donaldson, Yordy, Lohr, & Vanselow, 1996), only recently have numerous local, regional, and national initiatives enabled practices to start integrating behavioral health in primary care (Cohen et al., 2015a; Miller, 2015). A growing movement focuses on implementing robust integrated care—where behavioral health and primary care clinicians practice collaboratively to deliver patient-centered care (Brown Levey et al., 2012; Cohen et al., 2015a; Miller, Gilchrist, Ross, Wong, & Green, 2016; Miller, Mendenhall, & Malik, 2009; Miller, Talen, & Patel, 2013). The AHRQ Lexicon defines the integration of behavioral health and primary care integration as:

The care that results from a practice team of primary care and behavioral health clinicians, working together with patients and families, using a systematic and cost-effective approach to provide patient-centered care for a defined population. This care may address mental health, substance abuse conditions, health behaviors (including their contribution to chronic medical illnesses), life stressors and crises, stress-related physical symptoms, and ineffective patterns of health care utilization.

(Peek & National Integration Academy Council, 2013, p. 1)

As more practices and health systems move toward models that integrate behavioral health into primary care, there is need for payment reform that allows for clinicians to work more closely as a team in an integrated and collaborative fashion (Hubley & Miller, 2016). However, implementing these central tenets of the PCMH in real-world settings frequently proves problematic because clinical workflows and payment structures prohibit coordination and collaboration between clinicians caring for patients with behavioral health and physical health needs (Davis et al., 2015; Gunn et al., 2015; Massa, Miller, & Kessler, 2012).

Clinical and Payment Fragmentation

Clinical Fragmentation

The U.S. health care system has been built around a false dichotomy separating the provision of behavioral and physical health care (Institute of Medicine, 2001). The resulting care fragmentation requires many patients see multiple clinicians who may treat one health care need (e.g., diabetes) with little attention to other aspects of the patient’s life (e.g., depression, job loss, unaffordable housing, divorce). The training, education, and practice of medical and behavioral health clinicians in siloed clinical environments reinforces this fragmentation (Blount & Bayona, 1994; Blount, DeGirolamo, & Mariani, 2006; Blount & Miller, 2009; Hall et al., 2015; Institute of Medicine, 2001). Differences in the traditional scheduling patterns for primary care clinician appointments (i.e., focused 15-min encounters) compared with behavioral health clinicians (i.e., 45–50 min, multiple sessions) further create cultural distinctions in how care is structured (Davis et al., 2015). Clinician-centric care, rather than team-based or patient-centered care, has led to a less than ideal patient experience in obtaining “whole person” care (Epstein & Street, 2011) or achieving the Triple Aim. Driving much of this behavior are antiquated policies and payment models that are often not in support of the integrated team.

Payment System Fragmentation

Historically, medical and behavioral health services have been paid through two distinct channels and budgets that have utilized two separate sets of billing codes, reimbursement pathways, and reporting requirements (Kathol, Butler, McAlpine, & Kane, 2010; Knesper, Wheeler, & Pagnucco, 1984; Mauch, Kautz, & Smith, 2008). While some settings like the Veteran’s Administration and the Department of Defense have not paid for care in this fashion, the vast majority of nongovernmental health care delivery has done so (Kathol et al., 2010; Pomerantz et al., 2010). Traditionally medical and behavioral health clinicians have been reimbursed for specific services rendered during multiple face-to-face patient visits rather than using an outcomes-based approach, which might also allow comprehensive and coordinated efforts to address issues in one or a brief series of encounters using a format best suited to patient needs and preferences (e.g., in person or using remote monitoring technologies). In addition, some state policies limit billing for medical and behavioral health services on the same day, encouraging scheduling practices that are not supportive of patient-centered care (Miller et al., 2013). This volume-oriented reimbursement also has the effect of “crowding out” clinical interventions for individuals who require more focus, as well as prohibiting proactive planning, targeting, and follow-up that is necessary to improve the health of a population over time (Schroeder, Frist, & the National Commission on Physician Payment Reform, 2013).

A compounding challenge is that traditional payment strategies are highly variable across states and commercial plans because of differences in contracting for behavioral health services. This leads to fragmentation in services as well as an increase in administrative costs. For example, in Medicaid, behavioral health benefits and payments are most often treated as a separate line of service from the overall Medicaid medical benefit (Kathol et al., 2010; Kathol, Degruy, & Rollman, 2014; McAlpine & Mechanic, 2000). Contrast this with states that have chosen to consolidate various aspects of behavioral health services with physical health services, with the end goal of offering an integrated benefit. For example, Arizona has chosen to consolidate the behavioral health benefit within their overall health benefit for individuals with serious mental illness, increasing the likelihood that a Medicaid beneficiary could receive integrated care (Arizona Health Care Cost Containment System, n.d.). While this approach can be politically challenging, it does offer the most promise to seeing integrated services come to fruition in clinical practice; however, this approach is a rarity, and most states find that the responsibility for behavioral health is overseen by multiple separate agencies (Bachrach, Anthony, & Detty, 2014).

Additional fragmentation in states is created because of the allocation of Medicaid mental health service dollars to one behavioral health agency, which is tasked to provide all behavioral health services to Medicaid patients in a region. Whether managed behavioral health organizations are “carved-in” (owned by a medical insurer but sold as an independent insurance product) or “carved out” (owned by a standalone behavioral health insurance vendor), segregated benefit management by states has downstream unintended consequences in clinical delivery (Kathol et al., 2010). However, it is important to note that it is the independent claims adjudication process, not the owner of the behavioral health product or service, that creates clinical and financial separation at the patient and clinician levels. Specifically, contracted behavioral health organizations are required to complete certain intake processes and to meet the needs of the most complex patients, which can lead them to prioritize care for care for complex patient leaving little money for prevention work or care for less complex patients. These disjointed and fragmented financial structures construct unnecessary barriers to whole-person care (Kathol et al., 2010; Kathol et al., 2014; Kessler, Stafford, & Messier, 2009; Miller et al., 2013).

Commercial payers such as Anthem, United, and Blue Cross Blue Shield also have separate companies that manage the behavioral health benefits for their commercial product. This culture of separating behavioral health financing from medical financing has historical roots. Since the 1960s, commercial plans have treated behavioral health services as a separate benefit, making integrated care a challenge (Kathol et al., 2010). While there have been policies that aim to enhance insurance to support behavioral health services (e.g., Mental Health Parity Equity and Addictions Act), these policies do not change the overall payment system for behavioral health, but rather bring into parity the mental health benefit of an insurance product with the medical benefit (Cunningham, 2009; The Center for Consumer Information & Insurance Oversight, 2008). For clinical systems to integrate behavioral health into primary care, the financial reimbursement structures must also transform to support this integrated system, which is no small task given the different paradigms under which each currently operates.

The Value Opportunity for Integrated Care

One way to begin to change the current payment paradigm for how both public and commercial insurance plans provide behavioral health care is to highlight data showing the cost savings potential when care is integrated. For the field of psychology and behavioral health in general, these data are not new but are highly compelling. For years the cost offset literature has shown how psychotherapy can decrease medical utilization and improve outcomes (Blount et al., 2007; Eells, 1999). One of the most compelling examples of cost offset through integration comes from the Cherokee Health Systems (CHS) in Tennessee. The CHS serves approximately one tenth of the individuals in their region with a high proportion of Medicaid patients. The CHS utilizes an integrated PCMH model wherein psychologists address the behavioral health needs of the patients and consult with psychiatrists as needed to comprehensively address the more complex mental health needs of patients. Their results demonstrate increased primary care utilization, along with a 68% reduction in emergency care, 42% reduction in specialty care, and 37% reduction in hospitalizations when compared with other clinicians in the region. In addition, the CHS saw an overall reduction of 22% in costs over a 3-year period (Franko, 2015), demonstrating that employing a PCMH framework with integrated behavioral health can lead to substantial improvements in care and reductions in cost.

The cost of comorbid chronic medical conditions with behavioral health disorders is significant. For instance, the per member per month (PMPM) cost of asthma care is over three times higher when the patient has a comorbid serious and persistent mental illness or substance abuse disorder (Melek & Norris, 2008; Petterson et al., 2008). A report completed by Milliman for the American Psychiatric Association compares the health care costs PMPM of commercially insured members with various medical conditions without a comorbid behavioral health disorder to similar members who have a comorbid serious and persistent mental illness (SPMI), a non-SPMI mental health disorder, or a substance use disorder (SUD; Melek, Norris, & Paulus, 2014). The value opportunity for health care cost savings is the difference in health care costs between patients with no mental health or substance use disorders and those with identified behavioral health need. For example, a patient with diabetes and complications costs $1,821 PMPM with no mental health diagnosis, $3,366 with SPMI, $2,681 with nonsevere mental health diagnosis, and $3,678 with a SUD. The additional health care costs resulting from these behavioral health care comorbidities are largely medical costs, not behavioral health care costs, regardless of the severity of the behavioral health condition (Melek & Norris, 2008). This is because untreated or ineffectively treated behavioral health conditions lead to increased medical treatment and associated costs, not increased behavioral health care costs. For example, individuals within a commercial health plan who have co-occurring behavioral health conditions utilize significantly more inpatient general (nonbehavioral health) hospital and physical health emergency care costs as a result of the impact that persistent comorbid behavioral health symptoms have on their self-care, medical treatment adherence, diet, exercise, and other healthy lifestyles factors (Melek et al., 2014). Similar value opportunities exist for Medicare and Medicaid populations (Melek & Norris, 2008; Melek et al., 2014).

The aggregated cost of these value opportunities, including insured commercial, Medicare. and Medicaid populations in the United States, amounts to about $300 billion annually (Melek & Norris, 2008; Melek et al., 2014). Programs that have integrated behavioral health have been shown to save 9%–16% of this value opportunity in absolute health care savings, or an annual savings of $27–$48 billion (Melek et al., 2014). These data show that payment reform models could foster effective clinical and economic outcomes based on the integrated delivery of physical and behavioral health care. In the following sections, we describe current financial models and review alternative payment strategies that are currently being implemented in certain contexts to enable and sustain behavioral health within the PCMH.

Payment Reform

A number of payment models provide promise for behavioral health integration in the PCMH. These models often offer a range of behavioral health services in an integrated fashion and provide alternatives by incentivizing quality over quantity. For behavioral health clinicians to integrate and provide whole-person care alongside their interprofessional counterparts, a comprehensive understanding of payment models is necessary. This section describes current payments models (i.e., FFS) and provides examples of alternative payment strategies (e.g., modified FFS, bundled payments, global payments). Each will be described from the behavioral health clinician perspective, outlining their strengths and weaknesses, which can also be seen in Table 1.

Table 1.

A Comparison of Payment Models and the Associated Pros and Cons in Support of Behavioral Health Integration

Payment model	Description	Pros	Cons
Fee for service (FFS)	FFS system uses a retrospective payment where each item of service provided is reimbursed based on certain billing codes that are submitted as a claim to the health insurance company; behavioral health payments primarily come from a separate entity within an insurance company	Behavioral health services can receive compensation for their mental health services	Relegates behavioral health clinicians to deliver more traditional mental health interventions often independent of the team
Modified fee for service	Oftentimes a hybrid of FFS and non-FFS payments. For example, pay for performance (see below) and partial capitation.	Increases the ability of PCMH to engage in some value-based rather than solely volume-based care.	Still makes behavioral health its own service line and intervention rather than a part of the team
Pay for performance (P4P)	P4P holds clinicians accountable for the outcomes their care delivers. Such initiatives aim to incentivize processes and outcomes of care	Increases the likelihood that certain behavioral health conditions are addressed (e.g., depression)	Payment may not be sufficient to support the behavioral health member of the primary care team
Bundled payments	Bundled payments reimburse for a discrete course of treatment rather than paying for each discrete clinical interaction and procedure	Supports more of the team approach to specific conditions	Behavioral health often not considered as a part of the payment bundle
Global payments	A global payment system, or a capitated system, pays a predetermined per person rate to healthcare organizations, regardless of the delivered services	When behavioral health is a part of the service expectations through the global payment, there can be seamless and unfettered access to behavioral health; behavioral health becomes natural extension of primary care team	Challenge associated with assuming risk for patients with behavioral health; practice change and transformation

Traditionally, the U.S. health care system has utilized a FFS reimbursement structure. The FFS system uses a retrospective payment where each item of service provided is reimbursed based on certain billing codes that are submitted as a claim to the health insurance company. This system inherently incentivizes clinicians to deliver more care (Gosden et al., 2000), and has been linked to increased procedures based on the availability of clinician expertise/skills (Evans, 1974). As such, the FFS model is largely incongruent with Triple Aim goals and the fundamental purpose of the PCMH structure; it supports volume over value. Aside from the fact that FFS incentivizes clinicians to do more with little to no consideration for outcomes, other issues arise when one tries to leverage payment for behavioral health in integrated settings. For example, psychologists may only bill for patients with diagnosable mental health conditions, making some of the psychologists’ services like health behavior change interventions, consultation, and coordination ineligible for billing. FFS is one rate limiting factor on what psychologists can do in integrated PCMH settings as it often dictates what a clinician can and cannot do for the patient; this often leads to more traditional, 50-min long mental health encounters and limits nonmental health interventions.

Behavioral health clinicians have pushed for billing practices that can better support PCMH efforts. One example is the Health and Behavior Assessment codes. These codes allow nonphysicians, like psychologists, to bill for brief interventions that pertain to a patient’s health behavior change. These codes do not require an independent mental health diagnosis and are allowed on the same day of billing by other clinicians (Butler et al., 2008; Kessler, 2008). There has been much written in the literature about the role Health and Behavior Assessment codes can play in health care delivery to help advance more behavioral health in nontraditional behavioral health settings (Kessler, 2008; Kessler et al., 2014). For example, a psychologist working with a patient in primary care who is looking to lose weight may be an ideal candidate for these codes. However, while Health and Behavior Assessment codes create a revenue stream that can support some behavioral health services, they are within a FFS framework and accordingly are still encounter-based and limited in scope. Said differently, these codes remain volume- rather than value-based. The FFS model does not typically allow for the full range of patient and care team interactions that promote better patient support and outcomes, including facilitation of care transitions, informal consultations, and team huddles for planning purposes. Other models are needed to facilitate the full range of interactions.

Alternative Payment Structures

Alternative payment structures to FFS include Modified FFS, bundled payments, and global payment (i.e., capitation) models. To fully transform the health care environment in the United States to include the behavioral health expertise, payment structures need to support behavioral health clinicians and their full range of services within primary care settings. In addition, pay for performance techniques may supplement each of these structures to incentivize and reward clinicians for obtaining designated performance standards and patient outcomes (Unützer et al., 2012). Of note, there is a very limited literature base for financially sustaining behavioral health in primary care using alternative payment methodologies. Just a handful of papers discuss financing of integrated behavioral health (e.g., Hubley & Miller, 2016; Kathol et al., 2010; Kathol et al., 2014; Mauch et al., 2008; Miller, Gilchrist, Brown Levey, et al., 2016); thus, different payment models that support integrated behavioral health is an area ripe for research.

Modified FFS

Modified FFS is a general term that encompasses any modifications to the current FFS system to allow for payment that is not volume-based, such as a structured payment given to the clinician that is tied to the patient and not the visit. One example of a modified FFS approach is Pay for Performance (P4P). P4P initiatives are increasingly popular components of financial reform, particularly because the ACA has encouraged expansion of P4P approaches. Building off a FFS framework, P4P begins to hold clinicians accountable for what they are doing with the care they deliver. Such initiatives aim to incentivize processes and outcomes of care and can be applied under any additional alternative payment models. However, to date, there are no uniform definitions and mechanisms for P4P because these efforts differ widely in payment structure details and timing. For example, some programs supplement a capitated payment system with P4P by offering bonus payments to clinicians who meet certain care requirements or outcomes (Beaulieu & Horrigan, 2005; Hillman et al., 1998; Roski et al., 2003), while others offer an end-of-year bonus for achievement of selected outcomes (Hillman et al., 1999).

A P4P initiative would be a good fit for integrated behavioral health and primary care settings if behavioral health factors were incorporated into process and outcome measures of care. For example, one process measure could include the percentage of patients who are screened with the Patient Health Questionnaire-9 (PHQ-9) and the health outcomes associated with treatment subsequent to screening (e.g., depression-/anxiety-free days or reduction in medical service use). The P4P model already has shown some traction around behavioral health in the PCMH including widespread adoption in Washington State (Unützer et al., 2012). In the Washington example, key quality indicators for depression, including time to follow-up care and time to depression improvement, improved significantly as a result of tying payment to the same quality indicators. Said differently, the inclusion of P4P for depression identification and treatment significantly improved performance, indicating that P4P can have a dramatic effect on behavioral health in primary care settings.

Modified FFS, in particular P4P, is not without its critics. While modified FSS approaches may be quicker to gain traction because of their grounding in FFS, the dominant payment mechanism, research examining P4P has yielded mixed results. While some studies have found evidence for P4P promoting improvements in health care quality measures (e.g., improving influenza vaccination rates and cervical cancer screening rates), other studies have not found improved care quality or patient outcomes (Eijkenaar, Emmert, Scheppach, & Schöffski, 2013). Furthermore, few studies have examined behavioral health-focused quality measures, and significant concern exists about appropriately measuring these outcomes (Liptzin, 2009). In addition, if not appropriately implemented, P4P initiatives may have unintended consequences, such as physician resentment about P4P and disenrollment of noncompliant patients to avoid financial penalties (McDonald & Roland, 2009).

Partial capitation or care management payment could also fall under the model of modified FFS. Partial capitation is where some health care services are funded on a per capita basis, while others are supported through FFS payments. In other words, clinicians or practices would receive a lump sum of money to cover general services that may enhance the overall care while still billing out FFS for traditional clinical interventions. Such an alternative could enable PCMHs to invest in building the infrastructure and workflows for team-based care. However, to the extent that encounter volume remains the fundamental economic driver, the value of these investments may not be fully realized. Even with a sufficient, demonstrable return on investment, support for team-based care and population-oriented payment models is often abandoned by payers and employers (Kathol et al., 2010). An example of this approach is found in the Centers for Medicare and Medicaid Innovation Comprehensive Primary Care Initiative (CPCI; Centers for Medicare and Medicaid Services, n.d.). Under CPCI, primary care practices receive population-based (or PMPM) care management funds, which they can use to enhance a variety of services including pharmacy and behavioral health. The benefits are that primary care practices have more capital with which to be flexible in offering services to their patients. However, as with most programs where the PMPM payment is tied to care management, the dollars are in some cases insufficient to cover much beyond care management.

Bundled Payments

Bundled payments reimburse for a discrete course of treatment (e.g., a joint replacement procedure, cardiovascular surgery, drug or radiation intervention) rather than paying individually for each clinical interaction and procedure (Hussey, Ridgely, & Rosenthal, 2011). Bundled payments usually occur in the form of a lump sum payment for all professional and facility-based services that are projected to be medically necessary for treatment of an illness episode for a fixed time period. If the cost of the episode of care is less than the bundled payment, the clinician/hospital keeps the additional money; however, if the cost exceeds the payment then the clinician/hospital is financially responsible for the excess cost. In a PCMH setting, a bundled payment may be based on diagnosis rather than an illness episode. For example, a bundled payment may be for diabetes-related services provided to a patient with Type 2 diabetes within a specified period, such as follow-up visits with the physician for medication adjustments, dietary advice, and smoking cessation counseling. The goal of a bundled payment is to improve care quality by providing financial incentives to clinicians to eliminate clinically ineffective or duplicative services (Miller, 2009), and, indeed, research demonstrates that significant health care cost saving could be gained by undertaking a bundled payment approach (Cutler & Ghosh, 2012). In addition, a bundled payment system is thought to encourage collaboration and coordination of care, because multiple clinicians are jointly accountable for the care of each patient (Hussey et al., 2011), and there is some evident to support the increase in multidisciplinary collaboration with introduction of bundled payments (Struijs & Baan, 2011).

A bundled payment system could provide some support to integrated behavioral health and primary care environments; however, it is not without its shortcoming. This system often has FFS underpinnings, such that FFS income by medical clinicians is pooled to cover total program costs (including behavioral health), and, therefore, fails to incorporate the instrumental role of behavioral health clinicians on the team. For example, if a patient experiences a congestive heart failure episode and is admitted, the bundled payment likely would not include the outpatient follow up to address behavioral health and social issues, such as depression-based treatment nonadherence. Additional difficulties exist in determining the proper bundle of care for mental health episodes and for illness episodes for patients with comorbid mental and medical health diagnoses. For example, what level, frequency, and duration of care is appropriate for a patient in primary care with comorbid borderline personality disorder, posttraumatic stress disorder, and diabetes? Determining the appropriate care bundle and bundled payment for a hip replacement is more straightforward than determining the same for comorbid mental and chronic medical health conditions. Last, bundled payments incentivize provision of high quality care and minimization of redundant care and services within an episode, but the model does not incentivize the reduction of number of episodes (Miller, 2009), which is a key focus in delivering quality PCMH-based care.

Global Payments (Capitation)

A global payment system, or a capitated system, pays a predetermined per person rate to clinicians, practices, or health care organizations, regardless of the delivered services. Here the onus lies on the health care team to provide comprehensive care while minimizing unnecessary and redundant testing and limiting cost. Global payment structures offer latitude to utilize a variety of clinicians to meet the needs of the patient and, therefore, do not stifle interprofessional integration among behavioral health and primary care clinicians. The global budget allows the practice to create the best team and deliver whatever services are most appropriate for the population they are accountable for. Organizations with this financial arrangement often seek to maximize the teams that need to be in place to achieve the optimal outcomes or financial targets (Song et al., 2012). Some research has specifically demonstrated the global payment methodology may result in greater reduction in health care costs than a bundled payment approach (Cutler & Ghosh, 2012).

Within a PCMH, a capitated model would allow for a practice to choose how they leverage their financial resources in service to their patients. When given a budget to manage the population they are responsible for, the PCMH may choose to hire new clinicians like behavioral health clinicians to increase the likelihood that a patient clinically improves, decreases improper utilization of other health care services (e.g., emergency department), and is highly satisfied with his or her health care experience. When contrasted with the other payment options, a global payment system appears to offer the best opportunity to support fully integrated and behavioral health within the PCMH, but only when budgets for medical and behavioral health are consolidated. It is challenging if not impossible for global payment models to function successfully when independent medical and behavioral health service delivery budgets compete (Kathol et al., 2010; Kathol et al., 2014).

Because of limited research examining the impact of global payment models in integrated primary care, researchers the University of Colorado recently conducted a study to examine the clinical and financial outcomes associated with applying global payment methodology in integrated primary care. This study evaluated six integrated primary care clinics over the course of 18 months; three clinics were provided a global payment while three continued with their usual financial practices. The integrated primary care clinics that utilized a global payment model achieved higher behavioral health screening rates for their patients and saved money (via cost avoidance) when compared with the control clinics (Miller, Gilchrist, Brown Levey, et al., 2016). The results of this study are promising and demonstrate that a global payment methodology in concert with an onsite integrated behavioral health clinician may promote both cost-saving and high quality care delivery.

Global payment structures are not without weaknesses, and capitation is not a new concept. In the 1990s, health care underwent a substantial shift toward capitating payments to medical practices in an attempt to allow for more clinical flexibility while containing costs. However, many practices were not capable of managing the risk they accepted for their patients. This was because of multiple factors including an inadequate adjustment for risk, quality, and resource intensity. Because of this, some clinicians lost out financially and had to close their clinics (Kongstvedt, 2013). In addition, some practices tried to “game” the system whereby they limited services offered to patients in an attempt to increase their revenue without spending the dollars they had been given to serve their patients.

Those designing future global payment structures must learn from past mistakes and attend to potential pitfalls. Global payment systems may encourage clinicians to enroll higher patient volumes to increase income without hiring additional staff. Such a system results in impossibly large workloads with resultant nondelivery of needed care (Berwick et al., 2008; Song et al., 2012; Woodward & WarrenBoulton, 1984). The global payment model, not dissimilar to the results of managed care, may lead to rationing of services to contain patient costs. In addition, determining the base amount and necessary adjustments to account for patient complexity and resource intensiveness in PCMH settings is challenging because there is no credible FFS base from which to determine the budget. Most contemporary risk adjustment systems focus on near-term costs and thus fail to account adequately for the patients who are resource intensive and who are at significant future risk but who have not been diagnosed with “acute” diseases and disabilities that are conducive to predictive models. Some of these risks may be mitigated through an appropriate focus on outcomes for agreed measures of performance such as clinical screening and follow-up rates and “attribution” metrics, which focus upon evidence of ongoing contact with the empaneled population of patients. Other structural supports (e.g., creation of an active “patient advisory council” that provides direct feedback on clinical processes and patient experience) may also mitigate the historic pitfalls associated with this model of reimbursement.

Additional Considerations

Despite the recognition that change in payment is needed to better support and sustain integrated behavioral health in the PCMH, there are additional considerations. As with any system, changing one area can have unintended consequences in another. The following considerations are examples of those that may warrant attention as payment reform is pursued.

Changes to Clinical Practice

It is not possible to change payment without changing practice. Psychologists and other behavioral health clinicians, like their medical colleagues, need to change their practice to be consistent with a team-based approach. As described previously, there are some fundamental differences in delivering behavioral health in primary care settings. A recent paper by Davis and colleagues (2015) indicates staffing and scheduling patterns for behavioral health clinicians are associated with the approaches used to deliver integrated care. The authors point out how traditional funding structures frequently contribute to 50-min, prescheduled appointments even in practices working on integrated care. Behavioral health clinicians working in primary care settings need financial structures that enable them to be flexible and implement brief interventions that lead to outcome change, which may be different interventions than those often found in traditional long-term therapy. Payment reform without a recognition that both primary care and behavioral health clinicians will need to change their approach to practice will be insufficient to bring about practice change.

Health Disparities

Experts caution clinicians to be mindful of the potential to widen the health disparities gap as these reimbursement programs may encourage “cherry picking” of patients. If PCMH payments are based on health outcomes, clinicians may be motivated to include and exclude patients in their practice because of demographic and socioeconomic characteristics (Chien, Chin, Davis, & Casalino, 2007). In 2010, approximately 7.2 million individuals experienced unmet behavioral health need, growing from 4.3 million in 1997 (Roll, Kennedy, Tran, & Howell, 2013). Behavioral health disparities are associated with race, age, ethnic, and geographic factors (Bartels et al., 2004; Chow, Jaffee, & Snowden, 2003; Lasser, Himmelstein, & Woolhandler, 2006; Wang et al., 2005; Wells, Klap, Koike, & Sherbourne, 2001), and these gaps may widen if reimbursement programs fail to address them (Farber, Ali, Van Sickle, & Kaslow, 2015).

Culture Change

With an increasing number of primary care practices pursuing PCMH, it will be important to address a variety of health care culture barriers to behavioral health integration. For instance, most medical and behavioral health clinicians historically have been educated and trained in siloed environments without access to interprofessional team development (McDaniel, Belar, Schroeder, Hargrove, & Freeman, 2002). Although local and national training programs are emerging and competencies have been articulated for interprofessional practice (McDaniel et al., 2014), the majority of medical and behavioral health professionals are not trained to work in integrated settings (Beacham et al., 2017; Blount et al., 2006; Blount & Miller, 2009; Hall et al., 2015; McDaniel, 1995; McDaniel et al., 2002, 2014). Moreover, implementation of integrated care models requires robust leadership to change culture, redesigned clinical workflows with associated staffing, and altered scheduling (Davis et al., 2013). Changes to electronic health record structures, and even changing the physical layouts of practices to engender team-based care, are also critical elements (Cifuentes et al., 2015; Gunn et al., 2015). Implementing integrated behavioral health and primary care services in real-world settings requires substantial planning and effort—and is a developmental process and not a one-time change (Cohen et al., 2015a; Hall et al., 2015). Policy changes by states to adopt standardized, evidence-based assessment and reporting requirements could accelerate this process (Miller, 2015). Many states maintain unique, decades-old, “home grown” assessment and reporting requirements that preclude behavioral health clinicians from documenting their work in mainstream electronic health records.

Practice Transformation

Medical practices attempting to integrate behavioral health often do so in the face of other practice redesign efforts. As described above, integrating care requires an investment in quality improvement infrastructure and developing electronic health record systems to allow tracking and monitoring of quality indicators (Centers for Medicare and Medicaid Services, 2011; Pace et al., 2009; U.S. Department of Health and Human Services, 2010). Practices, particularly small and medium independent clinics, may struggle to make these changes. Quality improvement agents, such as externally supported practice facilitators (Nagykaldi, Mold, Robinson, Niebauer, & Ford, 2006), increasingly are being used to support these transformational initiatives, like integrating care. Many facilitators focus on improving the overall delivery processes in primary care (Nutting et al., 2010), and some projects may specifically focus on improving integrated care delivery. Facilitators in these settings may train staff and clinicians on the approaches to integrated care, help practices develop clinical screening processes and data documentation, discuss the relationships between payment structures and behavioral health clinician roles, and coach practice teams as they implement and refine integrated care clinical workflows (Balasubramanian et al., 2015; Cohen et al., 2015b).

Sustainability

Considering that many integrated programs have started with grant funding, a predominant issue with sustaining integrated care is determining how to finance such models beyond the funded project or grant period (Kathol et al., 2010). Indeed one challenge in adopting integrated care is that clinicians may hesitate, wondering how long the change will last and what they will do if they have to return to “traditional” models of care provision. Changes in payment structure and care policies are critical to bridging this gap between aspirations for integrated care in the PCMH and real-world contexts. As outlined previously, there are various models to consider to help sustain integrated behavioral health work, but many of these changes will also require policy change in support of this new approach to care.

Policy Recommendations

Transforming how we pay for health care without simultaneously making clinical delivery changes as well as operational changes are unlikely to lead to successful and sustainable transformation; therefore, the following recommendations should be considered in the context of other reform efforts (Peek, 2008).

Consider What Impact Carving Out Behavioral Health in All Forms and Permutations Does at All Levels and All Policy Processes

As described previously, carving out behavioral health benefits and payments has substantial consequences on patients looking to have a team address the majority of their health care needs. Essentially, this benefit design creates a one-size-fits-all mentality for patients seeking to gain access to behavioral health services from any provider who is not part of their behavioral health contract. Communities, states, and federal policies should all recognize what carving out behavioral health does and does not do in support of broader population health goals inclusive of behavioral health. Said differently, do these carve out policies and contracts simply advance the delivery of specialty behavioral health or assist in the promulgation of general behavioral health across the entire population? All health policies should be assessed in accordance to this question.

Create Incentives to Encourage Primary Care Clinicians to Work With Behavioral Health

Payment models that are mainly driven by individual clinician behavior often do not encourage collaboration. These payment models also do not incorporate the quality and outcome standards that are essential to the future team-based health care. For behavioral health to be more centrally involved in primary care and the PCMH, inclusive payment methodologies are needed. Payment should support integrated medical and behavioral health teams, and incentivize population health outcomes. Integrated behavioral health should be a captured investment opportunity that reduces much larger overall primary care spending, thus contributing to the Triple Aim goals.

One possible first step to support integrated clinical transformation is to begin to hold primary care clinicians accountable for certain behavioral health conditions. When financially motivated through any of the previous payment models, primary care may have the resources to invest in their care teams, perhaps bringing on behavioral health clinicians to help manage the population for which they will be financially at risk. From a policy perspective, on the journey to achieving more population-based payments, considering how to provide incentive to primary care to address behavioral health may be a substantial step toward more widespread integration of behavioral health into primary care.

Study Alternative Payment Models’ Impact on Integrated Teams

While payment reform undoubtedly will be more supportive of a team-based approach, it is important to learn which payment model works best for what team and in what context. Establishing an evidence base around how different models of payment influence the delivery of integrated behavioral health within the PCMH and how the clinical delivery influences the payment models will allow payers and policymakers to be more informed on which approach is best for them.

Pursue and Reinforce Policies in Support of Integrated Teams and Comprehensive Payment Reform at State and Federal Levels

Even the most innovative clinical models require medical practices to be creative in how they fund their behavioral health clinicians. Policies and programs that are built on FFS may only be incremental steps toward more robust transformation. To this end, professional associations committed to integration must simultaneously educate their members about the benefits of a new payment model for their services and push for more substantive policy changes (Manderscheid & Kathol, 2014). To support much of the innovation happening in health care around integration, novel payment methodologies should be considered that are more comprehensive in nature, focusing on the whole of health of individuals. With this end goal in mind, behavioral health can begin to position itself for more involvement in integrated settings.

Develop an Interprofessional Policy Agenda to Advance Payment Models in Service to Integrated Primary Care

All the previous policy recommendations could gain significant power and traction if they were pursued by multiple professional groups simultaneously. Uniting the larger behavioral health and primary care professional community around a shared policy agenda in service to more robust team-based integrated health care could have a significant positive effect. While professional associations may not always agree on what policies to advocate for, it is becoming increasingly apparent that uniting the larger health care field could be to the benefit of all. This policy agenda could be broad, but have specifics related to advancing behavioral health through policy and payment reform. For example, professional societies could collectively call upon the Medicare Payment Advisory Council (MedPAC) to study the impact of integrated behavioral health and primary care for Medicare beneficiaries. Having multiple professional organizations agree to the need to address this area could send a very powerful signal of the problem and potential solution to Congress, the country, and the organizations’ memberships.

Conclusion

As the nation moves forward with PCMH models of care that better integrate behavioral health, the authors challenge health care leaders to define integration and comprehensiveness as including behavioral health services in primary care. In addition to this inclusion as a central tenet to primary care, corresponding payment models must be developed, tested, and implemented to support seamless integration of behavioral health in PCMHs. There is increasing recognition that the future of health care is indeed integrated. The question remains, which approach to payment will best support a more comprehensive and effective model of team-based health care?

Biographies

Benjamin F. Miller

Kaile M. Ross

Melinda M. Davis

Stephen P. Melek

Roger Kathol

Patrick Gordon