Integrated Healthcare Delivery: Ten Differences That Make a Difference

Bern Heath, Ph.D., CEO, Axis Health System, April, 2012

Standing on the shoulders of those that have gone before us, (especially Cherokee Health Systems) Axis Health System has become a leader in the integration of healthcare. We have moved the concept and practice of integrated healthcare materially forward and in doing so have come to some key principles and reached some understandings that set our efforts apart from our earlier conceptualizations and much of the integration going on in Colorado and the nation. More importantly, we believe that these principles and understandings provide guidance for the continued evolution of integrated healthcare. Our concept of integrated, whole person, healthcare is population based, patient-centered, outcome driven and technology enabled. Health and wellness are not program components; they are a central tenet of our approach.

1. And All for One

Our patient-centered, outcome driven, technology enabled healthcare is care team delivered. Our concept of care teams is different from the long-standing multidisciplinary team and more current physician led versions in that it is not led by a physician or any assigned discipline. With the cultural “team” structure and institutional cross-training all participants are expected to contribute within and outside of their disciplines in the best interest of patient care. Patients are assigned teams in which they actively participate and the entire team is equally responsible for all aspects of the patient’s health.

2. Stepped Care

Barbara Mauer’s Four Quadrant model, introduced in 2002 as an NCCBH white paper and revised in 2006, contributed substantively to the early discussions and conceptualizations of healthcare integration. It is logical, it has the advantage of being readily understood in the context of traditional community mental health service delivery, and it is a useful starting place for the transformation to co-located, and in some instances more integrated, levels of service.

The Four Quadrant model places people in specific care settings, based on high or low behavioral health and primary care needs. In contrast, we base our decisions on stepped care principles. In stepped care a patient’s level of care is determined by specific health indicators. Moving levels of care based on indicators at each step, increases effectiveness and lowers the cost of care overall. Other problems with the Four Quadrant model are that it preserves the separation of physical and mental health, does not sufficiently accommodate the complexity of healthcare needs, supports a Specialty Behavioral Health structure that is not helpful in the long run, and keeps us tied to a narrow encounter service structure.

3. Baking a Cake

Using the traditional medical framework of primary, specialty and hospital care it was tempting, indeed seductive to fit behavioral health into this framework as specialty care. As we came to understand, when we conceptualized behavioral health as specialty care we took the wrong fork in the road. Behavioral health needs to be a part of all levels of care. More than that, it cannot be a grey container of salt and pepper at Subway that upon closer examination contains both white and black particles. In the transformation to integration we are baking a cake where antecedent parts of eggs, sugar, flour, and other ingredients no longer exist separately but as a single transformed whole…in our case, whole person healthcare.

4. Six Levels of Collaboration/Integration Implementation

A standardization of integration levels is critical to meaningful dialogue about service design as well as for research. Unless we can reliably categorize integration implementations, we cannot compare or evaluate outcomes adequately. Developers and implementers of integrated care really want to know what implementations and features of integrated healthcare lead to the best health outcomes. As we began to try our hand at integrating healthcare beyond the Four Quadrant Model and the concept of Specialty Behavioral Health, we quickly came to understand that there were many ways and levels at which healthcare could be integrated.

Doherty, McDaniel & Baird (1996) proposed the first classification of healthcare settings by level of collaboration and integration. That 5-level taxonomy has served us well over the years but is now being updated to improve our ability to make discriminations between levels and to add a sixth level; Full Collaboration in a Transformed Fully Integrated Healthcare System. The sixth level fills out the continuum with an integrated implementation in which providers have overcome barriers and limits imposed by traditional and historic service and funding structures. In this sixth level, antecedent system cultures dissolve into a transformed system viewed by staff and patients alike as a single system treating the whole person.

5. What We Have Here is a Failure to Communicate 

In order to support increasingly greater levels of integration, providers need to have easy and complete access to the full healthcare record. Yet, none of the major, enterprise-wide, electronic health records (EHRs) for behavioral health or for primary care can talk to, or communicate with, the other. These systems were designed for 1990’s and early 2000’s practices when integration was not seriously contemplated. With the Regional Health Information Exchanges in their infancy, this promising resource has not yet matured to a level that will solve this problem. In consequence, we have needed to create our own, internal Health Information Exchange (HIE) which is called the Personal Health Profile TM. It pulls select data from the behavioral health and physical health EHRs, as well as our population based screening tools to generate a profile that is readily accessible to all practitioners and meaningful to the patient as well.

6. Money Makes the World Go Around

Fragmented and siloed health records are not the only thing getting in the way of integrating healthcare. So, too, are funding structures. Traditional fee-for-service (FFS) billing does not support higher levels of integrated care. FFS structures stumble over same day billing restrictions and do not reimburse for consultations without the patient present, electronic contacts or a large volume of care management all of which are essential for improved health outcomes in an integrated healthcare system. FFS structures inadvertently provide incentives for ineffective, fragmented and churned visits rather than active management of care focused on health outcomes. Global payment structures, for example capitated programs, hold providers accountable for outcomes but do not limit or compromise the structure of service delivery which can be designed for maximum local impact. However, a cautionary note is called for here. Many states want it both ways. They want to provide a global payment, but require an accounting of individual services to justify the global payment. This is the worst of both worlds and will undermine effective, innovative integrated care and the global funding needed to support it. Global payment structures can save additional administrative costs both for the provider and the administrator by avoiding this “FFS in capitated clothing” pitfall.

7. Savings, Savings, Who’s Got the Savings?

In conjunction with global payment, integrated healthcare is our best vehicle for bending the cost curve (one of the 3 legs of Berwick’s Triple Aim). It does so by materially reducing costs through more effective primary care (better health outcomes). Cost savings inure not directly to the primary care sites, but to insurance companies who save by reduced hospital care and reduced needs for specialty care. Exercise programs reduce weight, risk for Type 2 diabetes, depression, hypertension and associated coronary disease, among other conditions. Creating cost savings for insurance companies provides integrated settings the opportunity for funding through a risk sharing/gain sharing compensation structure that returns a share of those savings to primary care which can then invest them in services, supporting a continued evolution of effective care.

8. Where not What (Site vs. Algorithm)

Both the Global Payment and the risk sharing strategies are built upon a fundamental shift in accountability mechanism from encounter based service data to population based outcome data. The Colorado Behavioral Healthcare Council’s (CBHC) Subcommittee on Integrated Data was tasked with deciding how to define integrated care services from a data standpoint. In the process of that effort the subcommittee reached a profound understanding. We best define integrated care by site/location rather than an algorithm of service code combinations. That is, a single service (e.g., blood pressure check or depression med check) provided in an integrated site/setting is considered an integrated service because it is provided in the context of that site’s whole person care. Conversely, multiple services provided in a single visit are not by definition integrated care as these services could be (and all too often are) provided by separate professionals without collaboration or integration. This definition of integration by site vs. algorithm sets the table for population based outcome assessments.

9. Outcome is the Thing

Our contractual compliance requirements and our internal productivity efforts are built upon an encounter, service code data structure. This structure also supports fee-for-service billing which (as noted earlier) does not support higher levels of integrated care delivery. If we move from FFS to a global payment structure it shifts compliance/accountability and productivity to population based (by site, payor source or both) outcomes. In short, we need to stop measuring units of service (volume) and instead measure the impact of our services (outcome). This better aligns our work to its goal (improved health) and provides a foundation for establishing cost savings – essential to making the case for healthcare reform and state innovation. The most challenging aspect is in moving accountability from individual staff productivity to population outcomes…yet this is the direction we believe aligns with integration, cost savings and improved health outcomes. We believe that the very heart of making the case for patient centered, integrated whole person healthcare is the measurement of population (by site and by payor source) health outcomes.

10. Just Because We Can, Doesn’t Mean We Should

A final note for consideration…Our concept of integrated healthcare is transformative. Axis is trying to shift the paradigm, not make incremental changes to the existing paradigm. We are regularly offered (through traditional practice management and emerging practices) opportunities for incremental improvements to a traditional community mental health model. While these opportunities can genuinely improve services there are times when we believe we should not proceed along those lines and instead focus on system change (i.e., the move to integrated healthcare) that would more broadly, reliably, and permanently address the problem. For example, instead of developing a centralized system to have local primary care physicians call a state-wide number for a psychiatric consult to prevent medication conflicts, we should focus on increased integration of local primary care sites that would accomplish the same thing more thoroughly and effectively. We must always keep our heads up and eyes on the larger goal of creating a broader, integrated and more effective healthcare system.