Who Should Own the Medical Home?
Approximately 130 million Americans have chronic illnesses, while 70 million have multiple chronic health problems. In 2008, 126 million outpatient follow-up visits provided to Medicare recipients were by family practice or an internal medicine specialist (IMS). Almost 25 percent of these visits were to an IMS, according to an American College of Cardiology (ACC) white paper, "The Patient-centered Medical Home (PCMH)." In addition, 17 percent of the 3.8 million Medicare new patient evaluations were provided by an IMS. Some architects of PCMH models suggest that a specialist team should lead patient management within the medical home. But regardless of who owns the patient—and by proxy, the PCMH—better coordination of care is needed.

The roundtable participants are:

  • Jack Lewin, MD, CEO of the American College of Cardiology (ACC)
  • Donald R. Lurye, MD, CEO of Elmhurst Clinic in Elmhurst, Ill., a National Committee for Quality Assurance-recognized medical home
  • David C. May, MD, PhD, chair of the ACC's Patient Centered Medical Home Committee and cardiologist/president at Cardiovascular Specialists in Lewisville, Texas
  • Mary Norine Walsh, MD, medical director of heart failure and cardiac transplantation at Care Group/St. Vincent's Health System in Indianapolis

Please define the model(s) of the specialist patient-centered medical home.

May: The concept of patient-centered care and the patient-centered medical home does not differ fundamentally between the primary care or specialty models of care. The medical home has its origins in pediatrics as a way to provide an archive for the patient's medical record. Over the years, the model has evolved to a single physician or a care team that is in charge of ensuring that care is patient-centered, well-communicated as well as socially and culturally appropriate.

As there are approximately 70 million people in the U.S. with multiple chronic problems, the concept of the IMS serving as the care team leader for those patients doesn't differ materially between the model of the primary care physician (PCP) being the physician leader and the specialty physician serving that role. The specialty physician in that instance would be responsible for overseeing the patient's general care. It is, however, true that patients in the specialty physician medical home would have advanced chronic illnesses in many instances, such as congestive heart failure or oncology. The physician and care team involved in the management continue to provide the same level of support globally for the patient. However, in the specialty physician environment, there would be the ability to manage that particular ongoing chronic problem.

Walsh: As a heart failure and transplant cardiologist, the medical home model exists already. Patients who have advanced heart failure are functionally in a medical home model, with the medical home being the transplant ventricular assist device center. In cardiovascular medicine, the heart failure model is probably the most robust example of a specialty medical home.

Lurye: From the primary care perspective, the medical home still is seeking a precise definition. The organizations providing recognition or accreditation emphasize different aspects. However, all medical home models should share the joint principles that are espoused by the major primary care societies: namely, a personal physician leading a team, focusing on the whole person to provide coordinated care by communicating well with other practices and entities with an emphasis on measured quality and safety. We hope this all triggers enhanced reimbursement.

However, if a specialist is taking care of specific chronic conditions, is that physician also making sure that patients get Pap smears, colonoscopies and flu shots on time? Is there an awareness of general disease burden within a patient population if it's not related to the condition that the specialist has the expertise to manage? The real answer isn't what you call it, but that the care is well-coordinated and connected to the other providers patients see.

Lewin: There needs to be a specialty version of the PCMH with the same requirements and caveats proposed for primary care. The specialty medical home would benefit patients with difficult-to-manage heart failure, arrhythmia or other cardiac conditions, in some cases congenital heart disease conditions, where it results in almost a reverse handoff. The specialist would see the patient on a more frequent basis to refer to primary care sources for other care services.

Also, in some locations we aren't going to have enough PCPs and teams available, and it might be necessary to have advanced practice nurses providing primary care services. Or, a combination of doctors who have an intern may manage some aspects of patient care. There are many different options, but the Patient Protection and Affordable Care Act recommended only one model.

How important is care integration and data interoperability in the medical home?

Lewin: Through the medical home model, we want to avoid having specialists more expensively manage people with problems that could be well-managed by primary care. Conversely, we need to avoid PCPs holding onto patients who really need to be referred to a specialist.

In an integrated system where PCPs and specialists are linked together by economic bridges as well as registry and quality of care connections that are totally transparent, we should not have this problem. However, most of the U.S. does not currently fall under this integrated model. The patient-centered medical home won't work effectively in a non-integrated environment where doctors have no economic linkage and are not linked by registries. We have some major coordination and efficiency problems that need to be addressed, in addition to ensuring sufficient PCPs and teams to meet the needs of the patients who are going to be identified in primary care settings.

Lurye: The future really belongs to those who can manage information flow. EMRs have the potential to push comprehensive and accurate information to the physician at all times but they must talk to each other, so caregivers don't have to hunt for records. This has worked in some communities, particularly in central Indiana, where there is a pretty robust health information exchange. Emergency rooms, for instance, can access data from a warehouse that is commonly housed by member hospitals and physicians. As Dr. Lewin suggests, it is more difficult to envision how all this works in a non-integrated system.

May: I agree with Dr. Lurye that the physician or the care team that assumes the mantle of the principal physician in the medical home model is going to be responsible for the oversight of the patient's total care, including Pap smears, mammograms and colonoscopies. While PCPs already do this, they don't necessarily accomplish those goals alone, as they require specialists for those other services. In an integrated model, monitoring chronic conditions is easier to accomplish, particularly if they share an EHR. There is fluidity in the specialist model in that a patient who is already seeing the specialist every two to three weeks for the management of ongoing problems becomes comfortable with that physician, and with that physician's care team. If that physician elects to assume the mantle of the primary physician, he or she takes responsibility of all other care issues as well.

There are still a variety of models to evaluate. Some IMS may elect to manage the entire patient care cycle, while others may elect to have a model in which advanced nurse practitioners are participants. If you have an integrated model, PCPs can pick up where the specialist left off. We are really discussing the flow of the patient back and forth. There is not an artificial wall created—one to the other. Over the next several decades, we are going to have to adapt to the fact that we don't have enough PCPs and we have a ready source of physicians who can assist in that process, namely the IMS. However, the medical home does not work well if the doctors serving as the principal physicians are not trained in a primary care or internal medicine environment.

Optimizing the medical home model for different groups will be different and fluid, and we may be open to those different approaches. Patient care must be a process in which we minimize unnecessary testing and eliminate unnecessary harm to the patient and maximize the patient's benefit, not just medically but also individually to each patient and their family—taking into consideration environment and societal contributors.

Lewin: We really need to make sure that care is patient-centered and not of physician design.

Lurye: I have encountered specialists who clearly do not wish to provide primary care services. For those physicians, there is going to be some tension. I am not sure how we solve that problem.

Walsh: I agree that not all subspecialists are going to be as interested in the medical home, particularly subspecialists who are procedurally based. However, for an oncology patient or a solid organ transplant patient, we have functioned for years in a medical home model without naming it because specialists deliver the flu vaccines and make certain that infections are treated. Clearly, it takes a village to treat the most complex patient, but the responsibility for these patients rests with the physician—the specialty physician. That said, not every cardiologist will feel comfortable with this specialist model and being responsible for every patient.

What shared-payment and shared-risk structure best facilitates the medical home?

Walsh: If it's directed by the patient, some specialists who are operationally functioning now as medical homes would be very interested in a shared model because we already are outcomes driven. We are assessed on not just quality measures that involve patient reporting, but also outcome measures that have nationwide benchmarks. As a community, advanced heart failure transplant specialists are familiar with this and would embrace it.

Lurye: In one sense, a medical home could be viewed as a core of an accountable care organization [ACO]. The question is how wide a circle is being drawn: Are you going to reimburse at a health system level that includes connected and disconnected PCPs and specialists? Or, are you looking solely at the outpatient world whether it's led by a specialty-driven model for a select group of complex patients or the more general primary care population?

Production-based payment models do not support the medical home concept well. Enhanced fee-for-service could allow a provider to hire case managers or a diabetic educator, but doesn't recognize that practices must reorganize to operate and survive once they are responsible for the health of the population. Currently, the only way to add revenue is to see more patients. In the future, there may be a tripartite form of compensation, which amounts to a capitation for care coordination activities, continued fee-for-service pay for productivity and rewards for quality achievements that some of the more specialized areas have been measuring for years. There are specialties that currently collect data in a uniform manner. One example is the STS [Society of Thoracic Surgeons] database in cardiothoracic surgery. This compensation model can be achieved, but it won't be one-size-fits-all.

Lewin: Five percent of patients consume 50 percent of the U.S. healthcare costs. These people require a high degree of coordination, maybe even more so than we propose in the medical home model. Then there is 10 to 15 percent of the population with multiple comorbidities who consume another large segment of the healthcare costs, who also need a medical home model. However, the additional 80 percent consume a relatively small amount of total healthcare costs—probably less than 20 percent. We need to first focus on those people who really require the care coordination, and there could be different options for those who are generally healthy.

For the 80 percent who consume modest amounts of healthcare dollars, we might have to be a little more flexible and not assign all those people to a medical home. Administratively, it's going to cost a lot of money, and it may not make much difference. Ideally, primary care practices should be revitalized and be well-funded to provide the right kind of teams to make them successful. Then, hopefully, economic and data integration will be able to determine the best patient care, while taking into consideration patient preference.

May: At a very fundamental level, we must enhance the value of cognitive care that U.S. physicians provide and reduce the outrageous value placed on some procedures. I am aghast that my full consultations are substantially less expensive than a single examination incorrectly ordered. Until we have a system that values the consultative and preventive care-driven role of physicians, we will continue to be driven by a procedure-oriented system. Appropriate use criteria help physicians, but do not replace trusted physicians who properly advise their patients.

We must be able to support and enhance the value of cognitive medical care, and it should be cost-effective. The U.S. is spending an inordinate amount of resources on a minority of the population. Getting our arms around that is the first step in returning to rational expenditure to the healthcare system. However, these efforts must go hand-in-hand with increasing the value of primary care services, which can be delivered by the PCPs, the family medicine physician, the interns, the nurse practitioners or a specialist care team.

How will hospitals, which will still assume the majority of the cost burden, fit into this model?

Walsh: Hospitals have their eye on this ball now more than ever and have moved to establish closer ties to their admitting physicians, whether those physicians are integrated or not because of the reimbursement structure. Also, with heart failure and other conditions, the hospitals have an intense focus on readmissions, where they expect to lose significantly unless they can reduce readmissions. They are exploring opportunities with clinic follow-up, home care, telemedicine and other strategies to monitor patients beyond hospital discharge. We are not talking about a potential payment structure change; we are already there with regard to the hospitals, as they are already invested.

Lurye: Structural aspects are important. We have a hospitalist team, and a structured way to communicate with primary care when patients are discharged. We also attempted, with some success, to position some of our PCPs at the most commonly used nursing homes to which our hospital discharges. Therefore, physicians seeing these patients can continue to access the outpatient chart from the EMR to ensure continuous availability of accurate patient data.

Lewin: If proper IT systems are in place, you can achieve true integration and value. Many hospitals, like Geisinger Health and Kaiser Permanente, converted case manager nurses to transition nurses. Nurses manage the transition from hospital to outpatient care, including going home with them to ensure their measures maintain stability, their home is safe and the handoff is organized. The nurse can make sure that whatever happens is transmitted to the next physician provider. In this model, everybody knows what's going on, such as what medications may better prevent unnecessary readmission.

How do providers respond to penalties versus incentives? How does the concept of ACO fit into these models?

Image source: N Engl J of Med April 21, 2010. (published online). *N denotes number of practices in each category. Percentages are weighed to be nationally representative.
Lewin: Based on the current Centers for Medicare & Medicaid Services [CMS] regulations, ACOs won't occur. However, looking at the next generation of providers, along with their demographics and attitudes, there is huge pressure to agglutinate to new integrated models. The pressure is huge with or without ACOs. The federal government may write the regulations over again, creating incentives for ACOs that could flourish. Even if we didn't have healthcare reform, and no one had ever heard of ACOs, this pressure to integrate is manifest now. This has to happen because we are wasting a lot of money, along with unnecessary morbidity and mortality, because we have failed to create these kinds of integrated models and coordinated systems of care.

Lurye: The CMS' pending approach to readmissions is all stick and no carrot. A looming threat is an unfortunate solution, and providers are left to figure out how to mitigate that threat. We also have to accomplish these goals in an environment where we are now being reimbursed the most we ever will see. How you reorganize and put together the requisite communication pathways supported by technologies in an environment like this isn't clear. The federal directive is: "We are not going to pay for these occurrences for certain conditions, but you have to figure out how to avoid that penalty." It is very troubling.

With readmissions, for example, it's not condition specific. If a heart failure patient falls, breaks an ankle and is readmitted, you will be penalized. This is true even if the patient is admitted to another hospital without your knowledge.

Walsh: I agree that healthcare is all stick right now. However, the stick is what has resulted in the hospital's interest in how the patient fares in the outpatient settings. I agree though there is not a carrot on the horizon.

May: We are in a very difficult transitional period. The current provision of care remains on the hospital side, fee for service, volume driven and unabated. Consequently, the current approaches to attempt to rectify our current model are by necessity stick. We are going to have a couple of decades, perhaps, of a fairly difficult time ahead. Right now, even those who are the vanguard for the concepts of coordinated care and patient-centered care are left to deal with the fact that our hospitals want volume. During this transition, hospitals are driving toward the concept of integration to protect their volume.

We will see a significant amount of turmoil as the policymakers come to grips with the different moving pieces. Currently, the ACO concept is a mental model that allows us to begin to look at the various pieces. The current regulations won't create ACOs, and they may morph into something else. But we, at this point, still have a very large section of a healthcare delivery system that continues to be obsessively driven by revenue streams generated by volume. Consequently, the way you stop this is to turn off the spigot, which by definition is the stick, and not the carrot.

Lewin: The carrots may appear later, and it will certainly happen fairly soon for those organizations that are integrated enough in partnership with payors to actually go at risk and undertake a global payment for managing either a fairly large bundle of services or for management of the whole population. Within these kinds of networks, there will be abundant opportunities to make care more effective and efficient, to reduce unnecessary admissions, reduce complications and reduce cost. In those circumstances, there will be incentives for physicians and hospitals that can achieve better efficiencies together, specifically when it improves the circumstances for patients as well. The outlook isn't gloomy for all providers, but it is for those providers that are clinging to fee-for-service, traditional funding and the current Medicare marketplace.

May: We need a paradigm shift in how we approach delivering healthcare. The stepwise efforts that we are recommending by necessity create friction at the surface where they interface with other models. This medical home model implies large-scale risk because providers would, for example, manage all patients in an entire North Texas area. At this point, this concept is very difficult to grasp, but it is the method by which we will achieve real cost savings and improved patient outcomes.

Lurye: When you talk about cost savings, you have to define the time periods that you are discussing. If you are living from annual budget to annual budget, you are not going to recognize the value of preventing breast cancer 20 years from now. Somehow, there has to be value recognized in that level of investment. In addition to those patients with chronic conditions, we also have to examine the underlying population that isn't sick yet—the 80 to 85 percent who do not consume many healthcare resources now, but may in the future have high cholesterol, diabetes, chronic obstructive pulmonary disease or heart disease. As a society, do we recognize the value of an investment that's going to pay off far in the future?