Emerging Models of Regionalization
Bob Bass, executive director of Maryland’s Emergency Medical Services (EMS) and trauma system and past president of the National Association of State EMS Officials, as well as the National Association of EMS Physicians, chaired the second panel. He said the first panel highlighted the methods of regionalization that have evolved in trauma care over the past three or four decades and provided a number of important lessons.
He noted that an integrated and coordinated approach—not to be confused with a centralized approach—can help in ensuring that the right patient gets to the right hospital at the right time and receives the right care. Accountability can be promoted through systems of verification, he said, and through data that examines processes and outcomes to ensure that the components of the system are working as they should.
The second panel focused on a number of time-critical conditions that may lend themselves to the same sort of regional approach that has been taken by trauma systems. These conditions include ST-elevation myocardial infarction, out-of-hospital cardiac arrest, acute stroke, and care of seriously ill and injured children.
ACUTE STEMI CARE
The first panelist was Joseph Ornato, cardiologist and emergency physician from the Virginia Commonwealth University in Richmond. Ornato said that the story of STEMI (ST-elevation myocardial infarction) dates back about as far as the trauma centers story—into the late 1960s, early 1970s. But he said it was not until about 1980 that we started to figure out what
was causing the majority of heart attacks. Before then, mortality if you had an acute STEMI was about 40 percent. Today, mortality in most cities is under 10 percent and usually it is in the 4 to 6 percent range. Clearly, he said, “we have made incredible progress.”
Over the past 30 years, we have learned that there are two major ways to open heart vessels: chemically or through mechanical means (percutaneous coronary intervention, or PCI). PCI is the current state of the art in treatment, but only one quarter of U.S. hospitals now have the capacity to provide it.
What has become obvious over the past 30 years, Ornato said, is that time is critical for patients. Just as “the golden hour” became a mantra in trauma care, “time is muscle” became the mantra for acute myocardial infarction (MI) care in the mid-to-late 1980s. Since then, we have really tried to better understand the trauma center model and apply its lessons to acute MI. A number of models have emerged, including Boston EMS, Minneapolis Heart Institute, and others. Now, the American Heart Association—analogous to the American College of Surgeons in this case—has helped identify the key components of a successful STEMI system and has launched a nationwide program called “Mission Lifeline.”
While a successful STEMI system has many similarities with trauma centers and regional systems of trauma care, Ornato continued, “There are also some very harsh differences.” Most general medical service hospitals rely upon cardiovascular care to stay alive financially. Therefore, a very important piece of the puzzle has been to ensure that we carve out an important role for medical centers that are not PCI centers—the functional equivalent of a Level I trauma center. “Non-PCI centers must be included as part of an integrated network,” he said, “and we have sought to strike a delicate balance, such that patients are not being diverted to PCI centers when it is not medically necessary.”
CONSIDERATIONS IN REGIONALIZING CARDIAC ARREST
Lance Becker, professor of emergency medicine and director of the Center for Resuscitation Science at the University of Pennsylvania, discussed cardiac arrest and related topics such as hypothermia and post-resuscitation care. Becker noted that there is wide variability in survival rates for cardiac arrest in communities in this country, ranging from 2 to 18 percent; however, we do not understand the causes of that variability. One goal of regionalization should be to aid in reducing variability.
Becker noted that cardiac arrest differs from most conditions, because about half of the cases in the United States take place inside hospitals. Moreover, it is a very time-sensitive illness. In some of the treatment modalities used for cardiac arrest, survival rates have been shown to differ based on
as little as 10-second intervals of time. “We’re not talking about a golden hour here,” he said. “We’re talking about a golden couple of minutes.” The regionalized system must be sensitive to that.
But Becker said that we need to get a handle on just what we mean by the term regionalization. One version, which he calls the Mecca model, involves bringing patients to a facility that has tremendous medical capacity. Another is to distribute some of that capacity out to the places where the patients are, and where they need them. A third model—virtual regionalization—involves distributing the expertise required to care for cardiac arrest cases so that local providers can provide the care in local facilities. In many cases the expertise needs to be brought to the patient, he noted, because in cardiac arrest we do not have the luxury of moving the patient to the Mecca.
Becker observed that while the procedures used to care for cardiac arrest patients are simple to perform (e.g., “thumping on a chest [hands-only CPR] is not difficult”), orchestrating an entire episode of cardiac arrest treatment is extremely difficult. “If you want to embarrass yourself sometime, just go to a mock code and run that code and you will be embarrassed.” He said that a lot of people “would be shocked” by the real quality of the care that is taking place in many cases.
An unintended consequence of a regionalized system of care, according to Becker, is that some centers become very very good, but others become less good, “i.e., really bad.” That should be taken into consideration. He argued that what is needed are several good demonstration projects to help figure out where regionalization’s pearls and pitfalls will turn out to be.
THE EMERGENCE OF STROKE AS AN EMERGENCY
Arthur Pancioli, professor and vice-chair of emergency medicine at the University of Cincinnati, said stroke is the leading cause of adult disability. There are almost 800,000 strokes per year in America, and statistically every member of the audience has a family member with stroke. “It is an enormous disease process,” Pancioli said.
However, until about 15 years ago, he said, stroke didn’t even rate as an emergency in most places, because nothing could be done about it. In fact, back then stroke was a viewed as a Level V triage by many ambulance systems—the same as a toothache. Now, primarily because of the emergence of an effective acute treatment, stroke has come to be viewed as an emergency condition.
Pancioli described stroke as a diverse disease process which truly requires a multidisciplinary approach. Even in the case of the simplest stroke, care that is well-coordinated at a local community hospital can make an enormous difference. The majority of stroke patients are cared for at the
equivalent of a Level III, IV, or V trauma center, and actually this is where they belong, he said. The personnel there have access to effective guidelines and can obtain support through communication.
In the case of patients whose conditions are a little more complex and who are a little bit sicker, Pancioli said, many do not need specialized procedures. The local physician just needs to be on the phone with an expert physician for guidance; commonly this is facilitated by having the expert review the imaging. Then there is a fairly small percentage of patients (possibly in the single digits) who need more extensive treatment and who should be seen at the most comprehensive stroke centers.
Pancioli observed that stroke is a very young disease process in terms of therapeutics, probably about 15 years behind acute STEMI. Stroke providers have been vastly less coordinated than providers of cardiac care, primarily because there has been a lot less money driving the process. Interventions are more cognitive-based than procedural. Pancioli concluded that “there has never been a disease that is more amenable to careful regionalization than stroke.”
A HUB-AND-SPOKE WHEEL MODEL FOR CHILDREN
Joseph Wright, senior vice president at the Children’s National Medical Center in Washington, DC, and member of the 2006 Institute of Medicine (IOM) committee on the Future of Emergency Care, pointed out that kids are not a disease process but they do account for about 20 percent of all visits to emergency departments in this country.
According to Wright, children are comparatively healthy overall and centers that can treat severely ill or injured children are relatively scarce. There are now about 90 freestanding children’s hospitals in this country. Children typically flow through the emergency care system by way of a hub-and-spoke wheel model, Wright said. When a child is in need of tertiary care, typically they are seen at the edges of the spoke-wheel and then flow to the hub.
So with regard to the care of children, regionalization is a process that is already functioning daily, Wright said. Significant research has shown that children do better in regionalized systems of trauma care, especially systems in which there are pediatric trauma centers exclusively committed to caring for children.
Because the vast majority of children are seen at the periphery of the hub-and-spoke wheel, we need to ensure that all of the 4,000 emergency departments in this country—some of which see fewer than 10 children a day—are prepared for pediatric emergencies. There needs to be a standardized floor of readiness. Emergency departments should also be able to handle major surges in demand, such as will occur with the H1N1 virus.
Finally, Wright discussed the federal Emergency Medical Services for Children (EMSC) program and the performance measures it has established. One of the performance measures is aimed at identifying a system of categorization, so that particularly the prehospital community knows which hospital emergency departments are prepared and equipped to handle certain types of pediatric cases and which are not. Only four states—California, Illinois, Oklahoma, and Tennessee—currently have a categorization system like this in place. States that are receiving funding from the EMSC program are directed to develop a categorization system, and this, Wright said, should be a topic of discussion for us.
Abhi Mehrotra, emergency physician at the University of North Carolina Hospitals and chair of the American College of Emergency Physicians’ (ACEP’s) Task Force on Categorization of Emergency Departments, said there seems to be a great deal of effort from specific disease islands—cardiac, trauma, stroke, and so forth—as opposed to the overall system of emergency care. He asked, how can we organize the models and categorize the components so that we do not duplicate efforts within each disease silo?
Bass replied that this is “a really, really good question.” He said that some of the later sessions hopefully will drill down into that a little bit more. But he added, “I think our view in Maryland is that we are a system of emergency care.” He said Maryland started with a trauma system, but evolved into a system of emergency care. The system is now able to incorporate “whatever comes along that is time-critical.”
Bass said the same infrastructure applies across the board—communications, data, verification, regulations, everything. If any additional specialty-care area is needed, it is actually pretty easy for the system to add it, although it can be difficult politically. For example, it might just include adding a set of verification standards and another column for STEMI centers in their Web-based application that tracks the hospitals and their capacity.
Ornato added that when new treatments and new hospital product lines arise in this country, you may have a dozen facilities raising their hand—having never done the therapeutic intervention before—asserting their right to receive their share of the patients. There’s no easy solution, he said, because we all understand the market forces. “We don’t [have] an environment in which someone up above is going to look at a community and pick [who should provide the interventions]. That would be easy, but that’s not reality.” Consequently, he said, we have “a fundamental system problem.”
Hospital Economic Incentives
David Seaberg of the University of Tennessee and ACEP focused on the economic incentives facing hospitals. He said the case for regionalization of traumatic pediatric burn care may be easier to make, because the profit margins on those cases are not very great. However, for cardiac, stroke, and other cases where the margins are more substantial, community hospitals will fight against regionalization—unless, of course, they might be recognized as the hospital to go to for that specific condition. Seaberg asserted that this reflects a truly inefficient system.
Ornato said a study in the journal Circulation illustrated the economic impact for a small community hospital, a medium-sized hospital, and an academic health center if community PCI centers were named, but that hospital was not included as one of them. The dollar losses ranged from about $150,000 per year for the small community hospital (which could be quite significant for a facility that size), up to at least $1 million for the larger facility. He said “this opened all of our eyes to the fact that, at least for STEMI, economics are really a very important consideration.” However, he said, the first priority has to be what is best for patients.
Wright said that early experience from California showed that efforts to categorize emergency departments relative to the care they could provide for children proved difficult. Hospitals responded, “If you take away pediatrics from us, what does that mean for our OB service? If it is perceived that we do not or cannot take care of children, what does that mean for our other services?” One lesson from that experience was to get early buy-in from hospital associations and really demonstrate with data that some places do better than others.
Assessing Burgeoning Capacity
David Sklar from the University of New Mexico and ACEP said that all four panelists indicated the need for additional research to demonstrate that, for their time-sensitive conditions, care can be improved if it is organized in certain ways. Sklar asked how we can encourage that type of research. Pancioli added in response, “What industry can you imagine would start a new product line without an incredibly careful and rigorous study of the outcome of that sale? It would never happen.” But he said he could imagine in medicine that we would design a very good regionalization system, but then not study the results. Why? Because there is an added cost in doing so that flows separately from the dollars that pay for care. Still, he said, not studying the results would be a serious mistake.
Becker said that in policy there is often an impulse to “just do something,” rather than do research. He said that New York City is doing a large experiment right now with respect to hypothermia. They have
decided they will begin to bypass patients with out-of-hospital cardiac arrest to centers that are cooling centers. Now, 17 hospitals across New York City have suddenly decided to become cooling centers (previously, there were only 4).
Becker said this represents a large natural experiment that is now under way. However, it will be very difficult to sort out the results of this experiment and determine whether or not there has been an effect. He argued that we need good studies that can help inform what it is about a system, or about the concept of regionalization, that confers this ability to reduce variability and improve survival.
Pancioli pointed out that already at this meeting we have repeatedly heard the call for “metrics, metrics, metrics.” He said, whatever we do with regionalization, if we don’t build in counting mechanisms to look at the outcomes of our efforts, we will make changes and never know if they’re good or not. He said, so now you’ve got 17 cooling centers. That could be great, or, due to the dilution of quality and expertise, “that may have been the worst thing in the world.”
Pancioli added that the reason New York ended up with 17 cooling centers is that no one wanted to lose the correlatives. “The reason people can’t stand the concept of strokes bypassing [their facility’s] stroke center,” he said “is because they’re going to lose all the weak-and-dizzies and all the syncopal patients and all the Medicare patients who pay money.” He continued, “If there is an economic incentive to do something, suddenly centers will pop up, and without metrics to measure their performance, there will not be a way to cull them back and get them focused back on what they should be doing.”
He added, “We rely heavily on EMS and whenever we discuss regionalization, we should thank our lucky stars for the prehospital providers who go out there and make really hard decisions in cornfields and intersections with profoundly undifferentiated patients.” He emphasized that “we need to educate them, give them good tools, give them feedback based on individual cases, accept overtriage and undertriage, and educate them toward the right level.”
Bass said that we now have several decades of experience with the trauma triage algorithm for prehospital providers and the Centers for Disease Control and Prevention (CDC), and others have put as much science into it as we possibly can. But, as a system, we are still struggling with it for patients who have conditions such as acute stroke and STEMI. If you are in southern Maryland and you’re 45 minutes from a PCI center but you’re 5 minutes from an ED, what is the appropriate decision about where to go? What should the cutoff be? Bass said that those of us who have regionalized systems of emergency care at the state level are looking for more guidance on the stroke patients and out-of-hospital cardiac arrests. Obviously, he said, that is a huge emerging issue.
Maryland has an inclusive system in that every ED in the state has a role in the trauma system. But the state has wondered how best to centralize the care of patients who have serious injuries. The state has 9 adult trauma centers and 47 hospitals overall. Bass said, “When we put out a request for applications for stroke centers a few years ago, it just came with a flood. All of the hospitals were very interested in being stroke centers, which meant having the pathways, doing the training, putting in the expense, submitting to a verification process, following the standards, submitting data, etcetera—for all of them.” He said at last count 35 of Maryland’s 47 hospitals have been designated as stroke centers. He reported that the state has been looking at the data and has found “terrific” results with regard to reduction in mortality. “It seems to me a very different model than the trauma model,” he concluded.
Sklar asked what might be learned from the experience of other countries. He said because the United States continues to struggle with whether health care should be considered a public good or a business, we face many political challenges that other countries do not. He said there may be lessons to learn from how they have addressed regionalization.
Ornato said that clearly these are global issues and the industrialized countries face very similar problems. He acknowledged that “some of our colleagues [in other countries] really are far ahead of us.” For example, in Scandinavia, researchers examined a regionalized PCI model for patients with STEMI and demonstrated better outcomes at those centers that are geared up to provide the care 24/7, as trauma centers are, and have a critical mass of volume. The study found that the PCI centers that had ramped up quickly and gained a critical mass of experience over a short period of time proved to be far superior in their performance.
Michael Sayre, chair of the Emergency Cardiovascular Care Committee for American Heart Association, said he would like to give Pancioli a chance to elaborate on the idea of decentralizing care and regionalizing expertise. Sayre argued that the stroke community has done a much better job than any of the other entities focused on here in spreading their expertise, both physically, by going to the referring hospitals themselves, or virtually through telemedicine.
Pancioli reiterated that stroke is a little more cognitive and a little less procedural than other diseases, though that is beginning to change. For the most part, though, stroke victims only need to be able to get to a hospital that is equipped with a phone or a telemetry device of some sort that aids communication. This is often the only technology that is needed to obtain assistance in making treatment decisions, such as whether thrombolysis or a more technical procedure is required.
Pancioli said that earlier in his career he had signed up for a stroke team, requiring him to be on call every fourth night serving an entire hospital system composed of 15 hospitals. Team members went to those hospitals—across Cincinnati and into adjoining areas in Kentucky—and still do. “We chose to do it on foot,” he said. He and his team still go to these hospitals and treat about 20 patients per month with tPA.
But, he said, there are another one to two dozen hospitals that are farther away, out of reach, that call and ask for advice on patients. What do those centers, emergency physicians, or other practitioners need, he asked? They need assistance in reviewing the patient’s clinical history and physical examination, and in reading the computed tomography (CT) scan. He said this is often done in the middle of the night without any problems by someone in Australia or New Zealand.
Pancioli noted that he is probably better in person than he is by phone, and is arguably better by videoconferencing than he is by phone (but not quite as good as in person). Still, he said, “We can do an awful lot of this just by bringing the expertise out there with technology. I don’t have to lay hands on every single patient I see. [However,] I think there’s a marginal difference. So when you can get into a center, you should.” In general, however, he called stroke a wonderful opportunity to take expertise out to distant places.
The Role of Emergency Medical Services Personnel
David Stuhlmiller, an emergency physician at Westchester Medical Center in New York, said that most hospitals advertise that they can take care of their communities, and most EMS agencies want to bring community members to their own community hospital. It is familiar to them, but it is not necessarily the best option.
Regionalization involves delivering the right patient to the right hospital at the right time. But it has taken many, many years to convince EMS to deliver trauma patients past their local hospital to the trauma centers. He asked, is it also going to take 20 more years to convince them to drive stroke patients to stroke centers, cardiac patients having STEMIs to STEMI centers, or children to children’s hospitals? How do we involve EMS in the decision to bypass the community hospital that wants these patients?
Bass said that in Maryland, it has actually been a task to hold EMS back. He said Maryland strives to make these decisions based on evidence, and sometimes it takes time for the data to settle in. But, he said, “Paramedics in Maryland are chomping at the bit to get patients to interventional centers.” Candidly, he said, sometimes they jump the gun. But the Maryland EMS providers know who does primary PCI, they know where to take the patients in their community, and they do it. Maryland is trying
to follow up with a formalized designation process and standards to ensure there is an appropriate interface, he said, but EMS personnel are leading us, not trailing us, in that respect.
Pancioli said the key is to empower EMS and assure them that it is okay to make these decisions, because there are a lot of barriers. If they are a hospital-based transport system, their hospital will say, “don’t you dare bypass us.” But if you empower them and inform them that they may save a life by doing this, they will drive right by that hospital.
The truth is, Pancioli added, that there is a lot of negative pressure on them, much of it economic. Clinicians need to give them the decision tools they need (e.g., the Cincinnati stroke scale or 12-lead electrocardiogram [ECG] to detect AMI) and empower them to make these decisions. Bass reiterated that the challenge in Maryland has not been the EMS providers. It has been those with financial interests and political clout who sometimes get in the way of doing the right thing for patients.
Stuhlmiller said that there are 47 transporting EMS agencies just in the County of Westchester, where he is based. Many of these are volunteer and they are community-minded. They will not drive 55 minutes outside of their service area for one individual if that means leaving their community with reduced coverage. With respect to the advanced life support (ALS) providers, Stuhlmiller said, the state can come down and say, “This is a stroke patient—take him to the stroke center.” Hopefully, they will say, “This is a STEMI patient—take him to a STEMI center.” But for the volunteers, this is not the case. Bass replied that “that is why you need a system-wide approach and protocols,” with everyone part of the same mechanism for designation, verification, and data collection. “It needs to be system-wide,” he emphasized.
The Uses of Data
John Holcomb, former commander of the U.S. Army Institute of Surgical Research and trauma consultant for the Army Surgeon General, asked the members of the panel whether they had all published findings on the quality outcomes in their communities based upon implementation of the systems they had described. He said he found at the Department of Defense that capturing that kind of information and being able to show that soldiers fare better if a system approach is used, allowed money in support of the system to flow more easily. He wondered whether the same dynamic might occur in communities, putting the pressure back on the funders and the businesses to support this type of system.
Wright responded, “You’re absolutely right. We have to be able to demonstrate these kinds of quality-of-life improvements in order to justify creation of these systems.” He said there has been some research through
the EMSC program showing highly positive results for children who were treated at centers that were able to perform ALS procedures. However, he acknowledged, “we’ve got a volume issue”—this study is just one example of what needs to be done throughout the country.
Ornato said a number of studies now very clearly show that organized systems of care for STEMI do measurably improve survival outcomes, and the systems are cost-beneficial. He provided an illustration of the impact that numbers can have. He said that his medical center began cooling patients and providing post-resuscitation care in 2003 and, until 2008, they were the only one of 12 hospitals in Richmond that cooled. But each year they received only their fair share of 5-10 cases. However, the data showed their survival rates jumping from 2 percent to 10 percent and continuing to climb, while rates elsewhere in the city were essentially remaining the same.
In 2008, as the EMS medical director in Richmond, Ornato said he decided, after consultation with many colleagues, to have the paramedics bring patients to his facility exclusively. He said that survival rates citywide have now reached 18 percent, which he attributed to the fact his hospital is now doing 70 cases a year instead of 5-10.
But, he said, about two months ago every community hospital in town declared at a regional EMS council meeting that they would be introducing a hypothermia protocol within a month or two, and they demanded to receive all the patients in their catchment area. Their major concern was that EMS would preferentially bring STEMI patients with no cardiac arrest to the post-resuscitation center, because it would be perceived as providing better care (however, this has not happened based on objective data review). Ornato said that at this point in the negotiations, it looks as if Richmond may wind up with one or two other centers—the maximum that existing volume can justify.
David Magid of Kaiser Permanente said that, with respect to STEMI care, the preliminary literature on volume and outcomes indicates that there is a threshold for minimum volume, but the threshold is not very high. He said both small and large hospitals have shown tremendous improvements in door-to-balloon times; this improvement was not only observed in large-volume centers.
The Importance of Public Buy-In
Joseph Waeckerle, editor emeritus of Annals of Emergency Medicine, commented that the speakers had forgotten an essential component in designing centers of excellence, and that is the consumer and the public. Unless you have public buy-in, he said, you will have difficulty in your community, your region, and your state.
Waeckerle observed that the public understands the center of excellence
concept. They don’t understand the science behind it, but they understand that they or their loved ones may have a better chance to live under this type of system, because these are perishable skills and you want to be able to go to a place that performs these tasks all the time. He cautioned, “When you design your systems, please don’t forget the consumer, the public. If we don’t educate them, we are going to fail.” Bass agreed that the perception of the citizens is a key piece. He said, “It’s a piece that we have to use to try to offset the political heat that we get when we try to do these things.”