Excellence in Stroke Care: Improving Care in Community Hospitals

Advances in the treatment and prevention of stroke have yielded significant results in recent years. In 2008, stroke became the fourth leading cause of death in the United States, down from the third leading cause of death, a position it held for decades. A number of factors are at play, including clinical advancements, increased awareness and increased adherence to evidenced-based care. Yet challenges remain in providing timely equitable and effective care.

To discuss best practices in stroke care within community hospitals, Health Forum convened a panel of industry experts Nov. 16 in Chicago for a roundtable discussion. Health Forum would like to thank all of the participants for their open and candid discussion, as well as Genentech for sponsoring this event.

MODERATOR (JOHN COMBES, M.D., AHA's Center for Healthcare Governance): Today's topic is best practices in stroke care for community hospitals. It's been about eight years since the Joint Commission launched its primary stroke care designation program. What changes have you seen over that period? How has stroke care evolved and improved through the use of the designation of primary stroke centers?

WENDE FEDDER, R.N. (Alexian Brothers Health System): First and foremost, is the ability to capture and access data to look at our clinical outcomes and compare them with our other hospitals, and with other primary stroke centers and community hospitals. We've really been able to see clinical outcomes improvement since we started with the guidelines program in 2005.

TIM SHEPHARD, R.N. (Bon Secours Health System): Looking at the clinical data has been helpful, but I think it's the recognition and prominence that the stroke center certifications and the work around neurosciences have promoted. Many community systems are seeing neurosciences as a viable alternative for them to develop as a service line that wasn't there before. It's moved neurosciences into the forefront, surpassing cardiology in some of our hospitals.

MODERATOR: Are you finding that your organizations are able to demonstrate through the stroke networks outcomes of care that are equivalent to even the best academic medical centers in the community?

SHEPHARD: That's true. If you rely on the data, community hospitals are achieving great things. Community hospitals, if they focus and maintain a focus, can achieve the same results as academic medical centers. It's a question of whether they go from stroke on to the next step in developing their neurosciences program — neurospine, neurosurgical spine, etc.

ANDREW MEADE (St. Luke's Episcopal Hospital): When we began our telemedicine network, we started with Beaumont Hospital. Within six months, they were faster than we at door-to-computerized tomography time. They don't have a lot of the same bureaucracy that we do and they don't see the same complexity of cases that we see. In many areas, they may not be able to handle the complex level of cases. But on the bread-and-butter stuff, I think they can be faster.

SHEPHARD: That's an excellent point.

MARILYN RYMER, M.D. (St. Luke's Hospital): I would take another view of the primary stroke center process. I certainly agree it's been positive overall. It's made people pay attention to data, to protocols and to standardization of care. Every hospital in our city is a primary stroke center. It's become a marketing tool.

I don't think just being a primary stroke center guarantees that patients will get the acute care they need. The continuum of care is well-organized within the primary stroke center guidelines. But we aren't doing as well when it comes to acute care.

MODERATOR: Would you say, though, that overall, the level of stroke care in this country in the past 10 years has improved tremendously, even though we haven't gotten to the reperfusion rates we'd like for them to be?

RYMER: I think "tremendously" might be too big a word.

SHEPHARD: But stroke has dropped to the fourth leading cause of death from the third leading cause of death in the United States.

DAVID GHILARDUCCI, M.D. (American Medical Response): I'll take it from a public health perspective. It's raised the bar for a community standard of care. Hospital administrators realize that we need to have neurologists on call. I think it's caused everybody to be more forward leaning. EMS is more forward leaning with stroke. Before, we'd say, "This is probably a stroke; we'll take the patient to the closest hospital."

Now we realize that we have only one chance to get it right. If the closest hospital is not a stroke center, we've lost valuable time. But I think the other issue, from an emergency physician perspective, is that there's more accountability on how we evaluate patients. There's just a lot more awareness now than there used to be.

FEDDER: The stroke center certification process in the past 10 years laid a nice foundation for the development of a system of care. Whether we are stroke ready, primary stroke, or comprehensive, I can't imagine that we'd be where we are today without it. We've definitely seen an improvement in outcomes.

CHERYL BUSHNELL, M.D. (Wake Forest Baptist Medical Center): As Marilyn touched on earlier, there is marketing exposure associated with certification. It's brought a healthy competition among providers that's sort of promoted stroke care. And in changing the goals, with door-to-needle times now being 60 minutes or less, it has upped the bar in terms of which hospitals can truly accomplish the kinds of process improvements necessary for getting the appropriate therapy as quickly as possible.

SHEPHARD: We're seeing a great deal of collaboration between larger and smaller hospitals to help standardize the level of care. Resources that may not have been available to smaller hospitals now are available through the systemization that stroke center networks have created.

On the other hand, more EMS networks are creating protocols that will bypass noncertified stroke centers. Some hospitals will lose some of their neurovascular volume as a result. You can project that this is better for the patient, but we also know that we need to maintain financially viable institutions in rural areas that may not be stroke centers. So one of the efforts that was taking place by the American Heart Association and the American Stroke Association was to develop a third tier of care, stroke-ready hospitals. But that changes the baseline standard of care, going one level below that to include other hospitals. There's a lot of debate about that taking place now.

BUSHNELL: Another argument for not bypassing some of our hospitals in our network is the availability of telestroke systems. It's a little bit of a different definition in terms of stroke-ready, because then they've got access to our stroke team. And they're highly trained in process. They stabilize patients and prepare them for transport for further treatment.

MODERATOR: I want to talk about that a little more, since the future looks like we're moving toward systems of care. Through telestroke services and through the support of the network, can you actually raise the level of care at smaller facilities?

RYMER: The systems of care process is a very viable option. We have a 25-bed critical access hospital that successfully treats stroke patients with appropriate medical therapy all the time with our help. I don't think all of these hospitals need to be stroke-certified, they just need their protocols in place, and then a relationship with a certified center to care for patients after they leave the smaller facility.

MEADE: That's the same tactic we take with our telemedicine network. We meet with organizations and ask them their goals. We ask about their level of neurology coverage. And then we work with them to establish programs that meet their goals and abilities.

MODERATOR: So, are you all working with hospitals that are outside of your primary network?

FEDDER: Our system has set up a collaboration with a nonsystem hospital in rural Illinois that has one of the highest stroke mortality rates in the state. We were able to get them to use some of the tools, like "Get With the Guidelines," and other things, to develop emergency department protocols so that they can treat the patient appropriately. If they have cases that are beyond their scope of practice, they transfer to us. You can see that their outcomes have improved in the emergency department. You don't have to be a certified primary stroke center to provide good stroke care.

SHEPHARD: One of the impetuses behind stroke certification is to standardize processes. Even when we're partnering with smaller hospitals outside of our system, we've taken the approach that any hospital can be a certified stroke center. We have a 98-bed hospital that is a certified stroke center using teleneurology because that's our standard of care within the system. I've seen some sites slide between certifications. You need that on-site validation to push the envelope. If you don't have a coordinator, if you don't have the data, if you don't have certification looming over your head, processes slide. And that happens in every hospital.

The most expensive components of a primary stroke center are data abstraction, collection and reporting, a 24/7 coordinator and ED training and processes. And those are the components that you have to have in place to deliver standardized care around acute stroke care. So if you can do those, you might as well go for certification. If you have to do the most expensive components, a hospital, by being a third tier, is not going to save any money, and they're going to lower the bar. The whole goal of stroke certification is to raise the bar. So I think taking that step down, while it provides that on-site validation, still lowers the bar in the extended care savings to a third-tier hospital.

FEDDER: I agree with that, but it's going to be difficult for some hospitals to become primary stroke centers. Critical access hospitals, for example, may not have the stroke volume or the technology to become certified. But they could partner with a system and meet some of those standards. However, it would have to be through development of some type of collaboration.

RYMER: The most important thing is that every hospital be stroke-ready, whatever that means for that hospital. So if there's no CT scanner and the family physician is the guy who's manning the ED and he's home, then that hospital needs a transfer plan. The ones that can do drip and ship need those protocols well-systematized and inline with the hub hospital. And then you work your way up from there.

But patients generally are going to go to the closest hospital. Even with these transport protocols that we're trying to get together in our states, they often land at the closest hospital. My firm belief is that every hospital needs to be stroke-ready.

MODERATOR: Tim raised the issue of the cost of these programs and the investment that hospitals have to make. Have you seen a documented reduction in cost of care?

SHEPHARD: Yes, absolutely. We have implemented what we call clinical transformation in the Bon Secours system where we track our highest volume margin diagnosis-related groups. Stroke is one of those. We have the volume metrics around cost per case, complications, etc. All of these things have come down as soon as we put the stroke center in place.

MODERATOR: That's good, but I guess we'll have to change our language to value. You're getting the costs down, but are you getting great outcomes, which is really adding value to the whole system?

SHEPHARD: For reperfusion, the availability of a therapeutic option for severe stroke patients who come into the hospital has increased, and the complication rate for these patients has decreased because the physicians have more time to make better decisions. As a result, complication rates are down. It's just giving physicians time to make more informed decisions about the course of treatment.

RYMER: If you become a stroke center, you actually become a neuroscience center by default. Stroke is sort of the leader in a way, and then you get brain tumors, seizures and all of those things that may even lead to neurosurgical procedures, which most hospitals find to be very profitable.

FEDDER: Well, in sheer volume, neuroscience is about 30 to 40 percent of our inpatient volume. That gets the attention of our senior leadership. If you don't treat stroke, your ED traffic is going to be nil. That's what fills your hospital.

GHILARDUCCI: There's a phenomenon that I call soft triage. There are protocols that direct EMS staff where to take which patient, but patients aren't always easy to categorize. So EMS makes a judgment call, which can be subjective. They may take the patient to the hospital they perceive to be the most responsive. They're intrinsically motivated to know that their patients are getting good care. So, generally, if they are concerned about a patient, they will say, "Let's just go a little farther to this other hospital." Most EMS systems aren't able to manage that. That sort of soft triage happens all the time, though. From a business perspective, it's important to consider.

MODERATOR: But don't you think some of that soft triage, even from the patient's perspective, comes from looking at people who seem to have mastered care? The care is coordinated and it's efficient. It's sort of a side benefit of what you do.

GHILARDUCCI: If an organization holds certification in some area of specialty, EMS knows that they're competent in that area. There's an assumption that you're competent in all areas, and that's not necessarily true. A comprehensive stroke center may not necessarily be located at a trauma center, but often those two things go together.

RYMER: EMS is part of the care team. To that end, we give the EMS crew a report on how well the patient did within 72 hours. It's wildly popular.

MEADE: And an appreciative letter from the physician — that always helps.

MODERATOR: This coordinated care approach sort of reminds me of Regina Herzlinger's focus factories in care. Is this now spilling over to other lines of your business within the organization?

MEADE: It's going to. Our chief quality officer has looked at our stroke program and said this is something we need to mirror elsewhere. They see we are tracking data and doing performance improvement like nobody else is doing in the hospital. And as medical home models become popular and payment becomes linked to quality, you can't ignore the success that stroke is having.

The guys running our CT have taken notice. They see what we're doing for stroke patients and how well it works. They've questioned why all of this isn't being done for everything else. The time to CT is not as good as it would be if we had a system of care. It sort of stands out as a sore thumb.

GHILARDUCCI: I'd like to touch on one topic that was brought up earlier about data. I think one of the biggest frustrations — and this would be my message to hospitals — is not to let HIPAA stand in the way of sharing information because it's for quality purposes. HIPAA certainly allows it, but it has become such a roadblock in terms of finding out patient outcomes for the pre-hospital folks and they often get discouraged. If there's one thing that will drive EMS away, it's not knowing what's happened to their patients.

MEADE: Yes, but you'd be surprised. At a previous job, I shared information with EMS and then the legal team found out about it. We spent the next three months sorting through what we'd share with them. It was really insulting to our EMS partners.

GHILARDUCCI: It becomes an obstacle.

MEADE: It became a three-month obstacle, and we really had to work to not insult our EMS partners. The numbers are amazing. We found that 99 percent of the time EMS is correct in their diagnosis. That's great information to share with them.

BUSHNELL: We focus on the acute stroke setting. Once the patient is admitted and past the hyperacute phase, we don't share information. But EMS should know what happens in the emergency department and whether or not the patient improved.

MODERATOR: You can share your aggregate outcomes, and you can talk about the patients who were brought there. Not by individual name, but in aggregate, and include the EMS team in any of your improvement efforts as well. It's important, too, that in this whole concept of coordinated care, there's a strong team approach throughout the continuum, so the HIPAA issue should not get in the way because you're all taking part in the care of the patient, and sharing outcomes is certainly allowable.

GHILARDUCCI: The Santa Clara County Public Health Department brings all of the stroke centers together once every two months to share data. And with the exception of Kaiser, none of them are part of a system. Only the organization knows which set is its data. The idea is that it provides a forum for people to look at each other's outcomes. It provides an opportunity for organizations to share best practices. One of the things that has come out of these meetings is that we've streamlined our intrahospital transfer process. We're basically leveraging our 911 system to move patients quickly from one place to the next, and we agreed on criteria as to which patients will qualify. I would recommend to hospitals that they collaborate. I know that your organizations compete, but there's also an opportunity to collaborate on a communitywide basis that would serve everybody's interest.

MODERATOR: Well, it sounds as though this was, for all of you, a great business decision — that it was great for patients and improved outcomes and care, and that almost everybody can play at some level. So what are the obstacles to getting this to happen? What are the roadblocks?

RYMER: It has a lot to do with the expertise and availability. We've had stroke center development teams visit us for the last 10 years, and you can almost tell who's going to make it and who isn't by whom they bring, and they're often missing the neurologist. There's been a lot of resistance in the neurology community because it's a change in lifestyle, basically. They used to be outpatient doctors and now, suddenly, they're in the ER more than the cardiologists are. We've got a wonderful young group of trainees who are embracing this and saying this is something they want to do. But the older practitioners are not necessarily embracing it as much. So there's that kind of resource scarcity. And we did have some pushback from ED for a while, but that's changing. The tide has changed because the data are there.

SHEPHERD: There is a constant shortage of highly skilled neuroscience expertise. Whether it's skilled neurovascular neurologists, skilled neuroscience nurses, neuro-interventional techs, neuro-interventional physicians, these resources are difficult to find. If a hospital has those resources and can share them with other hospitals for training purposes, for standardization, for data analysis, it's a real benefit. Finding a skilled neuroscience coordinator or stroke coordinator for a program can be a long search. We've actually had to grow some of ours, which requires intensive training, but we just couldn't find the right match for our system. So the scarcity of neuroscience expertise is another reason why you need to share those resources in hospitals that don't have them.

MEADE: I'm inserting a neurohospitalist into one of our community hospitals right now from our academic group because it just can't get a neurologist. The hospital is willing to pay full time for a neurohospitalist, who will probably see two or three patients a day. They can't get patients out without one. We've seen the same thing in psychiatry; we just can't get somebody to come in and consult on our patients.

BUSHNELL: I understand the reason for the shift toward using neurohospitalists, but there's going to be a negative consequence, and that is in the continuity of care. And we have a neurohospitalist. He doesn't see clinic patients, though. Luckily, we have a nurse practitioner who kind of bridges that continuum between inpatient and outpatient.

MEADE: From the financial point of view, I think we will be setting up a hybrid model where the neurohospitalist will spend the morning rounding on patients and the afternoons seeing patients in the clinic. From a financial point of view, the pro forma is going to look better from a physician perspective under this model.

ALISSA GORELICK, D.O. (Alexian Brothers Medical Center): From the ER perspective, the worst time in the ER is 3 a.m., when there's nobody in the hospital but you. We do have an ICU physician 24 hours a day. But in the ER, it's usually just me. And I run with the codes, and 25 beds are full, and there are 10 patients in the waiting room. And it's never black and white. The idea of a neurohospitalist is pretty amazing to me from an acute care standpoint because there's someone there to do the checks and balances.

Our neurologists are very responsive. They call back right away and help make decisions. But another pair of eyes is the best-case scenario.

MEADE: We set up our telemedicine program that way, where the patient comes in at 5 p.m., and we'll see the patient through until morning, when the neurohospitalist or the neurologist gets there to take over care. So that's where you can start developing that hybrid of care. But if you put one neurohospitalist in a hospital, and you don't provide any support, you'll burn him or her out. How many 3 a.m. calls can one person take? That's where you can mix in telemedicine. We'll do that in our system hospitals; telemedicine will provide coverage when the neurohospitalist is not there.

GORELICK: So are they there 24 hours a day, or are they just taking calls from home?

MEADE: It could be set up that way.

RYMER: We have what we call our code neuroteam. These are nurses who are neurocritical care trained nurses and are National Institutes of Health stroke scale-certified and have a lot of experience with acute stroke intervention. They provide coverage 24/7. They're our first responders to help the ED not just make the decision, because the neurologists are always available for that, but to help the ED nurses take care of the patient. ED nurses can use an extra set of hands to do the NIH stroke scale and to really be partnered with the ED staff. That has been a real satisfier for the ED and also for the neurologist who, if they can't get there within three hours, knows there are at least two sets of eyes looking at this patient.

SHEPHARD: We've implemented teleneurology within our EDs, even where we have employed physicians. So, it's easier to recruit a neurologist, because you don't have to do ED call. Their quality of life is better, and their patients' outcomes are better. It's been a boon to be able to recruit highly qualified neurologists into our system because they don't have to do ED call.

MODERATOR: I want to get back to Cheryl's issue: the discontinuity of care that some of this may lead to, especially as we move into an era where we're going to see bundled payment and organizations are going to be penalized for readmissions. How do you build in at least the immediate outpatient care after a patient comes through your primary stroke center?

BUSHNELL: Our issue was access to the follow-up appointments; it took literally six months to get a patient into the resident clinic. In hiring the nurse practitioner, I developed a template for all the things that patients should go through for prevention and follow-up and now our access is more like four to six weeks for the follow-up. That's had a tremendous impact on our outcomes. Our readmission rates have diminished somewhat because of the input of the nurse practitioners and we've also developed a transition coaching program that basically is calling patients after they go home to go over medications and appointments.

FEDDER: We don't have residents and we don't have neurohospitalists, so we do have gaps between levels of care, even with the team that we have. We've tried to be a bit creative with the transition from inpatient hospital to outpatient physician appointments. We started a program called the Stroke Nurse Navigation program. Navigation is not new to health care, but it is relatively new in stroke. We have two nurse navigators who see patients while they're in the hospital. They meet them and then follow them by phone at certain intervals, up to a year after discharge.

What we found is that you can hand patients all this wonderful information and literature and try to educate them while they are in the hospital. But patients and their families retain very little of that information, and they may not make their appointments with their primary care neurologist. So, we've employed this particular group of nurses in the Nurse Navigation program to help make sure patients are getting those appointments, their therapy appointments, going to rehab, etc.

MODERATOR: It seems the other advantage is that you can collect a lot of longitudinal data on these patients. As far as that's concerned, what have you seen in the longitudinal data for these patients who pass through the primary stroke center?

SHEPHARD: We've implemented "Get With the Guidelines" for data collection. It has 13 additional blank fields that you can use, so we've been capturing data at 30-day clinic visits, and then up to one year. We just implemented this in the last few months, and it's also being integrated into our electronic medical record as we expand that across the system. Once that's done, we should be able to capture that, especially as our employed neurologist group grows larger.

RYMER: We've been tracking our 90-day modified rankings for about three years on patients who receive acute therapy, either intravenous thrombolytics or some sort of intra-arterial therapy. It's a real challenge, though. It's a manpower issue to have the number of people you need just to do these phone calls. This is a huge issue.

MEADE: So, Marilyn, who's doing that for you? Who's doing the follow-up calls, discharge phone calls?

RYMER: We divide it up. The nurses attached to our interventional team do the intra-arterial follow-up calls, and neuronurses do the others. But we're patching it together. It's not as nicely organized as we would like.

MODERATOR: How do you assess your stroke programs? What is the ultimate measure of success from a clinical perspective?

SHEPHARD: We look at recurrent strokes and secondary prevention. Is the patient taking antiplatelets and antihypertensives? We have a Nurse Navigation program for our cancer patients, but we haven't implemented one for our stroke patients. But that's key to keeping patients out of the hospital. And with bundled payments, we're not going to be reimbursed for that. So it's good for the patient, it's good for the hospital. We measure success based on our recurrence. We compare the NIH Stroke Scale results for incoming and outgoing patients. But we don't have sufficient data on how the patient is faring long-term, and that would be very helpful.

BUSHNELL: The Adherence eValuation After Ischemic Stroke Longitudinal (AVAIL) registry, though, showed the recurrent stroke rate was about 8 percent. And these patients were actually pretty good at taking their medications and following their course of treatment. About 75 percent of patients were still in the same discharge regimen at three months, and 65 percent were on the same regimen at a year. That's pretty good, and you still have a recurrent stroke rate of 8 percent.

MEADE: It is difficult to reach out to every patient after discharge, with the resources available. But it is possible to identify which patients may be a problem post-discharge. At my previous place of employment, we reduced readmissions from 6 to 3 percent by focusing on those patients. I developed a business plan; we're devoting about $150,000 a year to reducing readmissions. If we focus on the 50 patients who come back repeatedly, we could cut $4 million in costs.

BUSHNELL: You are right, you can identify those patients. The more severe the stroke, the more cognitive impairment, the lesser the resources and social support; those are high-risk patients.

SHEPHARD: A business plan too often does not accompany the decision to develop a stroke program. Decisions are based on the desires of the neurologists and other factors. But organizations often don't quantify the patient benefit, the