En Español | Plain Text Version | Skip Navigation

You are here:  Home > News and Events > Proceedings

Improving the Chain of Recovery for Acute Stroke in Your Community: Task Force Report

Skip secondary menu

December 12-13, 2002

Task Force Report
Templates for Organizing Stroke Triage

David Persse, M.D.
Task Force Chair
Emergency Medical Services
City of Houston 		J.P. Mohr, M.D.
Columbia-Presbyterian Medical Center
New York, New York
Richard C. Hinton, M.D.
Task Force Co-Chair
University of Texas Southwestern Medical School
Presbyterian Hospital of Dallas 		Paul E. Pepe, M.D., M.P.H
University of Texas Southwestern
Medical Center at Dallas and the Parkland Health and Hospital System
Joe E. Acker III, M.P.H., E.M.T.-P.
University of Alabama
Birmingham 		Michael R. Sayre, M.D.
Ohio State University
Columbus
Todd J. Crocco, M.D.
West Virginia University
Morgantown 		Jane Wigginton, M.D.
University of Texas
Southwestern Medical Center at Dallas
James V. Dunford, M.D., F.A.C.E.P.
University of California at San Diego Medical Center 		Paula Willoughby, D.O.
Emergency Medical Services
Chicago Fire Department
James C. Grotta, M.D.
University of Texas Health Science Center at Houston 		Daniel J. Worman, M.D.
Medical College of Wisconsin
Milwaukee
Anne D. Leonard, R.N., M.P.H.
University of Texas Health Science
Center at San Antonio 		Judith R. Yates
Healthcare Association of San Diego and Imperial Counties

Our health care systems need specialized physician expertise and in some cases above-average diagnostic capabilities with the current recommendations for management of acute stroke. Specialized care is particularly important in the use of thrombolytic drugs, which ideally should be used by physicians with special expertise in acute stroke management and at facilities with organized stroke programs. This sentiment is supported by two successive articles focusing on the use of intravenous t-PA in stroke, which appeared in the March 1, 2000, issue of the Journal of the American Medical Association. In the first article (1), which reported the results of the STARS* study, the participating medical centers had well-established stroke programs and were in fact chosen from among centers already enrolled in the ATLANTIS** study (2). The favorable outcomes in acute stroke patients achieved by the STARS investigators were comparable to the outcomes from the original NINDS study, but with a reduction in the incidence of intracerebral hemorrhage (1). The second JAMA article presented data gathered from evolving community programs at 29 mostly non-stroke center hospitals in the Cleveland, Ohio, metropolitan area (3). In contrast to the STARS study, the Cleveland survey revealed a significant increase in the mortality rate in acute stroke patients receiving t-PA, with an almost threefold increase in the intracerebral hemorrhage rate. It is important to note that subsequently, when quality improvement programs were applied to the same Cleveland hospitals, outcomes improved significantly. This reinforces our conviction that acute stroke patients need to be treated by specialized physicians at designated stroke centers to assure optimal care, as well as to provide a safety net for potential complications. Furthermore, proper emergency management of all stroke patients, even those not receiving t-PA, prevents complications and improves outcomes. Even after delivery of specialized care in the acute care setting, hospitals supporting a specialized inpatient stroke unit have documented an improvement in patient outcomes, as well as the added benefit of overall cost reductions (4). These considerations led the Brain Attack Coalition to propose the establishment of stroke centers that could effectively deliver this specialized care and possibly duplicate the success of the trauma centers (5).

This chapter is offered as a guide for setting up systems to facilitate delivery of specialized stroke care, including emergency medical services (EMS), and networks of stroke center hospitals. The information in this "how to" chapter has been supplied by individuals who have experienced first-hand the challenges and rewards of establishing these centers of excellence. The Appendices at the end of this report include additional resources that may be of use to those establishing a new acute stroke program.

Hospitals should set a the timetable of 1 year following the publication of this document to assess their capacity to meet evolving standards in order to qualify as a stroke center and justify the triage of patients for specialized care. Definitions for primary and comprehensive centers can be found in the Choosing Your Level of Care section of this book, along with suggested requirements. It is important to acknowledge that not all hospitals have the capability or desire to establish stroke programs, just as many hospitals do not offer a trauma program. The decision to organize as a stroke center should be strictly voluntary. It is hoped that the information from these Task Force Reports will induce more hospitals to join in this national effort and establish stroke programs. Hospitals not equipped to safely deliver t-PA should not be forced to offer this type of treatment to patients at their facility. It is expected that these non-participating hospitals would decide not to actively attract stroke patients, and would cooperate with community efforts to triage appropriate patients to stroke centers. Most larger community hospitals have the proper training and facilities to correctly use t-PA, and hospitals that evaluate substantial numbers of stroke patients should consider becoming stroke centers.

________________________________________________________________________________

*STARS = Standard Treatment with Alteplase to Reverse Stroke.
**ATLANTIS = Alteplase Thrombolysis for Acute Noninterventional Therapy in Ischemic Stroke.

Return to top

Getting Started

Usually one or more physicians step forward to champion and guide the effort to establish a stroke center hospital or network of hospitals. While not obligatory, such leaders are often neurologists. In some cities, emergency physicians have assumed the leadership role in this regard. A physician leader must be convinced of the importance of stroke centers as an enhancement to patient care. In addition, there must be a commitment to see the process through to completion, as there are many challenges and success will not come immediately. However, leaders should be fortified with the knowledge that the concept of stroke centers is a sound one, and that with perseverance success will usually follow. At a local level, the leader or stroke director will have the responsibility of transforming his or her hospital into a facility capable of quickly evaluating and treating stroke patients. At a network level, participating hospitals must be organized into an efficient system that provides access to specialized stroke care for all or most of the population in a community.

The project cannot succeed without enlisting the assistance and support of many other interested parties, and it is important for the leadership to identify those likely to be helpful and those likely to benefit from the formation of stroke centers. These might be called stakeholders. At a hospital level, planning and execution must be a joint effort between physicians and hospital administrators. However, on a community-wide basis, other stakeholders might include medical societies, hospital councils, stroke survivors, people at risk for stroke in the community at large, state and national organizations, and political bodies. At any level, the EMS system is an integral part of the process. The EMS system includes dispatch agencies, first responder agencies, transport agencies, and medical control physicians and hospitals. Training programs for EMS providers and dispatchers are essential. Also, much thought and effort must be applied to designing a seamless communication link between EMS provider groups in the field, hospital emergency departments, and acute stroke teams. In some cases a dispatch service provides this link and directs the transport according to pre-established algorithms.

Return to top

Local Hospital Stroke Center Development

Physician Buy-In

It is usually important for the stroke director to recruit other physicians to join in and shoulder the responsibility of care. Considerations here include arranging a stroke call schedule as well as forming a stroke team. Depending on the size of the hospital, the stroke team might consist of neurologists, emergency medicine physicians, radiologists, neuro-interventionalists, neuroradiologists, internists, and neurosurgeons. A non-physician stroke coordinator, usually a nurse or physician assistant, is also an extremely important addition to the stroke team. These individuals may help with clinical responsibilities as well as data collection and outcome monitoring. It is always advantageous to have a neurologist perform a neurological assessment early in the process and to be closely involved in making the complex decisions regarding treatment. However, this is not always possible, and other willing physicians must be trained to assess the patient neurologically, with phone consultation from a neurologist if possible. Emergency physicians are usually the first physicians to examine the patient and are sometimes the only physician initially involved. Their complicity and partnership in the stroke program is of course essential, and these physicians must be thoroughly familiar with the established emergency procedures and treatment guidelines, including transfer procedures, for stroke. Radiologists must be available to interpret imaging studies on an emergency basis. More advanced treatment is available at comprehensive centers, including interventional neuroradiology and up-to-date neuroimaging capabilities such as digital angiography. Patients with stroke frequently have significant underlying medical conditions, and emergency availability of internal medicine consultation is invaluable. Finally, neurosurgical expertise is called for in cases of subarachnoid hemorrhage, intracerebral hemorrhage, and thrombolytic-related hemorrhages. At other times, other specialty consults, such as cardiology, will be necessary.

Return to top

Tips for Obtaining Physician Buy-In

Physicians may be reluctant to participate in the stroke team and to extend themselves on an emergency basis. This is particularly true of neurologists, whose practices are often office-based. Points to consider when recruiting neurologists include the following:

  • Of all the specialties, neurologists are most capable of accurately assessing stroke patient and making the tough treatment decisions that are sometimes required. Their input is very important for the delivery of quality care. It may be important to remind reluctant neurologists of this, and this realization might persuade them to make their expertise more available in the interests of patient care.
  • If the availability of neurologists on a hospital staff is limited, bedside emergency consultations on a continuous basis may not be feasible. In this case, the emergency physicians must develop expertise in stroke evaluation.
  • Radiology expertise may not be available on a continuous basis at some hospitals, and of course this is essential. However, the essentials of CT reading sufficient for t-PA patient selection can be learned by non-radiologists, especially neurologists, neurosurgeons, and emergency physicians. Teleradiology arrangements have in many cases circumvented this obstacle, and remote interpretation of radiological tests is now very common.
  • If the barrier to physician participation is not a manpower issue, but is strictly the inconvenience of receiving an emergency call, hospital medical boards may apply pressure by requiring emergency call participation as a requirement for hospital staff privileges.
  • If a hospital is truly committed to attaining distinction as a stroke center, a hospital stipend for physicians agreeing to take emergency stroke calls can be an attractive incentive. For instance, it is not uncommon for hospitals to provide some financial incentive to surgeons involved in trauma calls.
  • Convincing physicians that timely and quality acute stroke care, even apart from t-PA, results in improved outcomes will encourage their participation.
  • "Commando" systems are difficult to maintain. Ensuring appropriate financial and logistical support, adequate staffing, and renewal by training of younger stroke clinicians should help encourage and support the health care professionals involved in such efforts.

Return to top

Hospital Administration Buy-In

An initial reticence from hospital administration is not unexpected, and it is important for stroke leaders to see the endeavor through the eyes of the administrators. Implementation of a stroke program at a hospital requires considerable effort on the part of the hospital, as well as some expenditures. Upgrades to diagnostic imaging equipment may, in some cases, be necessary. Staffing of stroke services in many hospitals is not 24/7, thus hospital administration may be faced with additional staffing costs. In addition, t-PA itself is an expensive drug. Ideally, a stroke unit or neuro-ICU would be available, and this requires considerable allocation of resources. Outcome monitoring is an important part of any stroke program, and continuing medical education must be provided for all involved. Creating a stroke program is no small task for a hospital. Nevertheless, with patience and tenacity, the hospital will often see the value to the institution and its patients and make the decision to become a partner in this.

Return to top

Tips for Obtaining Hospital Buy-In

  • Hospitals with large medical staffs and advanced diagnostic capabilities are usually able to offer specialized care in many fields. It is in keeping with their position in the community for them to include excellence in stroke care in the services they offer, particularly if they wish to consider their institution a center of excellence. This argument may carry some weight in discussions with hospital administrators.
  • For all hospitals, the formation of a first-rate stroke program may enhance their image in a community, increase patient volume, and improve patient outcomes.
  • The procedures developed for a hospital stroke program often result in cost savings by reducing length of stay and minimizing medical complications (4).
  • Increased staffing costs and functions may, in some cases, be shared with other programs, for example, x-ray technicians can be trained to operate CT scanners.
  • The members of the board of trustees may have an ambitious long-range vision for the hospital and may, by virtue of having a keen sense of responsibility to the community, be very helpful to the stroke program effort.

Return to top

Stroke Center Systems Serving a Community

Providing more complete access to expert stroke care for patients in a geographical area requires close cooperation between physicians and administrators of the participating hospitals. The following suggestions may be helpful to those leaders planning to establish a community-wide coalition of stroke centers.

  • A steering committee or stroke council should be formed by the founding physicians and hospitals to oversee and monitor the activities of the network. Representatives from all participating EMS provider agencies must be included on the committee as well. It is also prudent to include other stakeholders such as the local medical society, hospital council, etc.
  • In designing the triage transport system, the steering committee should make every attempt to offer services to all neighborhoods in a community.
  • A quality assurance program should be in place to measure performance by participating hospitals and EMS provider agencies.
  • Outcome data should be shared among the participating hospitals and EMS providers; this will require resolution of any confidentiality issues.
  • An important function of the steering committee would be the development of basic qualifications for hospitals wishing to participate. The guidelines established by the Brain Attack Coalition might serve as a starting point (5). An invitation should go out to all hospitals in the community, and any hospital should have the opportunity to join in the effort at any time, provided it is sincere in attempting to comply with the standards established by the steering committee. The decision to become a part of the network should be a joint decision between the committee and the hospital candidate. The committee does, of course, have a responsibility to insist that the participating hospitals meet certain basic requirements in the interests of patient care.
  • Non-participating hospitals should understand that transport of patients to a stroke center within the first few hours of symptom onset is unlikely to have a financial impact on them. This is because the percentage of stroke patients presenting within a short time of onset of symptoms is relatively small.
  • EMS providers are usually willing participants. They do need reassurance that patients transported to stroke centers will receive prompt and specialized care. It is important to involve them as much as possible in the planning stages, as they can offer helpful suggestions in the design of the triage system.
  • All EMS provider agencies should have an opportunity to participate at any time if they are willing and able to comply with the policies established by the steering committee.
  • EMS providers, through community education programs, can help dispel reluctance and embarrassment about calling 911.
  • Opposition to the plan may come from hospitals, physicians, or sometimes other groups. It is important to have open discussion with these dissenting voices, since the objections are often based on misunderstandings.
  • It is critical to obtain the endorsement of the local medical society or other physician representative organizations. These bodies usually recognize the value of stroke centers and are likely to offer enthusiastic support. They are particularly useful in neutralizing political obstacles among physicians and hospitals. If the medical society is firmly behind the effort and considers the idea an enhancement to patient care, it is more difficult for hospitals or individual physicians to erect barriers.
  • The American Heart Association has long been behind this national initiative and the local chapter can be a helpful catalyst in launching stroke center networks. The Association will soon introduce a new product, the Acute Stroke Treatment Program, designed to guide hospitals through a step-wise process for establishing primary stroke center operations. The Program is based on and complements the recommendations for the establishment of stroke centers published in 2000 (5). The organization also hosts educational activities and promotes community awareness of stroke symptoms. The National Stroke Association is also a useful resource.
  • At a national level, organizations such as the American Academy of Neurology and the American Medical Association can urge Congress to allocate funds for the founding of stroke centers.
  • Media coverage is helpful, but only after the network is working smoothly and efficiently.
  • Finally, organized stroke survivor groups may be helpful in lobbying efforts with various groups including hospitals and physicians.

Return to top

EMS Stroke Triage Policies

Stroke Center System Coordination/Oversight

The key to effective stroke center/system development and implementation is to identify or create a single entity responsible for organizing the stroke system. The entity should have the ability to cross geopolitical lines and coordinate all participants: 911 centers, EMS response agencies, medical control physicians, and hospitals. The entity may be an organization already in existence or may be created for this sole purpose. The entity must be viewed as neutral to all parties and receive its policy direction from the steering committee.

Return to top

911 Call Center

Emergency medical dispatch and call-taking has improved greatly over the past decade. Standards exist for emergency dispatching, there is a National Standard Curriculum in place, and there are several dispatch accrediting agencies. Integral to effective stroke care is early identification of potential stroke patients and subsequent prioritization of the dispatch. Most EMS call-taking protocols allow for the caller to report what they believe the patient's problem may be. While some callers may be astute enough to correctly report that the patient is having a stroke, many stroke patients will initially be reported as being the victim of a fall or having altered mental status, diabetic problems, or cardiac problems. It is unrealistic to expect call-takers to accurately identify all potential stroke patients. However, sophisticated call-processing protocols generally include key questions designed to identify potentially critically ill patients. These questions usually pertain to whether or not the patient is awake and breathing or speaking normally. The uniform use of these key questions for all reported emergencies allows the dispatch center to identify potentially seriously ill patients, including stroke patients, even if the caller misidentifies the patient's actual medical problem.

When a caller reports that the patient's problem may be a stroke, the prioritization scheme of the dispatch center must reflect the time-dependent nature of stroke care. Just like dispatches to patients with serious trauma or acute myocardial infarction, dispatches to potential acute stroke patients must be given high priority, above less critical and less time-dependent emergencies.

Return to top

EMS Assessment of the Patient

Several studies have demonstrated that emergency medical technicians and paramedics are able to identify acute stroke patients with relatively good reliability. Prior to any specific stroke-identification training, San Francisco Fire Department paramedics correctly identified 61 percent of acute victims. Following a 4-hour training program on stroke and instruction on how to administer a modified NIH Stroke Scale, they correctly identified 91 percent of acute stroke victims (6). Other large EMS systems, including those in Los Angeles, Cincinnati, West Central Florida, Birmingham, Houston, and Dallas, have designed prehospital acute stroke evaluation tools for use by their emergency responders (Appendix A).

Each of these stroke-screening processes includes a brief and simple physical exam. The Cincinnati Prehospital Stroke Scale (CPSS) consists of observation for a unilateral facial droop when the patient is asked to smile; arm drift from a position of the arms being held out in front; and slurring of words, use of incorrect words, or inability to speak when asked to repeat the phrase "you can't teach an old dog new tricks" (7). The Los Angeles Prehospital Stroke Screen (LAPSS) consists of a physical examination that evaluates smile, arm drift, and grip as well as five inclusion criteria (exclusion questions concern age, duration of symptoms, glucose level, and history of seizures) (8). The LAPSS proved to be very accurate, with a sensitivity of 91 percent, a specificity of 97 percent, a positive predictive value of 86 percent, and a negative predictive value of 98 percent (9).

In addition to the physical exam, each of these prehospital stroke scales inquires about specific details of the patient's medical history. One of the most important parts of the history for the EMS caregiver to ascertain is the exact time of symptom onset. Paramedics and emergency technicians should be trained to use all available information sources to determine as exactly as possible when the patient's symptoms started. Patients are often unable to provide the time of onset as they have become confused or are frightened about what is happening to them. Family members or bystanders may remember what events were occurring when the patient began to develop symptoms. This type of information can be used to help them recall the time of symptom onset. A useful tool is to ask the bystanders what was on the television or what meeting had just started when the symptoms began. Whenever possible, it can be helpful to bring family members and/or witnesses of the acute event with the patient. This allows physician and nursing staff to obtain further historical details and to address issues of consent and advanced directives.

If there is a history of seizure disorder, and in particular if the patient has had a seizure in the past 24 hours, the diagnosis may be Todd's paralysis rather than stroke. Another co-morbidity to be considered is hypo- or hyperglycemia in diabetic patients.

In addition to the rapid recognition of a possible stroke, EMS workers can benefit their patient by providing supplemental oxygen, intravenous access, and cardiac monitoring while expediting rapid transport to the nearest stroke center. The intravenous fluid of choice should not include dextrose as it is well recognized that elevated levels of glucose are potentially harmful to at-risk cerebral tissue. Hypertension should not be treated in the field because it heightens the risk of hypoperfusion to the penumbra. The patient should be transported in a laterally recumbent position on the affected side to protect the affected limb, provided this does not cause any respiratory compromise. Finally, giving early notification to the destination hospital can be extremely beneficial by allowing health care personnel more time to mobilize the necessary resources for patient care.

Return to top

Decision-Making for Destination Selection

Once a patient is recognized as being a probable acute stroke patient, the decision must be made to transport the patient expeditiously to the most appropriate pre-identified hospital. Essential to the success of such a program is that the prehospital personnel know, with little or no delay, which facilities are able to appropriately care for the patient. In Houston, a criterion for hospital participation in the stroke care system is resource availability 24 hours a day, 7 days a week. In Dallas, on the other hand, four stroke center hospitals are available but participate in a rotational system on a weekly basis. Regardless of the system used, two important features must exist: (1) the system should be easy to understand, and (2) EMS providers should routinely and reliably know which hospital is the appropriate facility to receive the patient.

In the setting of rural prehospital care, it may be especially difficult to maintain a rotational system due to the time and distance variables that this would involve. Additionally, web-based hospital status systems may be difficult to maintain due to funding issues and the diversity of institutions that could potentially be involved. Consequently, innovative approaches may need to be adopted. For example, expanding the role of EMS dispatchers may allow confirmation of the destination hospital. After determining that an acute stroke has probably occurred, EMS personnel in rural areas could recruit dispatch personnel to confirm a hospital's availability and suitability to receive the presumed stroke patient. Communication over long distances can be difficult from the field. For this reason, rural EMS systems should consider having their dispatch centers communicate patient care issues to the destination hospital. This has the added benefits of (1) providing the receiving hospital with more time to mobilize necessary resources, (2) allowing time for alternative arrangements to be made during times of hospital crisis or unavailability, and (3) determining bypass of unavailable hospitals rather than relying on time-consuming transfer processes. This might include air-medical transport.

As hospital and ED overcrowding continues to be an active issue for emergency care systems, contingency plans must be developed by EMS and hospital providers. For example, if a participating stroke care hospital has determined the need to request diversion of incoming emergency patients due to ICU overcrowding, does this affect the hospital's ability to care for a new stroke patient? Many stroke patients, even those receiving t-PA, do not require ICU care and can be managed in specialized intermediate care settings. Diversionary status for stroke patients should be determined locally and independently of other diversionary conditions. EMS providers must have a way of knowing if a stroke care facility can continue to accept potential acute stroke patients if the intended receiving facility has requested ICU diversion. The same pre-planning must occur for ED diversion requests, trauma diversion requests, etc. Also, how are EMS providers to react if all stroke care hospitals are at full capacity? Are they to take the patient to a non-stroke care hospital or are all the stroke care hospitals to be considered open and acute stroke patients divided among each of them in a rotation? The steering committee should also consider provisions relating to patients' personal preferences about hospital destination. In today's health care financial landscape, patients often find themselves unsure of the impact on health care coverage if they are not taken to an "in-service" hospital. EMS care providers need to be adequately educated on this issue in order to correctly advise patients.

One city that has implemented such a system and measured the impact is Houston. Prior to the implementation of an EMS stroke triage program, a local group of university-based neurologists coordinated an aggressive stroke treatment program in four EDs. Supported by an American Heart Association grant, each hospital in the city was asked to participate in the program. Six of 29 invited hospitals agreed to share quality improvement data and offer acute stroke care 24 hours a day, 7 days a week. Paramedics were then trained how to identify possible acute stroke patients using a stroke screening tool. Prior to the start of the program, 46 percent of all acute stroke patients were being transported to one of four original centers covered by the University of Texas stroke team, with 50 percent arriving less than 2 hours after symptom onset. After the paramedics were trained in the stroke screening assessment, 70 percent of all apparent acute stroke patients were transported to one of the six stroke center hospitals. The University of Texas team continued to cover the original four hospitals, and two new hospitals were added with their own in-house stroke teams. The most common reason for a possible acute stroke patient to be transported to a non-stroke center hospital was patient insistence on transport to the non-stroke center hospital. As a result of this effort, door-to-needle times decreased from a mean of 68 + 28 minutes to 54 + 11 minutes across all six centers, and the proportion of stroke patients receiving t-PA increased from 7.4 to 10.8 percent.

Return to top

Model Systems for Stroke Response

One of the goals of this document is to provide models for communities wishing to improve the care available to patients with acute stroke. A number of different locations across the United States have developed systems that address many of the issues identified as inhibiting effective emergency response to these patients.

Urban Settings

Change is occurring too slowly for many patients. Some communities may find it helpful to identify elements from the following descriptions of stroke response systems that may be good models for them.

Return to top

Houston

Area Leadership Team

In 1999, James Grotta, a neurologist at the University of Texas Medical School in the Texas Medical Center, received a grant from the American Heart Association to develop a regional stroke system. The grant was titled "Can Paramedic Education Improve Stroke Outcome?" The specific aim was to determine if a program of paramedic education, including the identification of designated stroke center EDs, would result in improved urgent stroke management. The program was a collaborative effort between the stroke teams at the two Houston medical schools-University of Texas and Baylor-and the Houston Fire Department EMS. The grant paid for a study nurse to recruit hospitals to participate and help develop the paramedic and ED educational programs, and it provided funds to develop a paramedic educational program focused on acute stroke recognition and triage. Limited funds for hospital staff education were also provided.

There were only two conditions for hospitals to join the system. The hospital had to offer advanced stroke care 24 hours a day every day, and it had to collect patient outcome information and allow the study nurse to verify quality improvement processes. All 29 hospitals in the area were invited to join the system. Six agreed to participate. The organizers of the system tried, with limited success, to recruit hospitals to achieve a geographic balance.

The local chapter of the American Stroke Association's Operation Stroke program endorsed the program. All area hospitals had representatives on the Operation Stroke task force, which included a variety of different professionals including neurologists, emergency nurses, and emergency physicians. Participating EDs received recognition from this group as designated stroke centers, and recognition was given to paramedics who brought to stroke centers two or more patients who got treated. An evening program with stroke survivors was orchestrated by the American Stroke Association chapter and was attended by stroke center ED nurses, paramedics, EMS directors, and the University of Texas and Baylor stroke teams.

Return to top

Receiving Hospital Designation System

Many hospitals that wanted to participate found they were unable to do so since they did not have the internal resources to collect data on numbers of patients treated, process control variables (door to CT time, etc), and patient outcome information. In some hospitals, emergency physicians wanted to participate but neurologists were not available for 24/7 coverage.

Of the six hospitals that participated, two were the home hospitals of the two medical school stroke teams (Hermann and Methodist). At three large community hospitals (St. Luke's, Memorial Southwest, and Memorial Northwest), a hybrid system had been developed during the NINDS study in which University of Texas stroke team neurologists responded to treat patients. In one remaining hospital, an internal stroke team, coordinated from the ED, responded to treat patients.

During the term of the American Heart Association grant (1999-2002), the study nurse verified the capabilities of the participating hospitals. After the grant ended, the Operation Stroke task force assumed that responsibility. There is a well-organized system for collecting process and outcome information from the six participating hospitals, including times to treatment, the proportion of stroke patients treated, and the accuracy of stroke assessment by EMS personnel.

Return to top

EMS Triage Policy

The Houston Fire Department transports acute stroke patients to whichever hospital the patient or family designates, but paramedics routinely encourage acute stroke patients to go to one of the six stroke treatment centers. The Houston Stroke Scale, a variation of the LAPSS, is used to identify potential fibrinolytic treatment candidates. Paramedics call the medical control system when they recognize a possible candidate, and the paramedic at medical control pages the stroke team at the destination hospital.

Early stroke recognition and the stroke scale was taught to paramedics during the first year of the grant, and the training is routinely reinforced at least twice yearly by the University of Texas stroke team at regularly scheduled meetings of all paramedics.

Return to top

Receiving Hospital Response System

The response system in each of the six hospitals varies. In two hospitals, there is an internal stroke team, and a neurologist responds to the ED to assess the patient. In those hospitals, the emergency physicians are taking on increasing responsibility for assessing patients and initiating fibrinolytic therapy without a bedside assessment by the neurologist. In three large community hospitals, the neurologist from the University of Texas at Houston travels to the ED to assess the patient. Within each hospital, there are specific systems to mobilize the CT scanner and other support services.

Return to top

Cincinnati

Area Leadership Team

In the mid-1980s an emergency physician, William Barsan, and a neurologist, Thomas Brott, who worked at the University of Cincinnati College of Medicine began to collaborate under the mentorship of a senior research neurologist, Charles Olinger, to develop an emergency treatment for stroke. They decided to develop a regional stroke response team. The key idea was that the clinical researcher who was going to be enrolling a patient into the t-PA stroke trial would travel to the patient instead of having the patient travel to the medical center. Choosing that approach defused potential political opposition. Although the strategy added complexity for the researcher, it simplified emergency treatment of acute stroke patients.

This single regional team, called the Greater Cincinnati Northern Kentucky Stroke Team, was founded in 1987. Team members met with area neurologists to allay any concerns they might have about the program. The community neurologists were pleased that the patients would largely be cared for at community hospitals by the stroke team physicians for the first 24 hours and then have their care handed over to the community-based neurologists. The team was available 24 hours a day to assess and potentially treat patients with acute stroke in EDs of any hospital in the metropolitan area.

The Cincinnati stroke team researchers also obtained approval from institutional review boards at each of the receiving hospitals for clinical trials including the NINDS dose escalation and randomized trials. Once the original NINDS t-PA stroke trial ended and the results were known, the stroke team continued its commitment to patient care and began using t-PA to treat eligible patients outside the research process while still responding to more than 17 community hospitals in the region. The core group of treating clinicians includes stroke-trained neurologists and emergency physicians who share call duties equally. A major advantage of this system is that a small group of clinicians gains extensive clinical experience treating patients with acute stroke.

The stroke team also holds weekly case review and quality improvement meetings in which patient cases from the previous week are discussed and new team members are trained in acute stroke treatment. Another benefit is that the emergency physician at the community hospital never has to decide which neurologist the patient's primary care physician wishes to use since there is only one group of clinicians treating patients in the entire region. A single telephone number is used to access the system from any hospital at any time, simplifying the process for activating the team.

The team provides EMS agencies with a mechanism to activate the response system in which EMS contacts the regional on-line medical control at the University Hospital and asks to have the stroke team paged for a patient. The treating paramedic provides as much detail as possible about the stroke, including the time of onset, when known, and the estimated time of arrival at the receiving hospital. Then the medical control physician pages the stroke team and also notifies the receiving ED.

Return to top

Receiving Hospital Designation System

Hospital participation is voluntary, and all acute care hospitals within the greater Cincinnati area, with the exception of the Veterans Affairs Medical Center, participate. A de facto two-tiered treatment system has developed. In the base tier, patients are treated with intravenous fibrinolytic therapy and then remain at the community hospital for subsequent care. If the patient is thought to need specialized care such as intra-arterial treatment or the community hospital staff is not comfortable maintaining the patient after administration of intravenous fibrinolysis, then he or she is transferred to one of the two hospitals with stroke units.

Patients who are brought to the EDs by their families rather than EMS comprise about 40 percent of the acute stroke patients in the area. They are cared for using the same system as patients who arrive by EMS. However, the system response for patients brought by family members is not as efficient as it is for patients brought by EMS, primarily because without EMS participation no real advance notification of the system takes place. On the other hand, the distributed delivery model developed by the Cincinnati stroke team means that all patients, regardless of their choice of hospital, will have access to intravenous fibrinolytic therapy.

There is an ongoing epidemiologic study of stroke patients in the Greater Cincinnati area that measures a number of patient outcomes. Recently, stroke team researchers shared data collected through the Ohio Paul Coverdell National Acute Stroke Registry with all participating institutions and the community at large.

Return to top

EMS Triage Policy

Local EMS agencies generally take patients to the hospital chosen by the patient or family members, unless that hospital is on ED diversion. Acute stroke patients are not routed preferentially to any particular hospital. The paramedics use mobile telephones to contact the local medical control physician at the University Hospital or to make direct contact with the receiving hospital. Depending on the clinical status of the patient, the stroke team may be notified about the case before the patient arrives at the ED so that the stroke team physician can begin to respond to that hospital.

Return to top

Receiving Hospital Response System

A neurologist or emergency physician who is a member of the stroke team responds to the ED upon the request of the treating emergency physician. While responding to the hospital, the stroke team physician uses his or her mobile phone to mobilize hospital resources. The mobile telephone is programmed with the telephone numbers for various hospital departments such as CT and pharmacy. The ED staff is also integral to the process and is responsible for making a room available in the ED and notifying the CT technician that an acute stroke patient is arriving.

After the patient is assessed in the ED and the decision whether to treat with fibrinolysis is made, the stroke team physician communicates with the emergency physician and admitting neurologist. The stroke team physician will usually stay involved in the care of patients who receive fibrinolytic therapy for the first 24 hours of hospitalization. After that initial period, patient care is the responsibility of the patient's primary care physician and consulting neurologist.

Return to top

Dallas

Area Leadership Team

In 1999, Richard Hinton and Hal Unwin, neurologists and members of a stroke committee of the local chapter of the American Heart Association, began an initiative to establish a regional stroke response system in Dallas. James Atkins, then medical director of the Dallas EMS system, assisted them in this effort. Presentations were made to various groups including the Dallas County Medical Society (DCMS) and the Dallas-Fort Worth Hospital Council. Eventually, all potential physician and hospital participants were invited to a meeting hosted by the DCMS.

After receiving approval from the DCMS and enthusiastic support from EMS, all hospitals in the region were then invited to actively participate in a stroke center network. The purpose was to provide access to expert stroke care for all patients in the Dallas County region. Four hospitals agreed to participate in this Dallas Area Stroke Network, and a rotational arrangement for the participating hospitals was established with the help of Paul Pepe, current medical director of the Dallas Metropolitan BioTel (EMS), and Ray Fowler, deputy medical director for operations of BioTel and its EMS base station. A Dallas Area Stroke Council was formed to oversee this network. The council is composed of physicians and administrators from the participating hospitals, with representation from BioTel, the DCMS, and the American Heart Association. The network became operational on August 1, 2002. An open invitation remains to any other hospital in the Dallas area wishing to participate.

Return to top

Receiving Hospital Designation System

There are four hospitals that are prepared to receive acute stroke patients from the EMS system. Each of these hospitals is a self-designated, comprehensive stroke service hospital. Two are in the northern part of Dallas and two are in the south. Hospitals alternate weeks being on-call, with one hospital in the north and one in the south on-call each week. Then the other two hospitals take call for a week, although all four hospitals can care for acute stroke patients at any time.

Membership in the system is voluntary and other hospitals are still encouraged to join the network. According to their own stated capabilities, each of the four hospitals has a comprehensive stroke service.

EMS Triage Policy

Paramedics quickly assess the patient with a short stroke scale, developed by Paul Pepe, to identify patients with possible stroke. If the last time the patient was known to be normal was within 3½ hours of the arrival of the medics, then the patient is considered for transport to a stroke service hospital. The 3½-hour window was chosen in order to allow the opportunity for intra-arterial thrombolysis (available at all the participating hospitals under experimental protocols) for patients presenting after the 3-hour window for intravenous t-PA. As soon as the medics identify a patient fulfilling these criteria, the patient is given the choice of being transported to a stroke service hospital or another hospital. If the patient wishes to go to a stroke service hospital, the paramedics contact the regional base station (BioTel). BioTel confirms with the EMS crews which hospital is on-call and prepared to receive the patient. In turn, the BioTel staff notifies the receiving hospital, which then notifies its own stroke team. Even if the stroke service hospital is on EMS diversion because of overcrowding, stroke patients are not diverted.

Return to top

Receiving Hospital Response System

Each of the four treating hospitals maintains its own acute stroke treatment team, available on a continuous basis regardless of their rotational status at any particular time. In most of the facilities, the stroke team membership includes an emergency physician, a neurologist, a neuroradiologist, and an internal medicine hospitalist. Neurosurgery consultation is also available on a continuous basis, if necessary.

Return to top

Smaller Cities

Ann Arbor

A different treatment model is functioning in Ann Arbor, Michigan. Stroke research physicians at the University of Michigan developed a unique system for supporting emergency physician administration of fibrinolytic therapy at a number of hospitals in southern Michigan without requiring the stroke expert physicians to travel to those hospitals. A system like this could cover a large geographic area.

Return to top

Area Leadership Team

Phillip Scott and William Barsan, both emergency physicians at the University of Michigan, pioneered this approach in March 1996. Initially, the participating hospitals were four teaching hospitals affiliated with the University of Michigan.

A treatment guideline was developed with features specific to each of the participating hospitals. The guideline had a number of components such as inclusion and exclusion criteria for t-PA administration in acute stroke, blood pressure control parameters, informed consent documents, dosing charts, and standard ICU order sets. Emergency physicians, nurses, and other staff at the hospitals were trained to use the guidelines. All of the treating emergency physicians were board certified in emergency medicine and the majority were members of the teaching faculty at the University of Michigan/St. Joseph Mercy Hospital emergency medicine residency. Regional stroke team members were available for emergency telephone consultation, but contact with the team was initiated at the discretion of the treating emergency physician. CT scans were interpreted in real time by radiologists at each of the treating hospitals.

Return to top

Receiving Hospital Response System

Membership in the program is voluntary. Since this is a distributed system, all patients are able to be cared for whether they come to the hospital by EMS or private vehicle; about 30 percent of the treated patients come by private car. During the initial 18 months of the program, about 60 percent of the treated patients received either an in-person or telephone neurology consultation before treatment was initiated.

This system has the advantage of developing under the auspices of a regional stroke research team, and therefore outcomes were measured. One-year mortality among the first 124 patients treated in this distributed system between March 1996 and April 2001 was 27 percent, which is equivalent to the 24 percent 1-year mortality in the NINDS trial cohort.

Return to top

Birmingham, AL

Area Leadership Team

The process of organizing stroke care began in 1997 when Camilio Gomez, a neurologist, and Joe Acker, executive director of Birmingham Regional Emergency Medical Services System (BREMSS), agreed to serve as co-chairs of a stroke task force sponsored by the local chapter of the American Heart Association. Neurology, emergency medicine, EMS, hospital administration, nursing, public health, politicians, and stroke survivors were all at the table. After the task force developed the plan, which covered the six-county Birmingham metropolitan area, it was approved by the local American Heart Association chapter, the Birmingham Regional Hospital Council, and other groups. The plan then became a part of the BREMSS Regional Medical Control Plan which was adopted by the Alabama Committee of Public Health. One key to achieving success was the inclusion of stroke survivors on the committee. Their presence served to inhibit economic self-interest behavior by some participants.

Return to top

Receiving Hospital Designation System

The system includes all hospitals that are willing to participate. Participating hospitals sign a contract with BREMSS. To be stroke-ready a hospital must have current availability of ED, x-ray, operating room, stroke ICU bed, neurologist, CT scan, and neurosurgeon (or transfer agreement). Each hospital notes its current availability within a computer network, which is updated every 3 minutes or less. This provides the hospital the availability to determine "stroke readiness" based upon available resources. Twelve of 19 hospitals have been verified by a multidisciplinary site review team to receive stroke patients from EMS.

Return to top

EMS Triage Policy

EMS routes patients only to stroke-ready hospitals unless the patient requests another hospital. EMTs use the Stroke Observation Scale (SOS) triage system to identify stroke patients. The EMT then communicates with the Trauma Communication Center (TCC) and relays information on the patient. The TCC informs the EMT of the currently available stroke hospitals. The EMT, in conjunction with the patient, chooses a destination hospital. The chosen hospital is notified by TCC and a copy of the stroke patient report is electronically sent to the receiving hospital. An education program and a train-the-trainer process educated more than 2,500 EMTs and personnel from all EDs in the region.

The BREMSS performs quality improvement and reviews system, hospital, and prehospital performance. Through their contract with BREMSS participating hospitals provide the required outcome and process data. A feedback loop to the EMT who placed the patient in the stroke system and the TCC communicator who handled the call is also performed. Each EMT and communicator learns the outcome for each patient entered in the stroke system.

Stroke patients who arrive at a hospital ER by non-EMS means are not entered into the stroke system. However, if a non-participating hospital initiates an interhospital transfer, the stroke system assists with this process.

Return to top

Receiving Hospital Response System

Each of the 12 hospitals that receive patients from the stroke system has its own internal stroke team that is responsible for the care of acute stroke patients within that institution.

Return to top

Rural Settings

Morgantown, WV

Area Leadership Team

The rural area surrounding Morgantown, West Virginia, has developed an effective stroke treatment system. The champion of the system was David Libell, who serves as director of the Comprehensive Stroke Unit at West Virginia University (WVU). WVU's primary teaching hospital is Ruby Memorial, which is a large tertiary care center and the only university hospital in the state.

Return to top

Receiving Hospital Designation System

Within the city of Morgantown, there is only one other hospital. Patients who arrive at that hospital with symptoms suggestive of acute stroke are routinely transferred by ground ambulance approximately 1 mile to the WVU Medical Center. Participation in the system is voluntary. Ruby Memorial Hospital spent about $250,000 on a marketing campaign during the year 2000 to inform the public and rural hospital personnel that acute stroke treatment was available. When a patient with acute stroke arrives at a rural hospital, the referring emergency physician contacts medical control for the air medical service to facilitate patient transfer. The referring emergency physician speaks with the emergency physician on duty at Ruby Memorial to verify that the patient is possibly a candidate for either intravenous or intra-arterial stroke therapy. About two-thirds of the stroke patients cared for at Ruby Memorial arrive by ground EMS, while 17 percent are delivered by air medical transport, 15 percent are brought by friends or family, and 2 percent have some other mode of arrival.

There is no organized system for verification of stroke treatment capability at hospitals in West Virginia. There is a stroke unit at Ruby Memorial Hospital, and there is a hospital-based stroke care committee that includes representatives from hospital administration as well as all disciplines caring for stroke patients during the entire hospitalization. The committee meets quarterly and reviews quality improvement activities. Reports from the hospital are made available to EMS.

Return to top

EMS Triage Policy

Paramedics generally use either the Cincinnati Prehospital Stroke Scale or the LAPSS to identify patients with acute stroke. These patients are then transported to Ruby Memorial Hospital. If the patient or family insists on transport elsewhere, they will be accommodated if the facility is within reasonable distance. The paramedics contact medical control at Ruby Memorial about a potentially treatable stroke patient. The medical control paramedic activates the stroke response system if there is an obvious stroke. The medic can consult the emergency physician on duty to get advice as needed.

Return to top

Receiving Hospital Response System

Calling a single pager number sets off a group of pagers and activates the stroke team. Pagers are carried by the emergency medicine attending physician, an ED charge nurse, a stroke neurology attending physician, a pharmacist, laboratory personnel, the CT technician, and a "stat" nurse.

In the mid-1990s, a telemedicine demonstration project was in place for about a dozen hospitals throughout West Virginia. John F. Brick, who is chairman of the department of neurology at WVU, championed this effort. Medical professionals at WVU can use the system to evaluate patients remotely using cameras and audio equipment. Histories and physicals can be performed using the system. While CT scans can be read remotely using the system, acute stroke patients have not been treated using this system alone. Since this technology infrastructure remains in place, its increased role in the remote management of acute stroke patients is targeted for further study.

Return to top

Summary

The task force envisions the development of local, and eventually state and national guidelines for stroke care delivery, including prehospital stroke care. In anticipation of these developments, this document has been created to help communities begin to pursue these goals.

The task force urges that, within 1 year of publication of this document, each community should:

  1. Evaluate its stroke care system capabilities regarding:
  • Public awareness of the signs and symptoms of stroke
  • Prioritization of potential stroke patients within EMS dispatch protocols
  • Training of EMS professionals in recognition and treatment of stroke
  • Uniformity of prehospital stroke care protocols among all EMS provider agencies
  • Uniformity of transportation algorithms and destination protocols for stroke patients
  • Identification of hospital resources regarding stroke care
  1. Identify or create a community organization to implement and oversee the stroke care system.
  2. Ensure competency for all components of the EMS system and participating hospitals in assessing and treating patients with acute stroke.
  3. Prioritize dispatch of acute stroke patients similar to that assigned to patients with major injury and acute myocardial infarction.
  4. Develop triage protocols for preferential stroke patient transport (including inter-hospital transfers) to designated stroke center hospitals.
  5. Collect, analyze, and share EMS and stroke center hospital data among participating EMS systems and hospitals for purposes of quality improvement and patient outcome.
  6. Develop local guidelines for stroke care delivery, including prehospital stroke care.

Return to top

REFERENCES

  1. Albers G, Bates V, Clark W, et al. Intravenous tissue-type plasminogen activator for treatment of acute stroke. Standard treatment with alteplase reverse stroke (STARS) study. JAMA 2000;283: pp. 1145-1150.
  2. Clark WM, Wissman S, Albers GW, et al. Recombinant tissue-type plasminogen activator (Alteplase) for ischemic stroke 3 to 5 hours after symptom onset. The ATLANTIS study: a randomized controlled trial. JAMA 1999;282: pp. 2019-2026.
  3. Katzan I, Furlan A, Lloyd L, et al. Use of tissue-type plasminogen activator for acute ischemic stroke: the Cleveland area experience. JAMA 2000;283: pp. 1151-1158.
  4. Stroke Unit Trialists' Collaboration. Collaborative systematic review of the randomized trials of organized in-patient (stroke unit) care after stroke. BMJ 1997;314: pp. 1151-1159.
  5. Alberts MJ, Hademenos G, Latchaw RE, et al. Recommendations for the establishment of primary stroke centers. JAMA 2000;283: pp. 3102-3109.
  6. Smith WS, Corry MD, Fazackerley J, et al. Improved paramedic sensitivity in identifying stroke victims in the prehospital setting. Prehosp Emerg Care 1999;3: pp. 207-210.
  7. Kothari RU, Pancioli A, Liu T, et al. Cincinnati prehospital stroke scale: reproducibility and validity. Ann Emerg Med 1999;33: pp. 373-378.
  8. Kidwell CS, Saver JL, Schubert GB, et al. Design and retrospective analysis of the Los Angeles prehospital stroke screen (LAPSS). Prehosp Emerg Care 1998;2(4): pp. 267-273.
  9. Kidwell CS, Starkman S, Eckstein M, et al. Identifying stroke in the field. Prospective validation of the Los Angeles prehospital stroke screen (LAPSS). Stroke 2000;31(1): pp. 71-76.

Return to top

Return to Table of Contents

National Institute of Neurological Disorders and Stroke
National Institutes of Health
Bethesda, MD 20892

Last updated February 09, 2005