Task Force Report
Provider Support Systems for Acute Stroke
|Phillip A. Scott, M.D.
Task Force Chair
University of Michigan
|Dennis M.D. Landis, M.D.
University Hospital of Cleveland and Case Western Reserve University
|Sidney Starkman, M.D.
Task Force Co-Chair
UCLA Stroke Cente
|Steven R. Levine, M.D.
Mount Sinai School of Medicine
New York, New York
|John Y. Choi, M.D.
University of Texas Health Science Center
|David B. Matchar, M.D.
|Brian F. Connoly, M.D.
State University of New York at Buffalo
|Brett C. Meyer, M.D.
University of California at San Diego
|Karen L. Furie, M.D.
Massachusetts General Hospital
|Debra G. Perina, M.D.
University of Virginia Health System
|J. Stephen Huff, M.D.
University of Virginia Health System
|Jeffrey L. Saver, M.D.
UCLA Stroke Center
|Walter N. Kernan, M.D.
Yale University School of Medicine
New Haven, Connecticut
|Lee H. Schwamm, M.D.
Harvard Medical School and Massachusetts General Hospital
|Marian LaMonte, M.D., M.S.N
University of Maryland School of Medicine
The phrase "provider support systems" describes the organization of human and material resources necessary to solve a clinical problem. Developing support systems is an integral step in the second phase of translational research. In the first stage, the principal focus is on bringing new therapeutic approaches from the laboratory to clinical trials. In the second phase, the problem is to move from the clinical trial setting to accepted clinical practice.
Why Develop Support Systems ?
Provider support systems have proven remarkably effective in improving medical care in numerous areas in the United States. Regional trauma systems emphasizing the "golden-hour" have led to improvement in the outcome of patients suffering injury (1,2). Systems effective in reducing door-to-needle times in patients with myocardial infarction receiving thrombolytic therapy have also led to improved outcomes (3). Similarly, the progressive testing of chemotherapeutic agents in a systematic, sequential fashion has substantially improved quality of life and survival in patients with cancer (4).
Clinical trials have demonstrated that interventions exist to reduce the morbidity and mortality from stroke. These include primary treatment (thrombolysis in acute ischemic stroke ) and prevention of secondary complications of stroke (deep vein thrombosis, aspiration pneumonia, decubitus ulcer formation, etc). The use of other interventions, such as specialized stroke care units, is supported by evidence of improved long-term survival and independence for stroke patients treated in such settings compared to patients treated in a standard hospital setting (6). The widespread application of these disease management approaches is limited by various barriers, some of which can be overcome through effective support systems.
While it is not reasonable to expect every physician and health care institution to manage every aspect of stroke care, it is reasonable to expect pre-established plans to deal with stroke patients falling outside the expertise of an individual provider or health system. These systems should serve to enhance the providers' and facilities' own capabilities while providing for patients needing additional resources, thus improving the delivery of care for all stroke patients.
Every health care delivery system providing care for patients with acute stroke has a responsibility to develop and implement plans for meeting the requirements of each phase of stroke care. The burden of stroke, for individuals and society, compels us to develop effective support systems to deliver the optimum care to every patient with stroke. This document identifies multiple support systems to accomplish this task.
Specific Challenges for Support Systems
The use of thrombolytic therapy for acute ischemic stroke poses an especially difficult challenge in the development of support systems. Ischemic stroke is a clinical diagnosis based primarily on patient history and physical findings. Unlike cases of acute myocardial infarction or trauma, there are no immediately available confirmatory studies for stroke. Guidelines derived from successful clinical trials call for treatment when the diagnosis is established by a person with expertise in stroke; however, the vast majority of patients present to hospitals without specialized stroke teams, and are instead seen by frontline medical providers, such as emergency medical services (EMS) personnel, emergency physicians, family practitioners, and internists, who have limited expertise in diagnosing stroke and/or using thrombolytics in patients with stroke.
Even when physicians who are comfortable in diagnosing stroke are available, there may be other substantial barriers to treatment. Patients often fail to seek medical support in the initial hours of ischemic stroke, therefore delaying initial presentation. In-hospital delays may exist, including impediments at triage, during initial physician evaluation, and in obtaining required CT and laboratory studies. Given the potential obstacles encountered, it is not surprising that more than 7 years after the initial NINDS t-PA publication overall utilization rates are estimated in the 1 to 2 percent range for all patients with acute ischemic stroke (7). Even at centers with specialized stroke capability, only 7 to 9 percent of all acute stroke patients undergo thrombolytic therapy (8).
The barriers encountered in the development of systems to deliver hyperacute therapies to stroke patients highlight many of the obstacles to effective care for patients with either ischemic or hemorrhagic stroke in complex health-delivery systems. By developing support systems to overcome these barriers, it is anticipated that care of stroke patients in general will improve. It is encouraging that recent evidence suggests that development of support systems for providers serves to increase the use of thrombolytic therapy in stroke (9).
Areas for Support System Development
From a chronological perspective, there are multiple points during the care of acute stroke where pre-established systems should exist to enhance patient care. In some locations, all the resources for a given phase may be present. More often, however, institutions and caregivers require additional resources to provide optimum neurologic care. These resources may be as simple as advice over a telephone or as complex as automated image interpretation. Regardless of the precise details, it is usually possible to predict the resources required, and to plan in advance the manner in which the necessary resources will be obtained.
Presently, optimal stroke care requires rapid evaluation and management in the first few hours after the onset of clinical symptoms. The phases of stroke care for which advanced planning is required include:
Moreover, each health-care delivery system must monitor the success of its plans and be prepared to change, as necessary, to improve. As staff change over time, an institution should develop protocols and plans for training new personnel in the resources and techniques that have proven feasible at that site.
In general, knowledge used by medical personnel in the acute setting falls into at least three categories: training acquired in advance, information and training available upon demand, and information obtained in the acute care setting. Training and continuing education are discussed in more detail elsewhere (see chapter on Professional Education, page XX); the following discussions will emphasize the range of support systems and resources that may be used in the acute care setting. A list of suggested sites for additional information appears in the Appendix
Prehospital Care Support
Prehospital recognition of the potential stroke patient is essential for rapid initiation of treatment. Evidence from prehospital studies indicates that stroke patients accessing 911 emergency systems have significantly shorter times to physician evaluation and CT imaging of the brain than patients arriving at the hospital via other means (10). To exploit these benefits the following support systems are recommended for prehospital care.
EMS Training and Stroke Identification Support
Education on the signs and symptoms of stroke should be included in the initial and continuing education of all prehospital personnel. Since EMS providers function according to written guidelines developed by their medical directors, it is advantageous for such guidelines to contain a prehospital stroke identification instrument, e.g., Cincinnati Prehospital Stroke Scale, Los Angeles Prehospital Stroke Screen, or other similar scale (11,12). Prehospital stroke identification instruments can be rapidly administered and they enhance the ability of EMS providers to accurately identify stroke and notify the receiving facility in advance. All EMS providers should be familiar with such an instrument and employ it when assessing suspected stroke patients to allow early hospital notification.
EMS Transfer Support Systems
Stroke patients should be transported without delay to the closest appropriate facility. Guidelines for hospital destination decisions should be made in advance with the cooperation of area hospitals, emergency departments, EMS agency administrators, and EMS medical directors. Such decisions should result in the patient reaching a stroke-capable facility in the most timely manner. In cases where long ground transport time to the receiving hospital is anticipated, consideration should be given to aeromedical evacuation to preserve the potential for thrombolytic, endovascular, neurosurgical, and/or investigational therapy in appropriate cases. Data suggest that aeromedical transfer of stroke patients may increase the opportunity for acute intervention and can be accomplished safely (13-15).
Prehospital Management Support
While prehospital providers are knowledgeable about the symptoms of stroke, many are unaware of the appropriate use of supplemental oxygen, the potential detrimental effects of unnecessary dextrose administration, the need to avoid rapid lowering of blood pressure in stroke patients, and the presence of a therapeutic window for thrombolytic delivery (16). The development of EMS system guidelines requiring intravenous access and glucose measurement, avoidance of interventions to lower blood pressure, and methods to assist in identifying the symptom onset time (e.g., transporting a family member or event witness with the patient) and stressing the importance of early treatment is recommended.
EMS personnel routinely provide reports of individual patient assessment to local medical directors. Guidelines should exist to ensure that physicians at the receiving facility are made aware of the prehospital assessment of possible stroke, thus allowing internal hospital systems to prepare for rapid evaluation and treatment. Currently, such reports are given verbally; however, new technologies utilizing real-time video/still image linkages between ambulances and receiving centers may allow receiving physicians to remotely examine the patient during transport. This type of communication may shorten the decision-making process for acute therapies.
Hospital Care Support
Following arrival at the hospital, via EMS or other means, the stroke patient faces barriers within the hospital setting. These include many that are not amenable to hospital support system development: emergency department (ED) and intensive care unit overcrowding, reductions in hospital staff, and nursing shortages. A lack of access to consultants - particularly those with radiology and neurology expertise - is modifiable through specific support system development and is discussed in detail below.
Rapid radiologic assessment is necessary to determine patient eligibility for thrombolysis, neurosurgical evacuation of intracranial hematomas, and early treatment of aneurysmal subarachnoid hemorrhage. Hospitals that provide acute stroke care but do not have available expert CT interpretation on-site must develop a plan for urgent and accurate interpretation of brain CT scans. This may be accomplished via use of local resources, remote teleradiology systems, or patient transfer.
Teleradiology Support Systems
Teleradiology is a widely accepted practice for a variety of medical conditions. Traditionally, high-cost teleradiology solutions for CT interpretation have imported uncompressed brain images into a Picture Archiving and Communications System (PACS) workstation to be viewed on a 2K black and white high-resolution monitor. Newer low-cost teleradiology systems are now available for the transmission of wavelet compressed DICOM images (17). These images are compressed without loss of visual resolution and can be viewed using an internet browser on a conventional PC equipped with a cathode ray tube or liquid crystal display monitor (18-21). These systems allow local radiologists or neurologists to provide rapid interpretation of high-quality images from their homes or offices, or for outside physicians to provide interpretation services by contract. Because these systems handle much smaller datasets (a head CT scan is approximately 8 MGB, uncompressed) they can be effectively transmitted at low to medium bandwidth rates that are widely available. The impact of integrating these technologies may be substantial at non-urban hospitals with a high likelihood of first medical contact.
Rapid assessment, including neurologic examination, is necessary for optimum acute stroke care; it is especially urgent in determining eligibility for thrombolysis. Patients in areas without local expertise may have limited access to these treatments. Hospitals providing acute stroke care must have a plan for urgent and accurate patient evaluation. This may be accomplished via use of local resources, remote teleconsultation systems, or patient transfer. The delivery of rapid, efficient, and accurate neurological evaluation has been improved by the creation of "acute stroke teams" in various hospital settings. Centers that utilize such teams may have better outcomes compared with those administering t-PA ad hoc or to relatively few patients (22,23).
Teleconsultation can allow a stroke expert to remotely evaluate a patient's neurological condition, review brain images, and initiate a collaborative management plan with a referring physician. This location-independent infrastructure collapses barriers of time and distance to bring medical expertise to the bedside, potentially enabling delivery of care that would not otherwise be available.
Telemedicine for stroke, sometimes referred to as Telestroke, offers the potential for:
Telemedicine has been shown to be a feasible and effective method of performing consultative services and delivering remote neurologic assessment and treatment of stroke patients (24,25). Specifically, reasonably good inter-rater reliability of NIH Stroke Scale scores is preserved when performed over a telemedicine video link (26). These preliminary data support the potential role of a remote telemedicine link in the assessment and treatment of a patient's neurologic deficit.
Depending on the degree of information required, teleconsultation methods can range from telephone consultation to advanced hardware and imaging devices. At the simplest level, a telephone call to share clinical information can suffice, and this currently occurs in numerous settings where face-to-face acute evaluations are not possible. Limitations of this method are the inability to personally visualize the degree of clinical deficit, view the relevant neuro-imaging data, confirm the patient history, or have face-to-face interactions with patients, family members, and the treating bedside physicians. This method has been used to successfully support emergency physician delivery of t-PA in one model (27). Concerns remain, however, on the widespread acceptance of this strategy as an effective consultative mechanism in stroke.
High bandwidth data transmission allows real-time video consultation to support truly interactive patient evaluation and management. It can be performed with off-the-shelf commercially available systems that utilize integrated services digital network (ISDN) lines or digital subscriber lines (DSL) with internet/intranet protocols for data transfer. These systems can generate 30-frames-per-second video resolution and should support such features as remote pan / tilt / zoom to allow full remote control of the camera functions. These technologies have been tested and are currently feasible for the remote evaluation of patients (28).
Today's telemedicine systems serve point-to-point, hard-wired connections, but newer wireless and mobile systems are being developed and evaluated. Software based systems should be designed to facilitate the collection of critical patient data, assist in decision-support, and create documents to meet medical, legal, and regulatory requirements. Further research regarding these newer technologies should be addressed.
Implementation of teleconsultation requires installation of essential equipment, availability of the relevant telecommunications infrastructure, personnel training, technical support, and periodic maintenance. Teleconsultation must be integrated into the acute stroke evaluation process so that activation of the remote stroke team occurs in much the same manner as that of a local stroke team. When a potential acute stroke patient presents in the emergency or hospital setting, an initial evaluation by the bedside treating physician is performed. If special expertise is indicated, but not locally available, the remote system can be activated and the consultation can take place via telemedicine technologies.
Multiple practical barriers have thus far limited the widespread use of teleconsultation technology. The need for data transfer systems that can support high bandwidth transmissions (LAN/ ISDN/TCP-IP) has limited its use in some rural areas; however, this bandwidth is rapidly becoming ubiquitous. Current state medical board regulations define the site of health care delivery as the patient location, preventing physicians from other states from providing teleconsultation into states in which they are not licensed. Some malpractice carriers are wary of the liability risks associated with teleconsultation and may impose restrictions to coverage.
Additionally, most third-party payers have been slow to recognize these activities as eligible for reimbursement, whereas others, such as the Centers for Medicare and Medicaid Services, have provided reimbursement based solely on hospital location without reference to the availability of resources specific to stroke. Because these teleconsultation methods are not yet widely accepted in acute stroke care, many practitioners obtain informed consent from patients or their designated caregivers prior to initiating these consultation services, imposing additional time delays in the care of the patient with acute stroke.
Management Decision Support
Clinical management decision-making is often complex and involves the mechanics of health care delivery as well as institutional policies, regional practices, and referral patterns. The process becomes even more difficult when time limitations are added.
In stroke patients, once the initial neurologic assessment and diagnosis is established and radiologic interpretation obtained, the treating physician still faces fundamental issues regarding what, if any, interventional therapy is appropriate. Who is responsible for administering thrombolytics in eligible patients? Which patients with intracranial hemorrhage would benefit from surgical consultation? Who will assume care of the patient following ED evaluation or treatment? The wide variability in stroke experience among individual physicians and limitations in local emergency/hospital stroke care delivery systems - combined with the potential medicolegal risk of failure to treat or inappropriate treatment - has impeded widespread adoption of thrombolytic stroke therapy.
The availability of patient management support systems can assist the treating physician in achieving the essential care needed for patients with acute stroke. The need for such support can clearly be found in the following statement by the American College of Emergency Physicians regarding the use of t-PA in acute ischemic stroke:
"There is insufficient evidence at this time to endorse the use of intravenous t-PA in clinical practice when systems are not in place to ensure that the inclusion/exclusion criteria established by the NINDS guidelines for t-PA use in acute stroke are followed. Therefore, the decision for an emergency department to use intravenous t-PA for acute stroke should begin at the institutional level with commitments from hospital administration, the emergency department, neurology, neurosurgery, radiology, and laboratory services to ensure that the systems necessary for the safe use of fibrinolytic agents are in place" (29).
Options for management decision support may be conceptually divided into systems or elements available locally to the treating physician; those located in off-site (remote) locations; those facilitating patient transfer to a higher level of stroke care; and those providing patients access to clinical trial participation.
Local Decision Support
Local decision support identifies processes or elements which assist physician decision-making within a hospital or health care system. Pre-planned guidelines developed by all involved groups, such as family physicians, internists, neurologists, neurosurgeons, radiologists, and emergency physicians, are a critical initial step for developing a standardized approach to acute stroke management. All stake-holders, including nursing and ancillary services, should participate in their development.
Potential options for delivering such support include web-based and PDA-based tools (30) in addition to traditional paper-based formats. In the setting of emergent thrombolytic decision making, these can assist with information on patient eligibility, risk/benefit information for providers and patients, dosing information, and care guidelines. The web- and PDA-based tools have the added potential of supporting interactive management tools, allowing pre-programmed algorithms to support treatment decisions in the acute setting. Such protocols, in certain settings, can empower emergency physicians to initiate interventional care in the most expedient manner possible.
In circumstances requiring additional expertise in therapeutic decision making, the most familiar approach is the use of "in-house" specialty consultation. The reasons for specialty consultation in stroke are numerous and include: understanding co-morbid conditions, providing effective communication on the current diagnosis and treatment plan, discussing preferred treatment options, enlisting additional treatment assistance, and/or enhancing completion of a recommended course of action and disposition decisions (31). The use of standard consultation methods, whether by phone or in person, can be effective in this setting (27) but is dependent on specialist availability, technical capability, and experience.
An extension of this standard consultation support is found in hospitals using the "code-stroke" or "brain-attack" strategy to activate multiple resources, including neurologists - who often assume both ED decision-making and inpatient care roles. These teams have demonstrated effectiveness but are labor-intensive in resource utilization. Development of such teams within a hospital has been outlined by several groups (8,32,33).
Remote Decision Support
For facilities without the consistent local specialty consultation needed for a particular aspect of stroke care, processes must be in place to obtain off-site assistance. Such support might be as simple as telephone consultation with regional stroke centers (34) or as technically complex as the telemedicine solutions outlined earlier. The identification of regional stroke referral centers with access information should be immediately available in the ED. Just as medical staff must obtain assistance in managing patients with multiple trauma, so too must institutional and medical staff initiate pre-planning if time-limited interventional therapies are to be available to stroke patients.
Transfer Protocol Support
Hospital managers need to identify in advance the closest stroke referral centers, along with the most expedient means to transport patients needing a higher level of care. Protocols should exist to enable rapid contact with staff at the receiving institution, as well as ambulance agencies providing appropriate transfer level of care, so that little time is lost arranging acceptance and transport. Important administrative issues in patient transfer include:
All of these items must accompany the patient during transport. Transfer forms must include signatures from the patient or next of kin to ensure compliance with the Emergency Medical Treatment and Labor Act (EMTALA) requirements.
Clinical Trial Access
While developing systems for local and remote support of clinical decisions, consideration should be given to encouraging participation in clinical trials for acute stroke. Facilities offering participation in clinical trials are more likely to implement new therapies if they are proven to be beneficial. Participation in clinical trial networks promotes a sense of community commitment to improve delivery of acute stroke care. Clinical trials also provide the infrastructure to support acute stroke management and help highlight the importance of aggressive early care. Ideally, consideration of clinical trials should be integrated into clinical management algorithms, whether paper or electronic based. An electronic system offers the advantage of using automatic flags to identify patients eligible for research protocols. Manual systems can replicate this with intensive education and advertising about clinical trial eligibility.
Implementation and System Evaluation Support
To be effective in reducing stroke morbidity and mortality, support systems must not only be created, they must also be implemented. As outlined previously, multiple barriers at all levels of the health care system lead to sub-optimal recognition and early management of stroke. A concerted effort on the part of all health care workers and health care administrators will be necessary to overcome these barriers.
Delivery System Analysis and Implementation
To successfully implement an effective support system for frontline providers each component of the chain of recovery for acute stroke must be reviewed. This process also creates the expectation of high-quality performance from the various stakeholders in the delivery system. Recommendations for system analysis include focus groups (for medical staff and administrators) and the use of checklists, patient simulations, and/or external reviewers. Each of these processes can identify specific areas for improvement. A clearly identified local leader for the process may be instrumental in implementing a successful change in acute stroke management behavior.
Focus groups are effective in identifying barriers to delivery of stroke care and are commonly used in the ED setting to evaluate the level and consistency of care in other medical situations. It is a familiar process and requires few additional resources to conduct beyond a facilitator and staff time. Potential groups include EMS personnel; ED nursing, physician, and ancillary personnel; radiology technicians and physician staff; neurology, neurosurgery, and pharmacy staff; intensive care unit personnel; and cardiac arrest team representatives. Independent of a focus group evaluation checklists may be used to ensure availability of needed resources - triage assessment tools, emergent care pathways, inpatient care plans - for stroke patients.
Patient simulations using mock "Code Stroke" events offer the advantage of actually testing a system design and can be done for a variety of patient subtypes, including:
The mode of patient arrival or location at onset may reveal important system status information and should be varied in these simulations to include:
The benefits of such testing include identifying areas of suboptimal access to care and increasing the visibility of the therapeutic options in stroke in multiple health care settings. Such test situations also offer an effective means for regularly evaluating an established acute stroke treatment system. Finally, external review by individuals with acute stroke care expertise may assist in improving system design and care delivery.
Following development and identification of the support systems available to assist in the care of the stroke patient, their existence must be communicated clearly to all involved personnel. This can occur via regular staff meetings, mock drills, in-service meetings, and new-hire training materials. The opportunity to utilize repetitive forms of communications for educational reminders is encouraged, particularly email.
Additional physician education in the management of acute stroke will also be needed. In a 1999 survey of emergency medicine residents, 89 percent indicated their willingness to use t-PA with either telephone or bedside neurology consultation, while only 8 percent would administer the drug without a consult (35). A recent survey of graduating neurology residents found only 73 percent comfortable giving t-PA for stroke independently (36). Thus, it is not surprising to find a wide variation in attending physician practice with respect to the acute treatment of stroke. In addition to strengthening residency training programs in stroke care there is an additional need to provide continuing medical education materials to those already in practice.
The formal process of credentialing new staff and re-credentialing current staff is a potentially powerful tool in the effort to increase staff knowledge about stroke care pathways, algorithms, and resources. For physicians with a significant probability of caring for patients with acute stroke a clear attestation of their diagnostic and treatment capability is recommended.
From an institutional perspective, all health facilities providing stroke care need to establish policies supporting the chain of recovery for acute stroke. Opportunities exist for such organizations as the Joint Commission on Hospital Accreditation to champion this cause as they have done for other health care issues. Requirements for evidence of a hospital's stroke care plan should be established with criteria for outcome assessment and this should be linked to the institution's quality assurance / improvement program.
Establishing Support System Effectiveness
Considering the functional specifications of optimal stroke care, and the variety of system tools that can serve to fulfill those specifications, it may seem self-evident what needs to be done to improve local acute stroke practice. However, initial and ongoing assessment is indispensable for assuring that the implementation approach chosen is effective in achieving this goal.
Stroke care systems need to maintain a mechanism for ongoing assessment of performance. This can follow the structure/process/outcome formulation, focusing on aspects essential to satisfying the functional specifications of optimal stroke care. An example of a structure measure might be whether a person skilled in reading head CT scans is available 24 hours a day, or whether a "care map" or other description of expected stroke management is accessible to relevant care providers. Process measures could include the time from "door to needle" in patients receiving t-PA, rate of t-PA use, referral to physical therapy within 48 hours, or appropriate assessment for carotid testing. Outcome measures could potentially include blood pressure or functional status on discharge.
Some form of database or registry is valuable for assuring that the hospital has captured the population of stroke patients for whom stroke systems may be relevant. What is less clear is whether a stroke registry can serve all the needs for ongoing assessment for the majority of hospitals that do not have comprehensive electronic medical record systems. In the absence of an electronic medical record system, experience suggests that a special-purpose stroke registry should have a minimal number of data elements; the choice of elements should only be sufficient to guide the selection of a sample of patients for either a prospective evaluation or a retrospective chart review of a subset of patients. These may possibly be implemented through existing hospital quality assurance mechanisms.
Improved resources to evaluate the impact of the cumulative effect of the above strategies in improving acute stroke care in the United States are recommended. Research into the optimum methods of creating medical care delivery systems to reduce morbidity and mortality from stroke is also encouraged. Effective development, implementation, and monitoring of stroke support systems can improve the care of patients in the U.S. health system.
Appendix: Suggested Sites for Additional Stroke Support System Information
National Institute of Neurological Disorders and Stroke
Brain Attack Coalition
List of Professional Society Guidelines for Stroke Treatment
American Stroke Association
The Internet Stroke Center at Washington University
National Stroke Association, professional resources
Foundation for Education and Research in Neurologic Emergencies
American Academy of Neurology, stroke clinical assessment instruments
Return to Table of Contents
National Institute of Neurological Disorders and Stroke
National Institutes of Health
Bethesda, MD 20892
Last updated March 20, 2015