Despite the considerable national attention drawn to the need for improved secondary stroke prevention, a gap remains between evidence and application for stroke and other vascular events. Experience with the Coverdell stroke registry has shown that a minority of acute stroke patients receive the care recommended in established guidelines.1 Data collected from 4 registry centers in the United States showed a consistent lack of appropriate diagnostics, patient education, and initiation of drug therapies proven to reduce the risk of recurrent stroke.1

According to a report from the Committee on the Quality of Healthcare in America published in 2001, suboptimal treatment as well as inefficient use of health resources can be largely attributed to fragmentation of health care delivery in the management of various diseases in the United States.2 In response to these findings, the American Stroke Association (ASA) has established recommendations for the development of stroke systems of care. The objective of a systems approach is to integrate preventive and treatment services and provide patients with evidence‐based care.3

During hospitalization for acute stroke, immediate treatment must focus on minimizing stroke progression, avoiding common complications, and preventing recurrent stroke. Prior to discharge, patients need to be educated about the importance of lifestyle modifications and pharmacotherapies to reduce their risk of a recurrence of the stroke and other atherosclerotic vascular events.3 As the physicians who focus on inpatient care, hospitalists are likely to be responsible for participating in and coordinating the multidisciplinary team that provides treatment and services to stroke patients. Hospitalists also must facilitate the transition from inpatient to outpatient care. Hospitalists are in a position to help educate stroke patients about prevention strategies throughout the hospitalization period. These functions provide hospitalists with the opportunity to lead, coordinate, and participate in stroke systems care at their institutions.

The present article discusses the components of stroke systems care recommended by the ASA and the best‐practices recommendations from the recent hospitalist roundtable discussion on routine acute stroke care. The national treatment guidelines and clinical trials supporting the recommendations of the hospitalist roundtable participants have been discussed in the article in this supplement by Dr. Likosky et al, as well as in the patient scenarios article in this supplement by Dr. Lee et al. Some of the anticipated barriers and pitfalls that may be encountered, along with potential solutions, are also discussed. Hospitalists may be able to use this review to adapt feasible components of the systems care for stroke management to improve care at their institutions.

WHAT IS STROKE SYSTEMS CARE?

A stroke system is coordinated stroke care along the entire continuum from primary prevention to rehabilitation. Postemergency department inpatient care for patients with acute stroke, also referred to as subacute care, is only one component of the community‐based stroke systems of care recommended by the ASA (Fig. 1).3 In this model, regional stroke systems identify hospitals that are acute stroke capable and determine that those institutions use clinical pathways that reflect well‐established standards of care and nationally recognized guidelines.3 In this broad sense of the term, stroke systems function to organize and coordinate the various agencies and health care providers responsible for caring for patients with stroke, from the first call to emergency services through postdischarge medical care and rehabilitation (Table 1). The subacute phase of care provides the bridge from management of the medical emergency to discharge and is central to secondary stroke prevention.

Figure 1

RATIONALE FOR HOSPITAL‐BASED STROKE SYSTEMS

The Preventing Recurrence of Thrombo‐embolic Events through Coordinated Treatment (PROTECT) program provides proof of concept.4 The PROTECT program was implemented at a large teaching hospital to improve diagnosis, treatment, and secondary prevention for patients with ischemic stroke.4 Four medication goals and instruction in 4 lifestyle interventions were chosen as indicators of program impact. In the first year after PROTECT was started, 100% of eligible patients received instruction in all 4 areas of lifestyle change prior to discharge.4 In the year following implementation of PROTECT, the rate of appropriate prescribing of antithrombotics was 98%. Appropriate prescribing of angiotensin‐converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs), statins, and thiazide diuretics was significantly increased from pre‐PROTECT levels.4 After 3 months of follow‐up, patient adherence to therapy remained high.5 The final results of the PROTECT program are not yet available; however, it is reasonable to expect that increased use of evidence‐based therapy and good patient adherence to these proven therapies will have led to better patient outcomes, including lower rates of recurrent stroke.

Patient outcomes data are available for a related initiative for treatment of patients hospitalized with myocardial infarction. Compared with the year prior to implementation of the Cardiac Hospitalization Atherosclerotic Management Program (CHAMP), more patients who were involved in the CHAMP intervention achieved low‐density lipoprotein cholesterol levels P < .001). In addition, these patients achieved a 57% reduction in recurrent myocardial infarction.6

These 2 studies indicate a benefit of establishing hospital‐based stroke systems; however, these studies are the initial steps, and each has limitations. For example, neither study was a prospective, randomized trial with a concurrent control group.4, 6 In addition, PROTECT data were not evaluated by independent audit but by individuals who were aware of the program goals, and limited data were available regarding contraindications to therapy.4 CHAMP did not assess adherence to nonpharmacologic interventions or the effect of surgical interventions.6 Large, randomized, controlled trials are needed to better understand the impact of such systems. Although larger evidence‐based trials are needed, it is important to review available information on stroke systems to adapt those components that align with each institution's available resources.

ESTABLISHING HOSPITAL‐BASED STROKE SYSTEMS

Several barriers exist to establishing a stroke systems care program, as detailed in Table 2. The support and involvement of the hospital administration is essential to success, as is multidisciplinary agreement that such a program will benefit patients.

Other potential points of resistance revolve around the financial impact of implementing a stroke systems approach to care. The proposed stroke systems care plan is consistent with meeting nationally recognized quality improvement standards; however, the current health care market forces demand accountability for health care expenditures. Increasingly, payers are turning to the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and the National Committee for Quality Assurance (NCQA) evaluations to determine quality of care at various institutions. These programs encourage the use of standardized treatment protocols consistent with the concept of systems approach to care. Moreover, stroke care is a JCAHO quality measure and thus may have a financial impact on hospitals. It is possible that implementing standardized procedures for stroke care may reduce the cost of care. Information about the JCAHO Disease Specific Certification for Acute Stroke Care can be accessed at http://www.jointcommission.org/.7

Once there is agreement that a stroke system should be developed, a multidisciplinary team should be established. A multidisciplinary team may include hospitalists, neurologists, neurosurgeons, emergency medicine physicians, diagnostic and interventional radiologists, nurses, physiotherapists, occupational therapists, speech and language therapists, and social workers. However, the components of the multidisciplinary team may vary depending on the available staff and financial resources at different stroke centers. Assuring all participants in the system that their input is valued can improve communication among stroke specialists, hospitalists, and primary care clinicians. This team is responsible for evaluation of current procedures and development of algorithms, discharge forms, patient education, and preprinted orders.

The task of developing a cohesive plan for stroke care may appear onerous. Existing diagnostic and treatment procedures may be poorly designed or organized. However, multiple online sources provide tools for every aspect of stroke systems care. Information about evidence‐based stroke care practices is available as part of the American Heart Association (AHA)/ASA Get with the GuidelinesStroke program and can be accessed at http://www.strokeassociation.org.8 Standardized admission orders, patient education materials, and data‐monitoring methods are available at this site. Hospitalist‐specific information may also be found in the Stroke Resource Room of the Society of Hospital Medicine Web site (http://www.hospitalmedicine.org).9 A similar array of tools can be found at the PROTECT Web site (http://strokeprotect.mednet.ucla.edu).10 These materials are designed to facilitate integration of secondary prevention practices into stroke systems care.

A stroke system of care is a dynamic process. The multidisciplinary team may also be responsible for continuous monitoring and reporting of the efficiency and impact of the system and providing feedback to other staff and administration. Protocols should be revised regularly to account for new evidence‐based treatments and to streamline their use. The Canadian Stroke Systems Coalition recommends that a comprehensive and efficient system include prevention, prehospital and emergency care, hospital care, rehabilitation, reintegration into the community, surveillance, and research.11 Hospital staff should be educated in core competencies in hospital medicine as well as any changes to protocols made over time. Protocols that facilitate communication among health care providers should also be developed, and hospitalists may play a central role in this process. Accurate and timely transfer of patient information from the emergency department to the stroke center or ward is imperative.

FOCUSING ON INPATIENT CARE

Clinical pathways for inpatient care should be designed to limit stroke progression as much as possible.3 The Brain Attack Coalition (BAC) provides a resource for clinical pathways implemented at various institutes in the United States, including the Stanford Stroke Center, the Cleveland Clinic Foundation, and Thomas Jefferson University Hospital, among others (http://stroke‐site.org/pathways/pathways.html).12 Between 30% and 40% of acute stroke patients will show signs of stroke progression in the first week following stroke onset. Stroke can affect multiple systems, and symptom progression as a result of cerebral edema, hemorrhagic conversion, or additional strokes may require surgical intervention and warrants consultation with a neurologist.13

A neurologist should be available to the stroke system patients at all times, and ideally, all acute stroke patients should be evaluated by a neurologist specializing in the evaluation and treatment of patients with stroke.14 There are several stroke scales available to evaluate stroke patients, including the Barthel Index, the Glasgow outcome scale, the Modified Rankin Scale, the National Institutes of Health Stroke Scale, and the Hunt and Hess Classification of Subarachnoid Hemorrhage (http:// www.stroke‐site.org/stroke_scales/stroke_scales. html).15 These scales aid in the assessment of severity of stroke and may be helpful for hospitalists and other health care providers to recognize when a neurology consult is needed. The team should agree on criteria that indicate a neurological emergency, such as stroke progression, and other reasons for immediate consultation with a neurologist, such as worsening of stroke, stroke in a young patient, or uncertain diagnosis. These criteria should be included as an established part of the clinical pathway.

Common complications of stroke, such as myocardial infarction, deep vein thrombosis, pulmonary embolism, urinary tract infections, aspiration pneumonia, dehydration, poor nutrition, skin breakdown, and metabolic disorders, should be anticipated, and preventive steps should be taken. The measures to prevent the above complications of stroke need to be initiated in the emergency department.3

Management of existing comorbid conditions is another key part of subacute stroke care. Given that 85% of all hospitalists have a background in internal medicine, management of comorbid conditions such as diabetes and hypertension is an area in which hospitalists have professional competence. Patient history and use of prescription medications prior to stroke should be reviewed whenever possible and incorporated into short‐term and long‐term treatment plans. Patients with diabetes in particular may benefit more from rigorous control of blood pressure and lipids compared with other patients.16

Secondary stroke prevention should start as early as considered safe. Diagnosis of stroke subtype, often accomplished in the emergency department, establishes suitability for antithrombotics and optimal management strategy. Patients who receive a diagnosis of stroke secondary to cardioembolic atrial fibrillation should be treated with an anticoagulant after the acute period. Aspirin can be used for those individuals unable to use anticoagulants.16 For those individuals with stroke of noncardioembolic origin, particularly those with atherosclerosis and lacunar or cryptogenic infarcts, antiplatelet agents are recommended.14

A multimodal prevention strategy is recommended to manage blood pressure and dyslipidemia poststroke. An algorithm for managing blood pressure soon after stroke has been developed by the PROTECT program (Fig. 2).10 Antihypertensives, usually a combination of an ACE inhibitor and a thiazide diuretic, can be initiated at low doses 48‐72 hours after stroke. A longer delay is recommended for patients with large infarcts or evidence of uncontrolled hypertension. ARBs may be substituted for ACE inhibitors.10 Target blood pressures should be determined using the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.17 In general, even a reduction of 10/5 mm Hg has been shown to be beneficial.16

Figure 2

Statins are recommended for all patients with elevated serum lipids unless treatment with statins is contraindicated. The recommended target level for low‐density lipoprotein cholesterol is below 100 mg/dL for individuals with coronary heart disease and symptomatic atherosclerosis. A target below 70 mg/dL may be appropriate for patients at very high risk.16

Prior to discharge, patients or their caregivers should be given prescriptions adequate to cover the time until postdischarge follow‐up visit. The responsible persons need to be made aware that some medications such as antihypertensives will require dosage adjustments by an outpatient physician, and the timing of the follow‐up visit may need to be arranged accordingly.

The importance of stroke risk reduction should be part of predischarge patient education, along with a list of the warning signs of stroke. Adherence to the treatment regimen, including lifestyle changes and medications, should be emphasized. Patients or their caregivers should be educated about identifying adverse events and a plan to address them. Understanding that some adverse effects (eg, headache with aspirin plus extended‐release dipyridamole) are likely to be transient may prevent unnecessary discontinuation of treatment and reduce anxiety.

Patient and caregiver education can be reinforced by providing standardized patient education materials that can be found in the Stroke Resource Room at the Society of Hospital Medicine Web site (http://www.hospitalmedicine.org),9 PROTECT (http://strokeprotect.mednet.ucla.edu),10 ASA (http://www.strokeassociation.org),8 and AHA (http://www.americanheart.org)18 Web sites.

Transfer of patient information to outpatient health care providers is a critical step in stroke systems care. Notes indicating any need for medication dose adjustment must be included. Discharge summaries should be available to primary care providers, neurologists, and rehabilitation specialists prior to follow‐up visits. The use of electronic forms that can be faxed or sent by E‐mail can shorten delivery time considerably. In lieu of electronic delivery, physician letters can be used, and prototypes are available at the resource Web sites. Whenever possible, a follow‐up phone call to the primary care physician provides the best means to ensure clear communication.

SUMMARY

Hospitalists are well qualified to lead quality focused patient care initiatives at their institutions. Use of standardized protocols to reduce the risk of secondary stroke is proven to increase appropriate prescribing at discharge, which in turn improves patient adherence to evidence‐based therapy. Multidisciplinary communication, including communication with outpatient clinicians, facilitates the transition from inpatient to outpatient health care providers.

In addition to improving patient care, use of standardized protocols is tracked by JCAHO and offers assurance to payers that a particular hospital and its staff are committed to quality care. Establishing protocols is made relatively easy by the online availability of materials that can be adapted to various hospital settings.

Despite the considerable national attention drawn to the need for improved secondary stroke prevention, a gap remains between evidence and application for stroke and other vascular events. Experience with the Coverdell stroke registry has shown that a minority of acute stroke patients receive the care recommended in established guidelines.1 Data collected from 4 registry centers in the United States showed a consistent lack of appropriate diagnostics, patient education, and initiation of drug therapies proven to reduce the risk of recurrent stroke.1

According to a report from the Committee on the Quality of Healthcare in America published in 2001, suboptimal treatment as well as inefficient use of health resources can be largely attributed to fragmentation of health care delivery in the management of various diseases in the United States.2 In response to these findings, the American Stroke Association (ASA) has established recommendations for the development of stroke systems of care. The objective of a systems approach is to integrate preventive and treatment services and provide patients with evidence‐based care.3

During hospitalization for acute stroke, immediate treatment must focus on minimizing stroke progression, avoiding common complications, and preventing recurrent stroke. Prior to discharge, patients need to be educated about the importance of lifestyle modifications and pharmacotherapies to reduce their risk of a recurrence of the stroke and other atherosclerotic vascular events.3 As the physicians who focus on inpatient care, hospitalists are likely to be responsible for participating in and coordinating the multidisciplinary team that provides treatment and services to stroke patients. Hospitalists also must facilitate the transition from inpatient to outpatient care. Hospitalists are in a position to help educate stroke patients about prevention strategies throughout the hospitalization period. These functions provide hospitalists with the opportunity to lead, coordinate, and participate in stroke systems care at their institutions.

The present article discusses the components of stroke systems care recommended by the ASA and the best‐practices recommendations from the recent hospitalist roundtable discussion on routine acute stroke care. The national treatment guidelines and clinical trials supporting the recommendations of the hospitalist roundtable participants have been discussed in the article in this supplement by Dr. Likosky et al, as well as in the patient scenarios article in this supplement by Dr. Lee et al. Some of the anticipated barriers and pitfalls that may be encountered, along with potential solutions, are also discussed. Hospitalists may be able to use this review to adapt feasible components of the systems care for stroke management to improve care at their institutions.

WHAT IS STROKE SYSTEMS CARE?

A stroke system is coordinated stroke care along the entire continuum from primary prevention to rehabilitation. Postemergency department inpatient care for patients with acute stroke, also referred to as subacute care, is only one component of the community‐based stroke systems of care recommended by the ASA (Fig. 1).3 In this model, regional stroke systems identify hospitals that are acute stroke capable and determine that those institutions use clinical pathways that reflect well‐established standards of care and nationally recognized guidelines.3 In this broad sense of the term, stroke systems function to organize and coordinate the various agencies and health care providers responsible for caring for patients with stroke, from the first call to emergency services through postdischarge medical care and rehabilitation (Table 1). The subacute phase of care provides the bridge from management of the medical emergency to discharge and is central to secondary stroke prevention.

Figure 1

RATIONALE FOR HOSPITAL‐BASED STROKE SYSTEMS

The Preventing Recurrence of Thrombo‐embolic Events through Coordinated Treatment (PROTECT) program provides proof of concept.4 The PROTECT program was implemented at a large teaching hospital to improve diagnosis, treatment, and secondary prevention for patients with ischemic stroke.4 Four medication goals and instruction in 4 lifestyle interventions were chosen as indicators of program impact. In the first year after PROTECT was started, 100% of eligible patients received instruction in all 4 areas of lifestyle change prior to discharge.4 In the year following implementation of PROTECT, the rate of appropriate prescribing of antithrombotics was 98%. Appropriate prescribing of angiotensin‐converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs), statins, and thiazide diuretics was significantly increased from pre‐PROTECT levels.4 After 3 months of follow‐up, patient adherence to therapy remained high.5 The final results of the PROTECT program are not yet available; however, it is reasonable to expect that increased use of evidence‐based therapy and good patient adherence to these proven therapies will have led to better patient outcomes, including lower rates of recurrent stroke.

Patient outcomes data are available for a related initiative for treatment of patients hospitalized with myocardial infarction. Compared with the year prior to implementation of the Cardiac Hospitalization Atherosclerotic Management Program (CHAMP), more patients who were involved in the CHAMP intervention achieved low‐density lipoprotein cholesterol levels P < .001). In addition, these patients achieved a 57% reduction in recurrent myocardial infarction.6

These 2 studies indicate a benefit of establishing hospital‐based stroke systems; however, these studies are the initial steps, and each has limitations. For example, neither study was a prospective, randomized trial with a concurrent control group.4, 6 In addition, PROTECT data were not evaluated by independent audit but by individuals who were aware of the program goals, and limited data were available regarding contraindications to therapy.4 CHAMP did not assess adherence to nonpharmacologic interventions or the effect of surgical interventions.6 Large, randomized, controlled trials are needed to better understand the impact of such systems. Although larger evidence‐based trials are needed, it is important to review available information on stroke systems to adapt those components that align with each institution's available resources.

ESTABLISHING HOSPITAL‐BASED STROKE SYSTEMS

Several barriers exist to establishing a stroke systems care program, as detailed in Table 2. The support and involvement of the hospital administration is essential to success, as is multidisciplinary agreement that such a program will benefit patients.

Other potential points of resistance revolve around the financial impact of implementing a stroke systems approach to care. The proposed stroke systems care plan is consistent with meeting nationally recognized quality improvement standards; however, the current health care market forces demand accountability for health care expenditures. Increasingly, payers are turning to the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and the National Committee for Quality Assurance (NCQA) evaluations to determine quality of care at various institutions. These programs encourage the use of standardized treatment protocols consistent with the concept of systems approach to care. Moreover, stroke care is a JCAHO quality measure and thus may have a financial impact on hospitals. It is possible that implementing standardized procedures for stroke care may reduce the cost of care. Information about the JCAHO Disease Specific Certification for Acute Stroke Care can be accessed at http://www.jointcommission.org/.7

Once there is agreement that a stroke system should be developed, a multidisciplinary team should be established. A multidisciplinary team may include hospitalists, neurologists, neurosurgeons, emergency medicine physicians, diagnostic and interventional radiologists, nurses, physiotherapists, occupational therapists, speech and language therapists, and social workers. However, the components of the multidisciplinary team may vary depending on the available staff and financial resources at different stroke centers. Assuring all participants in the system that their input is valued can improve communication among stroke specialists, hospitalists, and primary care clinicians. This team is responsible for evaluation of current procedures and development of algorithms, discharge forms, patient education, and preprinted orders.

The task of developing a cohesive plan for stroke care may appear onerous. Existing diagnostic and treatment procedures may be poorly designed or organized. However, multiple online sources provide tools for every aspect of stroke systems care. Information about evidence‐based stroke care practices is available as part of the American Heart Association (AHA)/ASA Get with the GuidelinesStroke program and can be accessed at http://www.strokeassociation.org.8 Standardized admission orders, patient education materials, and data‐monitoring methods are available at this site. Hospitalist‐specific information may also be found in the Stroke Resource Room of the Society of Hospital Medicine Web site (http://www.hospitalmedicine.org).9 A similar array of tools can be found at the PROTECT Web site (http://strokeprotect.mednet.ucla.edu).10 These materials are designed to facilitate integration of secondary prevention practices into stroke systems care.

A stroke system of care is a dynamic process. The multidisciplinary team may also be responsible for continuous monitoring and reporting of the efficiency and impact of the system and providing feedback to other staff and administration. Protocols should be revised regularly to account for new evidence‐based treatments and to streamline their use. The Canadian Stroke Systems Coalition recommends that a comprehensive and efficient system include prevention, prehospital and emergency care, hospital care, rehabilitation, reintegration into the community, surveillance, and research.11 Hospital staff should be educated in core competencies in hospital medicine as well as any changes to protocols made over time. Protocols that facilitate communication among health care providers should also be developed, and hospitalists may play a central role in this process. Accurate and timely transfer of patient information from the emergency department to the stroke center or ward is imperative.

FOCUSING ON INPATIENT CARE

Clinical pathways for inpatient care should be designed to limit stroke progression as much as possible.3 The Brain Attack Coalition (BAC) provides a resource for clinical pathways implemented at various institutes in the United States, including the Stanford Stroke Center, the Cleveland Clinic Foundation, and Thomas Jefferson University Hospital, among others (http://stroke‐site.org/pathways/pathways.html).12 Between 30% and 40% of acute stroke patients will show signs of stroke progression in the first week following stroke onset. Stroke can affect multiple systems, and symptom progression as a result of cerebral edema, hemorrhagic conversion, or additional strokes may require surgical intervention and warrants consultation with a neurologist.13

A neurologist should be available to the stroke system patients at all times, and ideally, all acute stroke patients should be evaluated by a neurologist specializing in the evaluation and treatment of patients with stroke.14 There are several stroke scales available to evaluate stroke patients, including the Barthel Index, the Glasgow outcome scale, the Modified Rankin Scale, the National Institutes of Health Stroke Scale, and the Hunt and Hess Classification of Subarachnoid Hemorrhage (http:// www.stroke‐site.org/stroke_scales/stroke_scales. html).15 These scales aid in the assessment of severity of stroke and may be helpful for hospitalists and other health care providers to recognize when a neurology consult is needed. The team should agree on criteria that indicate a neurological emergency, such as stroke progression, and other reasons for immediate consultation with a neurologist, such as worsening of stroke, stroke in a young patient, or uncertain diagnosis. These criteria should be included as an established part of the clinical pathway.

Common complications of stroke, such as myocardial infarction, deep vein thrombosis, pulmonary embolism, urinary tract infections, aspiration pneumonia, dehydration, poor nutrition, skin breakdown, and metabolic disorders, should be anticipated, and preventive steps should be taken. The measures to prevent the above complications of stroke need to be initiated in the emergency department.3

Management of existing comorbid conditions is another key part of subacute stroke care. Given that 85% of all hospitalists have a background in internal medicine, management of comorbid conditions such as diabetes and hypertension is an area in which hospitalists have professional competence. Patient history and use of prescription medications prior to stroke should be reviewed whenever possible and incorporated into short‐term and long‐term treatment plans. Patients with diabetes in particular may benefit more from rigorous control of blood pressure and lipids compared with other patients.16

Secondary stroke prevention should start as early as considered safe. Diagnosis of stroke subtype, often accomplished in the emergency department, establishes suitability for antithrombotics and optimal management strategy. Patients who receive a diagnosis of stroke secondary to cardioembolic atrial fibrillation should be treated with an anticoagulant after the acute period. Aspirin can be used for those individuals unable to use anticoagulants.16 For those individuals with stroke of noncardioembolic origin, particularly those with atherosclerosis and lacunar or cryptogenic infarcts, antiplatelet agents are recommended.14

A multimodal prevention strategy is recommended to manage blood pressure and dyslipidemia poststroke. An algorithm for managing blood pressure soon after stroke has been developed by the PROTECT program (Fig. 2).10 Antihypertensives, usually a combination of an ACE inhibitor and a thiazide diuretic, can be initiated at low doses 48‐72 hours after stroke. A longer delay is recommended for patients with large infarcts or evidence of uncontrolled hypertension. ARBs may be substituted for ACE inhibitors.10 Target blood pressures should be determined using the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.17 In general, even a reduction of 10/5 mm Hg has been shown to be beneficial.16

Figure 2

Statins are recommended for all patients with elevated serum lipids unless treatment with statins is contraindicated. The recommended target level for low‐density lipoprotein cholesterol is below 100 mg/dL for individuals with coronary heart disease and symptomatic atherosclerosis. A target below 70 mg/dL may be appropriate for patients at very high risk.16

Prior to discharge, patients or their caregivers should be given prescriptions adequate to cover the time until postdischarge follow‐up visit. The responsible persons need to be made aware that some medications such as antihypertensives will require dosage adjustments by an outpatient physician, and the timing of the follow‐up visit may need to be arranged accordingly.

The importance of stroke risk reduction should be part of predischarge patient education, along with a list of the warning signs of stroke. Adherence to the treatment regimen, including lifestyle changes and medications, should be emphasized. Patients or their caregivers should be educated about identifying adverse events and a plan to address them. Understanding that some adverse effects (eg, headache with aspirin plus extended‐release dipyridamole) are likely to be transient may prevent unnecessary discontinuation of treatment and reduce anxiety.

Patient and caregiver education can be reinforced by providing standardized patient education materials that can be found in the Stroke Resource Room at the Society of Hospital Medicine Web site (http://www.hospitalmedicine.org),9 PROTECT (http://strokeprotect.mednet.ucla.edu),10 ASA (http://www.strokeassociation.org),8 and AHA (http://www.americanheart.org)18 Web sites.

Transfer of patient information to outpatient health care providers is a critical step in stroke systems care. Notes indicating any need for medication dose adjustment must be included. Discharge summaries should be available to primary care providers, neurologists, and rehabilitation specialists prior to follow‐up visits. The use of electronic forms that can be faxed or sent by E‐mail can shorten delivery time considerably. In lieu of electronic delivery, physician letters can be used, and prototypes are available at the resource Web sites. Whenever possible, a follow‐up phone call to the primary care physician provides the best means to ensure clear communication.

SUMMARY

Hospitalists are well qualified to lead quality focused patient care initiatives at their institutions. Use of standardized protocols to reduce the risk of secondary stroke is proven to increase appropriate prescribing at discharge, which in turn improves patient adherence to evidence‐based therapy. Multidisciplinary communication, including communication with outpatient clinicians, facilitates the transition from inpatient to outpatient health care providers.

In addition to improving patient care, use of standardized protocols is tracked by JCAHO and offers assurance to payers that a particular hospital and its staff are committed to quality care. Establishing protocols is made relatively easy by the online availability of materials that can be adapted to various hospital settings.

Despite the considerable national attention drawn to the need for improved secondary stroke prevention, a gap remains between evidence and application for stroke and other vascular events. Experience with the Coverdell stroke registry has shown that a minority of acute stroke patients receive the care recommended in established guidelines.1 Data collected from 4 registry centers in the United States showed a consistent lack of appropriate diagnostics, patient education, and initiation of drug therapies proven to reduce the risk of recurrent stroke.1

According to a report from the Committee on the Quality of Healthcare in America published in 2001, suboptimal treatment as well as inefficient use of health resources can be largely attributed to fragmentation of health care delivery in the management of various diseases in the United States.2 In response to these findings, the American Stroke Association (ASA) has established recommendations for the development of stroke systems of care. The objective of a systems approach is to integrate preventive and treatment services and provide patients with evidence‐based care.3

During hospitalization for acute stroke, immediate treatment must focus on minimizing stroke progression, avoiding common complications, and preventing recurrent stroke. Prior to discharge, patients need to be educated about the importance of lifestyle modifications and pharmacotherapies to reduce their risk of a recurrence of the stroke and other atherosclerotic vascular events.3 As the physicians who focus on inpatient care, hospitalists are likely to be responsible for participating in and coordinating the multidisciplinary team that provides treatment and services to stroke patients. Hospitalists also must facilitate the transition from inpatient to outpatient care. Hospitalists are in a position to help educate stroke patients about prevention strategies throughout the hospitalization period. These functions provide hospitalists with the opportunity to lead, coordinate, and participate in stroke systems care at their institutions.

The present article discusses the components of stroke systems care recommended by the ASA and the best‐practices recommendations from the recent hospitalist roundtable discussion on routine acute stroke care. The national treatment guidelines and clinical trials supporting the recommendations of the hospitalist roundtable participants have been discussed in the article in this supplement by Dr. Likosky et al, as well as in the patient scenarios article in this supplement by Dr. Lee et al. Some of the anticipated barriers and pitfalls that may be encountered, along with potential solutions, are also discussed. Hospitalists may be able to use this review to adapt feasible components of the systems care for stroke management to improve care at their institutions.

WHAT IS STROKE SYSTEMS CARE?

A stroke system is coordinated stroke care along the entire continuum from primary prevention to rehabilitation. Postemergency department inpatient care for patients with acute stroke, also referred to as subacute care, is only one component of the community‐based stroke systems of care recommended by the ASA (Fig. 1).3 In this model, regional stroke systems identify hospitals that are acute stroke capable and determine that those institutions use clinical pathways that reflect well‐established standards of care and nationally recognized guidelines.3 In this broad sense of the term, stroke systems function to organize and coordinate the various agencies and health care providers responsible for caring for patients with stroke, from the first call to emergency services through postdischarge medical care and rehabilitation (Table 1). The subacute phase of care provides the bridge from management of the medical emergency to discharge and is central to secondary stroke prevention.

Figure 1

RATIONALE FOR HOSPITAL‐BASED STROKE SYSTEMS

The Preventing Recurrence of Thrombo‐embolic Events through Coordinated Treatment (PROTECT) program provides proof of concept.4 The PROTECT program was implemented at a large teaching hospital to improve diagnosis, treatment, and secondary prevention for patients with ischemic stroke.4 Four medication goals and instruction in 4 lifestyle interventions were chosen as indicators of program impact. In the first year after PROTECT was started, 100% of eligible patients received instruction in all 4 areas of lifestyle change prior to discharge.4 In the year following implementation of PROTECT, the rate of appropriate prescribing of antithrombotics was 98%. Appropriate prescribing of angiotensin‐converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs), statins, and thiazide diuretics was significantly increased from pre‐PROTECT levels.4 After 3 months of follow‐up, patient adherence to therapy remained high.5 The final results of the PROTECT program are not yet available; however, it is reasonable to expect that increased use of evidence‐based therapy and good patient adherence to these proven therapies will have led to better patient outcomes, including lower rates of recurrent stroke.

Patient outcomes data are available for a related initiative for treatment of patients hospitalized with myocardial infarction. Compared with the year prior to implementation of the Cardiac Hospitalization Atherosclerotic Management Program (CHAMP), more patients who were involved in the CHAMP intervention achieved low‐density lipoprotein cholesterol levels P < .001). In addition, these patients achieved a 57% reduction in recurrent myocardial infarction.6

These 2 studies indicate a benefit of establishing hospital‐based stroke systems; however, these studies are the initial steps, and each has limitations. For example, neither study was a prospective, randomized trial with a concurrent control group.4, 6 In addition, PROTECT data were not evaluated by independent audit but by individuals who were aware of the program goals, and limited data were available regarding contraindications to therapy.4 CHAMP did not assess adherence to nonpharmacologic interventions or the effect of surgical interventions.6 Large, randomized, controlled trials are needed to better understand the impact of such systems. Although larger evidence‐based trials are needed, it is important to review available information on stroke systems to adapt those components that align with each institution's available resources.

ESTABLISHING HOSPITAL‐BASED STROKE SYSTEMS

Several barriers exist to establishing a stroke systems care program, as detailed in Table 2. The support and involvement of the hospital administration is essential to success, as is multidisciplinary agreement that such a program will benefit patients.

Other potential points of resistance revolve around the financial impact of implementing a stroke systems approach to care. The proposed stroke systems care plan is consistent with meeting nationally recognized quality improvement standards; however, the current health care market forces demand accountability for health care expenditures. Increasingly, payers are turning to the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and the National Committee for Quality Assurance (NCQA) evaluations to determine quality of care at various institutions. These programs encourage the use of standardized treatment protocols consistent with the concept of systems approach to care. Moreover, stroke care is a JCAHO quality measure and thus may have a financial impact on hospitals. It is possible that implementing standardized procedures for stroke care may reduce the cost of care. Information about the JCAHO Disease Specific Certification for Acute Stroke Care can be accessed at http://www.jointcommission.org/.7

Once there is agreement that a stroke system should be developed, a multidisciplinary team should be established. A multidisciplinary team may include hospitalists, neurologists, neurosurgeons, emergency medicine physicians, diagnostic and interventional radiologists, nurses, physiotherapists, occupational therapists, speech and language therapists, and social workers. However, the components of the multidisciplinary team may vary depending on the available staff and financial resources at different stroke centers. Assuring all participants in the system that their input is valued can improve communication among stroke specialists, hospitalists, and primary care clinicians. This team is responsible for evaluation of current procedures and development of algorithms, discharge forms, patient education, and preprinted orders.

The task of developing a cohesive plan for stroke care may appear onerous. Existing diagnostic and treatment procedures may be poorly designed or organized. However, multiple online sources provide tools for every aspect of stroke systems care. Information about evidence‐based stroke care practices is available as part of the American Heart Association (AHA)/ASA Get with the GuidelinesStroke program and can be accessed at http://www.strokeassociation.org.8 Standardized admission orders, patient education materials, and data‐monitoring methods are available at this site. Hospitalist‐specific information may also be found in the Stroke Resource Room of the Society of Hospital Medicine Web site (http://www.hospitalmedicine.org).9 A similar array of tools can be found at the PROTECT Web site (http://strokeprotect.mednet.ucla.edu).10 These materials are designed to facilitate integration of secondary prevention practices into stroke systems care.

A stroke system of care is a dynamic process. The multidisciplinary team may also be responsible for continuous monitoring and reporting of the efficiency and impact of the system and providing feedback to other staff and administration. Protocols should be revised regularly to account for new evidence‐based treatments and to streamline their use. The Canadian Stroke Systems Coalition recommends that a comprehensive and efficient system include prevention, prehospital and emergency care, hospital care, rehabilitation, reintegration into the community, surveillance, and research.11 Hospital staff should be educated in core competencies in hospital medicine as well as any changes to protocols made over time. Protocols that facilitate communication among health care providers should also be developed, and hospitalists may play a central role in this process. Accurate and timely transfer of patient information from the emergency department to the stroke center or ward is imperative.

FOCUSING ON INPATIENT CARE

Clinical pathways for inpatient care should be designed to limit stroke progression as much as possible.3 The Brain Attack Coalition (BAC) provides a resource for clinical pathways implemented at various institutes in the United States, including the Stanford Stroke Center, the Cleveland Clinic Foundation, and Thomas Jefferson University Hospital, among others (http://stroke‐site.org/pathways/pathways.html).12 Between 30% and 40% of acute stroke patients will show signs of stroke progression in the first week following stroke onset. Stroke can affect multiple systems, and symptom progression as a result of cerebral edema, hemorrhagic conversion, or additional strokes may require surgical intervention and warrants consultation with a neurologist.13

A neurologist should be available to the stroke system patients at all times, and ideally, all acute stroke patients should be evaluated by a neurologist specializing in the evaluation and treatment of patients with stroke.14 There are several stroke scales available to evaluate stroke patients, including the Barthel Index, the Glasgow outcome scale, the Modified Rankin Scale, the National Institutes of Health Stroke Scale, and the Hunt and Hess Classification of Subarachnoid Hemorrhage (http:// www.stroke‐site.org/stroke_scales/stroke_scales. html).15 These scales aid in the assessment of severity of stroke and may be helpful for hospitalists and other health care providers to recognize when a neurology consult is needed. The team should agree on criteria that indicate a neurological emergency, such as stroke progression, and other reasons for immediate consultation with a neurologist, such as worsening of stroke, stroke in a young patient, or uncertain diagnosis. These criteria should be included as an established part of the clinical pathway.

Common complications of stroke, such as myocardial infarction, deep vein thrombosis, pulmonary embolism, urinary tract infections, aspiration pneumonia, dehydration, poor nutrition, skin breakdown, and metabolic disorders, should be anticipated, and preventive steps should be taken. The measures to prevent the above complications of stroke need to be initiated in the emergency department.3

Management of existing comorbid conditions is another key part of subacute stroke care. Given that 85% of all hospitalists have a background in internal medicine, management of comorbid conditions such as diabetes and hypertension is an area in which hospitalists have professional competence. Patient history and use of prescription medications prior to stroke should be reviewed whenever possible and incorporated into short‐term and long‐term treatment plans. Patients with diabetes in particular may benefit more from rigorous control of blood pressure and lipids compared with other patients.16

Secondary stroke prevention should start as early as considered safe. Diagnosis of stroke subtype, often accomplished in the emergency department, establishes suitability for antithrombotics and optimal management strategy. Patients who receive a diagnosis of stroke secondary to cardioembolic atrial fibrillation should be treated with an anticoagulant after the acute period. Aspirin can be used for those individuals unable to use anticoagulants.16 For those individuals with stroke of noncardioembolic origin, particularly those with atherosclerosis and lacunar or cryptogenic infarcts, antiplatelet agents are recommended.14

A multimodal prevention strategy is recommended to manage blood pressure and dyslipidemia poststroke. An algorithm for managing blood pressure soon after stroke has been developed by the PROTECT program (Fig. 2).10 Antihypertensives, usually a combination of an ACE inhibitor and a thiazide diuretic, can be initiated at low doses 48‐72 hours after stroke. A longer delay is recommended for patients with large infarcts or evidence of uncontrolled hypertension. ARBs may be substituted for ACE inhibitors.10 Target blood pressures should be determined using the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.17 In general, even a reduction of 10/5 mm Hg has been shown to be beneficial.16

Figure 2

Statins are recommended for all patients with elevated serum lipids unless treatment with statins is contraindicated. The recommended target level for low‐density lipoprotein cholesterol is below 100 mg/dL for individuals with coronary heart disease and symptomatic atherosclerosis. A target below 70 mg/dL may be appropriate for patients at very high risk.16

Prior to discharge, patients or their caregivers should be given prescriptions adequate to cover the time until postdischarge follow‐up visit. The responsible persons need to be made aware that some medications such as antihypertensives will require dosage adjustments by an outpatient physician, and the timing of the follow‐up visit may need to be arranged accordingly.

The importance of stroke risk reduction should be part of predischarge patient education, along with a list of the warning signs of stroke. Adherence to the treatment regimen, including lifestyle changes and medications, should be emphasized. Patients or their caregivers should be educated about identifying adverse events and a plan to address them. Understanding that some adverse effects (eg, headache with aspirin plus extended‐release dipyridamole) are likely to be transient may prevent unnecessary discontinuation of treatment and reduce anxiety.

Patient and caregiver education can be reinforced by providing standardized patient education materials that can be found in the Stroke Resource Room at the Society of Hospital Medicine Web site (http://www.hospitalmedicine.org),9 PROTECT (http://strokeprotect.mednet.ucla.edu),10 ASA (http://www.strokeassociation.org),8 and AHA (http://www.americanheart.org)18 Web sites.

Transfer of patient information to outpatient health care providers is a critical step in stroke systems care. Notes indicating any need for medication dose adjustment must be included. Discharge summaries should be available to primary care providers, neurologists, and rehabilitation specialists prior to follow‐up visits. The use of electronic forms that can be faxed or sent by E‐mail can shorten delivery time considerably. In lieu of electronic delivery, physician letters can be used, and prototypes are available at the resource Web sites. Whenever possible, a follow‐up phone call to the primary care physician provides the best means to ensure clear communication.

SUMMARY

Hospitalists are well qualified to lead quality focused patient care initiatives at their institutions. Use of standardized protocols to reduce the risk of secondary stroke is proven to increase appropriate prescribing at discharge, which in turn improves patient adherence to evidence‐based therapy. Multidisciplinary communication, including communication with outpatient clinicians, facilitates the transition from inpatient to outpatient health care providers.

In addition to improving patient care, use of standardized protocols is tracked by JCAHO and offers assurance to payers that a particular hospital and its staff are committed to quality care. Establishing protocols is made relatively easy by the online availability of materials that can be adapted to various hospital settings.

The risk of recurrent stroke is high following an ischemic stroke or transient ischemic attack (TIA).16 Within the first 90 days following an initial TIA, between 4.8% and 18.3% of individuals will have an ischemic stroke, with many experiencing an ischemic event within the first 27 days.14 The risk of subsequent stroke in a stroke survivor is high as well4.2% at 6 months, 6.5% at 1 year, and 11.8% at 3 years.5 The management of these patients poses substantial challenges for the health care professional. Prevention of secondary stroke, with its risk for greater morbidity and mortality, is a priority. However, depending on the cause of the event, patient comorbidities, and other factors, the most effective therapeutic strategies may differ. For example, cardioembolic strokes, which constitute approximately 20% of ischemic strokes, are treated with anticoagulants, whereas strokes of noncardioembolic origin are usually treated with antiplatelet agents.7, 8 Other risk factors or variables such as recent stent placement or reduced left ventricular ejection fraction (LVEF) may affect therapeutic decisions as well, although in many cases clear data are not available to direct these difficult decisions. Thus, although antiplatelet agents, including aspirin, clopidogrel, and aspirin plus extended‐release dipyridamole, prevent strokes, the choice of agent depends on the individual patient risk profile. A number of challenging patient scenarios are explored in this article with the goal of providing a context for some of the more recent trial data.

RECENT STENT PLACEMENT

In 2004, there were approximately 663,000 percutaneous coronary interventions (PCIs).9 Stenting after PCI is a common procedure and is used in more than 70% of coronary angioplasty procedures. The addition of stenting to the PCI procedure has improved the outcome for patients, reducing the need for revascularization.10 Because restenosis of the area following stent placement is common, drug‐eluting stents are also used to allow slow release of antiproliferative agents such as sirolimus or paclitaxel.11, 12

Studies such as Percutaneous Coronary InterventionClopidogrel in Unstable Angina to Prevent Recurrent Events (PCI‐CURE) and Clopidogrel for Reduction of Events During Observation (CREDO) have supported the use of up to 8 months of clopidogrel plus aspirin following coronary interventions.13, 14 The European Society of Cardiology PCI guidelines state that in regard to PCI procedures, clopidogrel is superior to aspirin. The guidelines recommend 34 weeks of clopidogrel following stenting in patients with stable angina but up to 12 months in patients receiving brachytherapy. Among patients who have received drug‐eluting stents, clopidogrel therapy should be continued for 612 months. In contrast, aspirin therapy (75100 mg/day) should be continued for life in all these patients.10 In patients who have had a nonST segment elevation myocardial infarction (MI) or who have unstable angina, these guidelines recommend the continuation of clopidogrel (75 mg/day) plus aspirin (100 mg/day) for 912 months after a PCI procedure.10

However, although clopidogrel plus aspirin reduces the incidence of major ischemic events in the period immediately following a stenting procedure, some have suggested that long‐term use of clopidogrel is not supported by the evidence.14 It has been proposed that the sustained beneficial effect of clopidogrel given in the immediate postoperative period may account for much of the long‐term benefit, as has been shown to be true of the glycoprotein IIb/IIIa antagonists.14 However, others caution that in the case of drug‐eluting stents, inhibition of endothelialization of the stent struts by the embedded agents makes these stents more susceptible to thrombosis formation, particularly if therapy with clopidogrel plus aspirin is interrupted.12 It is believed that late stent thrombosis, which has a high mortality rate, is more common with drug‐eluting stents than with bare‐metal stents.12, 15 As a result, many cardiologists recommend at least 12 months of dual antiplatelet therapy with aspirin plus clopidogrel for patients who have received drug‐eluting stents.12 However, given the results of the recent Management of Atherothrombosis in High‐risk Patients with Recent Transient Ischemic Attack or Ischemic Stroke (MATCH) and Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trials,16, 17 in particular, the high incidence of bleeding events in the clopidogrel plus aspirin group, there are concerns about longer‐term or lifelong therapy with this combination in a population at risk for recurrent stroke.

What about the patient who has undergone a coronary stent placement in the past 12 months and experiences a subsequent ischemic stroke or TIA? The patient should be continued on clopidogrel plus aspirin for the recommended time, as premature discontinuation of antiplatelet therapy increases the risk of stent thrombosis.18 No data are currently available to support decision making regarding these patients. However, it has been suggested that among patients given drug‐eluting stents, extended use of clopidogrel at 6, 12, and 24 months is associated with reduced risk of death or death/MI.18

LOW EJECTION FRACTION

Patients who have had a stroke or TIA and have underlying left ventricular dysfunction are at increased risk of a cardioembolic stroke.8 The reduction in stroke volume creates a condition of stasis in the ventricle that increases the likelihood of coagulation and thromboembolic events.8, 19 Evidence indicates that the risk of stroke is inversely correlated with LVEF; LVEF of 29%35% carries a cumulative 5‐year stroke risk of 7.8%, and LVEF of 28% or below carries a 5‐year risk of 8.9%.8, 20, 21 Data from the Survival and Ventricular Enlargement (SAVE) study showed an 18% increase in the risk of stroke for every 5% decline in LVEF,19, 21 and the Studies of Left Ventricular Dysfunction (SOLVD) trial found a 58% increase in thromboembolic events for every 10% decrease in LVEF among women (P = .01).19, 22 Among patients with low LVEF who have had a stroke, the 5‐year recurrent stroke rate may be as high as 45%.19, 23

Although it would appear that stroke associated with left ventricular dysfunction and a low LVEF may potentially be cardioembolic in origin, risk reduction for recurrent stroke has not been adequately investigated as a primary end point in clinical trials, particularly in the absence of atrial fibrillation.24 Thus, the question of whether antiplatelet or anticoagulant therapy would be more effective has not yet been answered. However, results of secondary end point analyses in the SOLVD and SAVE trials suggested that patients had a lower risk of sudden death, thromboembolism, and stroke with antiplatelet therapy.21, 2426 In an observational analysis of prospectively collected data on patients enrolled in the SAVE trial, use of aspirin reduced the overall risk of stroke by 66% in patients with an LVEF below 28%.21 Warfarin is the standard of care for stroke prevention in atrial fibrillation, and the 2 conditions often coexist. In those patients, warfarin is the recommended therapy.24

In patients with sinus rhythm and a low LVEF, the choice is less clear. The results of the Warfarin/Aspirin Study in Heart failure (WASH) failed to establish efficacy or safety for aspirin in preventing all‐cause mortality, nonfatal MI, and nonfatal stroke in patients with heart failure. Patients treated with aspirin were significantly more likely to be hospitalized for cardiovascular events, especially worsening heart failure.27 The trial found no significant difference for the composite end point between the 3 treatment groups: aspirin, warfarin, or no antithrombotic treatment. However, this was a small trial, and the findings were far from definitive, as the study was designed primarily to be a feasibility study to aid in the design of a larger outcomes study.24 Because of the inconsistent results and lack of well‐designed studies regarding the benefit of aspirin or anticoagulation for secondary stroke prevention in patients with LVEF in the absence of atrial fibrillation, further study is needed.

More recently, results were presented from the Warfarin and Antiplatelet Therapy in Heart Failure Trial (WATCH), which randomized patients with heart failure, sinus rhythm, and LVEF of 35% or below to either aspirin 162 mg, warfarin (target international normalized ratio [INR] 2.53.0), or clopidogrel.28, 29 Two major comparisons were plannedwarfarin versus aspirin and aspirin versus clopidogrel.28 Whereas warfarin therapy was open‐label because of the need to check blood levels, antiplatelet therapy was given in a double‐blind manner. After a mean follow‐up of 23 months, no significant differences were found for the primary composite end point of all‐cause mortality, nonfatal MI, and nonfatal stroke, which occurred in 20.5% of those on aspirin, 19.8% on warfarin, and 21.8% on clopidogrel. However, for the secondary end point of stroke, there was a strong trend favoring warfarin over aspirin: stroke occurred in 0.7% of patients taking warfarin versus 2.1% of those taking aspirin (P = .06).24, 29 However, the WATCH investigators concluded that the question of warfarin's value for patients with low LVEF and sinus rhythm remained unresolved.29

In the absence of clear data, the American Heart Association (AHA)/American Stroke Association (ASA) guidelines on stroke prevention in this patient population recommend either warfarin (INR 2.03.0) or antiplatelet therapy, including aspirin (50325 mg/day), aspirin plus extended‐release dipyridamole (200 mg twice daily), or clopidogrel (75 mg/day).8 Patients with coexisting atrial fibrillation should be treated with warfarin, or if unable to tolerate that agent, aspirin 325 mg/day.8

The Warfarin Versus Aspirin for Reduced Cardiac Ejection Fraction (WARCEF) trial may provide more definitive answers on the best approach for reducing the risk of recurrent stroke in patients with low LVEF. The study will compare warfarin (INR 2.53.0) and aspirin (325 mg/day) in the prevention of all‐cause mortality and all strokes (ischemic and hemorrhagic) in patients with an LVEF of 35% or below but no atrial fibrillation.30 The study has a target enrollment of 2860 patients, who are being recruited at 70 North American and 70 European sites, and it will include patients with recent stroke or TIA.28 The results are anxiously anticipated.

INTRACRANIAL STENOSIS

Stroke patients with symptomatic intracranial atherosclerosis have a high risk of recurrent strokein the range of 10% per yearand this accounts for approximately 8% of ischemic strokes.8, 31, 32 Intracranial stenosis appears to be more common in African Americans and Hispanics than in white patients.31

Recurrent stroke prevention in patients with intracranial stenosis was explored in the Warfarin‐Aspirin Symptomatic Intracranial Disease (WASID) study, a multicenter, double‐blind trial. Patients with angiographically verified 50%99% stenosis of a major intracranial artery who had experienced either a stroke or TIA were randomized to either warfarin (target INR 2.03.0) or high‐dose aspirin (1300 mg/day). The primary end point was ischemic stroke, brain hemorrhage, or death from vascular causes other than stroke.33 Mean follow‐up was 1.8 years, and enrollment was stopped after 569 patients had been randomized because of concerns about the safety of warfarin in this patient population.33 The primary end point occurred in 22.1% of those treated with aspirin and 21.8% of those treated with warfarin.33 There were no significant differences between the 2 treatment groups for any of the prespecified secondary end points, including ischemic stroke in any vascular territory and ischemic stroke in the territory of the stenotic intracranial artery.33

The rate of death was significantly higher in the warfarin group (9.7%) than in the aspirin group (4.3%; P = .02). Patients in the warfarin group had higher rates of death from both vascular and nonvascular causes.33 Major hemorrhage was significantly more common in the warfarin group (8.3%) than in the aspirin group (3.2%; P = .01). The investigators concluded that warfarin should not be used as first‐line prevention of recurrent stroke in patients with intracranial stenosis. However, there was a significant association between an INR less than 2 and increased risk of ischemic stroke and major cardiac events (P < .001) as well as a significant increase in major hemorrhages in patients with INRs greater than 3 (P < .001).33

The failure of many patients in the study to remain within the therapeutic INR casts doubt on these results to some extent, although this may actually mirror a common real‐world scenario. Patients were within the therapeutic INR goal only 63% of the time. Furthermore, a nonstandard high dose of aspirin (1300 mg/day) was used, which also may have affected the results.34 Others looking at this data have suggested that aspirin remains an imperfect therapy, with an unacceptably high risk of ischemic stroke and other vascular events, and that anticoagulation may play a role in the period immediately following ischemic stroke or TIA with transition to antiplatelet therapy.34 This would require additional investigation.34

The current AHA/ASA guidelines recommend that for patients with noncardioembolic ischemic stroke or TIA, antiplatelet agents rather than oral anticoagulants be used to reduce the risk of recurrent stroke (class I, level A). Aspirin (50325 mg/day), the combination of aspirin and extended‐release dipyridamole, and clopidogrel are all acceptable options for initial therapy (class IIa, level A).8 The combination of aspirin and extended‐release dipyridamole is suggested instead of aspirin alone (class IIa, level A), and clopidogrel may be considered instead of aspirin alone (class IIb, level B).8 However, data are insufficient at this point to make evidence‐based recommendations between antiplatelet options other than aspirin.8 In patients with significant intracranial stenosis whose symptoms persist despite medical therapy, including antithrombotics, statins, and antihypertensives, endovascular therapy with angioplasty and/or stent placement is an option, but it remains investigational and its value is uncertain.8

CAROTID STENOSIS

Asymptomatic carotid stenosis greater than 50% has been found in 7% of men and 5% of women older than 65 years.35, 36 Among those with asymptomatic carotid stenosis greater than 50%, there is an annual risk of stroke of up to 3.4%.35 In such patients, the benefit of carotid endarterectomy (CEA) is highly dependent on the surgical risk, and if complication rates exceed 3.0%, benefit is eliminated.35 The AHA/ASA guidelines recommend that patients be given treatment for all identifiable risk factors, including statins for dyslipidemia, antihypertensives for hypertension, and aspirin as an antiplatelet agent. In select patients with high‐grade asymptomatic carotid stenosis, CEA performed by a surgeon with a morbidity/mortality rate below 3% is recommended.35 In asymptomatic patients with greater than 70% carotid stenosis, CEA can be an effective therapy. Trial data indicate that the overall 5‐year risk of any stroke or perioperative death is 11.8% for deferred surgery versus 6.4% for immediate endarterectomy (P < .0001).35, 37 Unfortunately, data on the value of stents or angioplasty compared with CEA in this patient population are limited.35

In patients who have had a recent TIA or stroke, carotid stenosis would be considered symptomatic. In these patients, the benefit of CEA is strongly associated with the degree of stenosis. Data from the Carotid Endarterectomy Trialists' Collaboration and North American Symptomatic Carotid Endarterectomy Trial (NASCET) have shown that in patients with stenosis greater than 70%, CEA reduces the absolute 5‐year risk of ischemic stroke by 16.0% (P < .001), whereas in patients with 50%69% stenosis, the 5‐year absolute risk reduction is 4.6% (P = .04). In those with stenosis of 30%49%, there is no effect, and CEA in patients with less than 30% stenosis increases the risk of stroke.38, 39 In patients with 50%69% stenosis, benefit is achieved only if patients at highest risk are selected.40 Recent data have also questioned the typical 4‐ to 6‐week delay before performing a CEA following a nondisabling stroke. Rothwell et al. found that surgery performed within 2 weeks of such a stroke was not associated with increased operative risk.41 Moreover, benefit from CEA fell rapidly within the first few weeks after a TIA or stroke, particularly in women, perhaps reflecting the high risk of recurrent stroke in the period immediately following an initial event.41

Angioplasty or stents have been investigated as alternatives to CEA, but the evidence to date has been disappointing. The Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS) demonstrated preventive efficacy and major risks similar to those found for CEA after 3 years of follow‐up in 504 patients with carotid stenosis.42 However, a more recent study was stopped prematurely after 527 patients had been enrolled because of a higher incidence of disabling stroke or death at 30 days in the stenting cohort (3.4%) compared with the CEA cohort (1.5%). The 30‐day incidence of any stroke or death was 3.9% after CEA and 9.6% after stenting, yielding a relative risk of 2.5 for stenting.43 The Stent‐Protected Angioplasty Versus Carotid Endarterectomy in Symptomatic Patients (SPACE) trial has also failed to find benefit for carotid stenting and/or angioplasty in comparison with CEA.44

The AHA/ASA guidelines recommend CEA in patients with ipsilateral severe (70%99%) stenosis and a recent TIA or ischemic stroke (within 6 months). Surgery should be performed by a surgeon with a perioperative morbidity/mortality rate less than 6%.8 In patients with 50%69% stenosis, the advisability of CEA depends on patient factors such as age, sex, comorbidities, and severity of symptoms. Surgery should be performed within 2 weeks of an ischemic event. In patients with severe stenosis in whom CEA would be difficult to perform, carotid angioplasty or stenting may be recommended if performed by practitioners with a morbidity/mortality rate less than 4%6%.8 The Seventh ACCP Conference also recommends that patients undergoing CEA receive aspirin 81325 mg/day prior to and following the procedure.7

ATHEROSCLEROSIS OF THE AORTIC ARCH

Atherosclerosis of the aortic arch contributes significantly as an independent factor to risk of embolic stroke.7 Such plaques can be detected using transesophageal echocardiography; those that are thicker than 45 mm, exhibit ulceration, or have mobile components place individuals at higher risk for stroke.7, 45 The stroke risk associated with aortic arch plaques greater than 5 mm is as high as 33% per year.7, 46

However, data from large‐scale randomized clinical trials on the efficacy of therapeutic interventions in this condition are lacking. Two small trials found efficacy for warfarin in patients with mobile thrombi in the thoracic aorta. In one, patients given oral anticoagulants had better outcomes than those treated with antiplatelet agents, and in the other, warfarin proved to be more effective than no treatment.47, 48 A retrospective trial that looked at 519 patients treated with warfarin, antiplatelet agents, or statins found there was a protective effect of statins, with an absolute risk reduction in embolic events, including ischemic stroke, TIA, and peripheral embolization of 17%, and a relative risk reduction in embolic events of 59%. The odds ratio for embolic events was 0.39 for statins, 0.77 for antiplatelet agents, and 1.18 for warfarin.49 The French Study of Aortic Plaque in Stroke found no significant difference in risk of events between those treated with warfarin and those treated with aspirin; however, this study was not designed as a therapeutic trial, and few patients received warfarin, casting doubt on this finding.45

Given the paucity of data, suggestions for treatment of patients with an aortic arch atheromata are difficult. Certainly, statin therapy, which would address general atherosclerotic risk reduction, can be initiated. Warfarin appeared to be more effective than antiplatelet agents in several of the studies; however some have expressed concern about the possibility of anticoagulation increasing the risk of cholesterol embolism in these patients.7

SYMPTOMATIC CORONARY ARTERY DISEASE

For patients with a history of ischemic stroke or TIA who have symptomatic CAD, their condition must be managed for both stroke and CAD risks. In patients with stable or unstable angina and a history of stroke or TIA, similar risks must be managed. The acute treatment of ACS or symptomatic CAD cannot be adequately addressed here; however, it may involve a number of therapeutic modalities, including PCI, ‐blocker therapy, glycoprotein IIb/IIIa inhibitors, anticoagulant therapy, angiotensin‐converting enzyme (ACE) inhibitors, and clopidogrel plus aspirin, depending on the exact nature of the syndrome.5054 The long‐term management and, in particular, prevention of recurrent stroke in the setting of symptomatic CAD are the focus here. As with a patient with a history of CAD and a recent TIA or stroke (as discussed earlier), patients with symptomatic CAD and TIA or stroke must be managed for multiple risk factors. NCEP guidelines recommend aggressive cholesterol lowering with statin therapy. Hypertension must be addressed as well, and long‐term therapy with ‐blockers and ACE inhibitors has been shown to reduce mortality in patients with ACS and is recommended by the AHA/ASA.5355

Once the acute ACS period has resolved, it is reasonable to address the question of the best possible antiplatelet therapy for long‐term stroke prevention. Long‐term use of clopidogrel plus aspirin is not advisable given the increased risk of bleeding events noted in the MATCH and CHARISMA trials.16, 17 At this point, it would be reasonable to start the patient on aspirin 75150 mg/day, which reduces risk of stroke up to 25%,56, 57 aspirin plus extended‐release dipyridamole, which reduces risk by about 37%,57, 58 or clopidogrel 75 mg/day, which reduces the relative risk for stroke alone by 7.3% compared with aspirin.59 In patients who cannot tolerate or are allergic to aspirin, clopidogrel is a reasonable choice.8

ANTIPLATELET FAILURE

Patients who have failed antiplatelet therapythat is, have gone on to have a recurrent strokeare particularly difficult. It is important to remember that any therapeutic intervention only reduces stroke risk; it does not eliminate it. Keeping that in mind, it is essential to reevaluate and reconsider both the original diagnosis and the etiology of the stroke or TIA. A number of diagnostic alternatives should be considered, including sensory seizure and migraine equivalents, as well as other etiologies, such as atrial fibrillation or cerebral amyloid angiopathy. Therapy may have to be adjusted accordingly, but the patient remains at increased risk for stroke recurrence, and thus preventive therapy is critical.

Several key points should be remembered. As outlined previously in this article, if the stroke is still thought to be noncardioembolic in origin, a reduction in the risk of stroke has not been found for those patients receiving warfarin, an increased dose of aspirin, a combination of antiplatelet agents and warfarin, or clopidogrel plus aspirin.8, 16, 31, 60, 61 However, if atrial fibrillation has developed in the patient, the recommendation is warfarin (INR 2.03.0) or, if anticoagulants cannot be taken, aspirin 325 mg/day.8 Risk factors should be reassessed and managed, with agents and lifestyle changes to control hypertension and dyslipidemia. Antiplatelet agents should be continued in patients with noncardioembolic stroke. Acceptable antiplatelet agents include aspirin (50325 mg/day), aspirin plus extended‐release dipyridamole, and clopidogrel. The combination of aspirin plus extended‐release dipyridamole is suggested over aspirin alone. If the patient cannot tolerate or is allergic to aspirin, clopidogrel is a reasonable alternative.8 The decision of which antiplatelet agent to use should be based on the individual patient's risk factor profile.8 The temptation to put patients on anticoagulation therapy because of a wish to do more should be avoided, as this is likely to expose patients to increased risk without known benefit.60, 61

Consider a common case scenarioa patient with a known history of hypertension and TIA presents with a 30‐minute episode of left arm numbness. The patient has been adherent to his prescribed medications, including aspirin 81 mg/day. What is the appropriate approach to acute treatment at this time? This is a common scenario in emergency departmentsnew‐onset TIA while taking aspirin 81 mg/day. There are advocates for several different treatment regimens in these patients: increasing the aspirin dose to 325 mg/day as a new treatment; discontinuing aspirin and initiating clopidogrel 75 mg/day; discontinuing aspirin 81 mg/day and initiating aspirin 325 mg/day plus clopidogrel 75 mg/day; or discontinuing aspirin 81 mg/day and initiating a combination of aspirin 25 mg plus extended‐release dipyridamole 200 mg twice daily. It is clear that patients with the same disease are treated differently in different institutions. What is the appropriate evidence‐based treatment in this case? The answer is clearno evidence supports increasing the dose of aspirin as a new treatment for this case or initiating aspirin 325 mg/day plus clopidogrel 75 mg/day.16, 17 Based on the literature, for a patient who has recently had another cerebral ischemic event while on treatment, it would make sense to consider switching to another agent. Three agents are recommended by the guidelines: aspirin, clopidogrel, and aspirin plus extended‐release dipyridamole. If treatment 1 were to fail, it would not be against the evidence to initiate treatment 2 or 3.

PATIENTS ON WARFARIN

Data from the Warfarin‐Aspirin Recurrent Stroke Study (WARSS), a large‐scale recurrent stroke prevention trial conducted in 2206 patients, demonstrated that there was no survival benefit for noncardioembolic stroke survivors who were treated with warfarin.60, 61 Yet there are patients still taking warfarin to reduce stroke risk who do not have atrial fibrillation. Unless a patient is allergic to or intolerant of antiplatelet agents such as aspirin, clopidogrel, or dipyridamole, they should not be treated with warfarin for noncardioembolic stroke risk.8 The results of other studies of anticoagulation in recurrent stroke prevention, including the European/Australasian Stroke Prevention in Reversible Ischaemia Trial (ESPRIT),62 the Stroke Performance for Reporting the Improvement and Translation (SPIRIT) trial,63 and the WASID study,33 have yet to demonstrate a role for warfarin in prevention of noncardioembolic stroke.

Given these trial results, patients currently on warfarin who do not have a cardioembolic risk factor should be placed on antiplatelet therapy with aspirin, aspirin plus extended‐release dipyridamole, or clopidogrel 35 days after discontinuing warfarin therapy. However, it would be advisable to evaluate these patients for atrial fibrillation, as patients with that risk factor should remain on warfarin.8

SUMMARY

In clinical practice, health care providers often must manage patients with complex profiles. Multiple risk factors and comorbidities complicate treatment of these individuals, and robust clinical data are often lacking as clinical trials rarely include such individuals. Guidelines offer recommendations, but these too are often based on extrapolations from clinical trial data. This is particularly true of patients at risk for ischemic stroke, as the primary underlying causevascular diseasehas systemic implications and comorbidities that often complicate treatment.

In general, antiplatelet therapy should be used to prevent recurrent stroke in patients with TIA or noncardioembolic stroke, whereas anticoagulation therapy should be used in patients with cardioembolic stroke such as that caused by atrial fibrillation. However, therapy must be individualized to account for the patient's full risk profile. Conditions such as dyslipidemia and hypertension must be addressed as well, as these not only give rise to stroke but also to the CAD, coronary heart disease, and ACS that may coexist with stroke. Among patients deemed suitable for antiplatelet therapy, class IIa, level A evidence supports the use of aspirin 50325 mg/day, the combination of aspirin and extended‐release dipyridamole, and clopidogrel for secondary prevention of stroke.8

The risk of recurrent stroke is high following an ischemic stroke or transient ischemic attack (TIA).16 Within the first 90 days following an initial TIA, between 4.8% and 18.3% of individuals will have an ischemic stroke, with many experiencing an ischemic event within the first 27 days.14 The risk of subsequent stroke in a stroke survivor is high as well4.2% at 6 months, 6.5% at 1 year, and 11.8% at 3 years.5 The management of these patients poses substantial challenges for the health care professional. Prevention of secondary stroke, with its risk for greater morbidity and mortality, is a priority. However, depending on the cause of the event, patient comorbidities, and other factors, the most effective therapeutic strategies may differ. For example, cardioembolic strokes, which constitute approximately 20% of ischemic strokes, are treated with anticoagulants, whereas strokes of noncardioembolic origin are usually treated with antiplatelet agents.7, 8 Other risk factors or variables such as recent stent placement or reduced left ventricular ejection fraction (LVEF) may affect therapeutic decisions as well, although in many cases clear data are not available to direct these difficult decisions. Thus, although antiplatelet agents, including aspirin, clopidogrel, and aspirin plus extended‐release dipyridamole, prevent strokes, the choice of agent depends on the individual patient risk profile. A number of challenging patient scenarios are explored in this article with the goal of providing a context for some of the more recent trial data.

RECENT STENT PLACEMENT

In 2004, there were approximately 663,000 percutaneous coronary interventions (PCIs).9 Stenting after PCI is a common procedure and is used in more than 70% of coronary angioplasty procedures. The addition of stenting to the PCI procedure has improved the outcome for patients, reducing the need for revascularization.10 Because restenosis of the area following stent placement is common, drug‐eluting stents are also used to allow slow release of antiproliferative agents such as sirolimus or paclitaxel.11, 12

Studies such as Percutaneous Coronary InterventionClopidogrel in Unstable Angina to Prevent Recurrent Events (PCI‐CURE) and Clopidogrel for Reduction of Events During Observation (CREDO) have supported the use of up to 8 months of clopidogrel plus aspirin following coronary interventions.13, 14 The European Society of Cardiology PCI guidelines state that in regard to PCI procedures, clopidogrel is superior to aspirin. The guidelines recommend 34 weeks of clopidogrel following stenting in patients with stable angina but up to 12 months in patients receiving brachytherapy. Among patients who have received drug‐eluting stents, clopidogrel therapy should be continued for 612 months. In contrast, aspirin therapy (75100 mg/day) should be continued for life in all these patients.10 In patients who have had a nonST segment elevation myocardial infarction (MI) or who have unstable angina, these guidelines recommend the continuation of clopidogrel (75 mg/day) plus aspirin (100 mg/day) for 912 months after a PCI procedure.10

However, although clopidogrel plus aspirin reduces the incidence of major ischemic events in the period immediately following a stenting procedure, some have suggested that long‐term use of clopidogrel is not supported by the evidence.14 It has been proposed that the sustained beneficial effect of clopidogrel given in the immediate postoperative period may account for much of the long‐term benefit, as has been shown to be true of the glycoprotein IIb/IIIa antagonists.14 However, others caution that in the case of drug‐eluting stents, inhibition of endothelialization of the stent struts by the embedded agents makes these stents more susceptible to thrombosis formation, particularly if therapy with clopidogrel plus aspirin is interrupted.12 It is believed that late stent thrombosis, which has a high mortality rate, is more common with drug‐eluting stents than with bare‐metal stents.12, 15 As a result, many cardiologists recommend at least 12 months of dual antiplatelet therapy with aspirin plus clopidogrel for patients who have received drug‐eluting stents.12 However, given the results of the recent Management of Atherothrombosis in High‐risk Patients with Recent Transient Ischemic Attack or Ischemic Stroke (MATCH) and Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trials,16, 17 in particular, the high incidence of bleeding events in the clopidogrel plus aspirin group, there are concerns about longer‐term or lifelong therapy with this combination in a population at risk for recurrent stroke.

What about the patient who has undergone a coronary stent placement in the past 12 months and experiences a subsequent ischemic stroke or TIA? The patient should be continued on clopidogrel plus aspirin for the recommended time, as premature discontinuation of antiplatelet therapy increases the risk of stent thrombosis.18 No data are currently available to support decision making regarding these patients. However, it has been suggested that among patients given drug‐eluting stents, extended use of clopidogrel at 6, 12, and 24 months is associated with reduced risk of death or death/MI.18

LOW EJECTION FRACTION

Patients who have had a stroke or TIA and have underlying left ventricular dysfunction are at increased risk of a cardioembolic stroke.8 The reduction in stroke volume creates a condition of stasis in the ventricle that increases the likelihood of coagulation and thromboembolic events.8, 19 Evidence indicates that the risk of stroke is inversely correlated with LVEF; LVEF of 29%35% carries a cumulative 5‐year stroke risk of 7.8%, and LVEF of 28% or below carries a 5‐year risk of 8.9%.8, 20, 21 Data from the Survival and Ventricular Enlargement (SAVE) study showed an 18% increase in the risk of stroke for every 5% decline in LVEF,19, 21 and the Studies of Left Ventricular Dysfunction (SOLVD) trial found a 58% increase in thromboembolic events for every 10% decrease in LVEF among women (P = .01).19, 22 Among patients with low LVEF who have had a stroke, the 5‐year recurrent stroke rate may be as high as 45%.19, 23

Although it would appear that stroke associated with left ventricular dysfunction and a low LVEF may potentially be cardioembolic in origin, risk reduction for recurrent stroke has not been adequately investigated as a primary end point in clinical trials, particularly in the absence of atrial fibrillation.24 Thus, the question of whether antiplatelet or anticoagulant therapy would be more effective has not yet been answered. However, results of secondary end point analyses in the SOLVD and SAVE trials suggested that patients had a lower risk of sudden death, thromboembolism, and stroke with antiplatelet therapy.21, 2426 In an observational analysis of prospectively collected data on patients enrolled in the SAVE trial, use of aspirin reduced the overall risk of stroke by 66% in patients with an LVEF below 28%.21 Warfarin is the standard of care for stroke prevention in atrial fibrillation, and the 2 conditions often coexist. In those patients, warfarin is the recommended therapy.24

In patients with sinus rhythm and a low LVEF, the choice is less clear. The results of the Warfarin/Aspirin Study in Heart failure (WASH) failed to establish efficacy or safety for aspirin in preventing all‐cause mortality, nonfatal MI, and nonfatal stroke in patients with heart failure. Patients treated with aspirin were significantly more likely to be hospitalized for cardiovascular events, especially worsening heart failure.27 The trial found no significant difference for the composite end point between the 3 treatment groups: aspirin, warfarin, or no antithrombotic treatment. However, this was a small trial, and the findings were far from definitive, as the study was designed primarily to be a feasibility study to aid in the design of a larger outcomes study.24 Because of the inconsistent results and lack of well‐designed studies regarding the benefit of aspirin or anticoagulation for secondary stroke prevention in patients with LVEF in the absence of atrial fibrillation, further study is needed.

More recently, results were presented from the Warfarin and Antiplatelet Therapy in Heart Failure Trial (WATCH), which randomized patients with heart failure, sinus rhythm, and LVEF of 35% or below to either aspirin 162 mg, warfarin (target international normalized ratio [INR] 2.53.0), or clopidogrel.28, 29 Two major comparisons were plannedwarfarin versus aspirin and aspirin versus clopidogrel.28 Whereas warfarin therapy was open‐label because of the need to check blood levels, antiplatelet therapy was given in a double‐blind manner. After a mean follow‐up of 23 months, no significant differences were found for the primary composite end point of all‐cause mortality, nonfatal MI, and nonfatal stroke, which occurred in 20.5% of those on aspirin, 19.8% on warfarin, and 21.8% on clopidogrel. However, for the secondary end point of stroke, there was a strong trend favoring warfarin over aspirin: stroke occurred in 0.7% of patients taking warfarin versus 2.1% of those taking aspirin (P = .06).24, 29 However, the WATCH investigators concluded that the question of warfarin's value for patients with low LVEF and sinus rhythm remained unresolved.29

In the absence of clear data, the American Heart Association (AHA)/American Stroke Association (ASA) guidelines on stroke prevention in this patient population recommend either warfarin (INR 2.03.0) or antiplatelet therapy, including aspirin (50325 mg/day), aspirin plus extended‐release dipyridamole (200 mg twice daily), or clopidogrel (75 mg/day).8 Patients with coexisting atrial fibrillation should be treated with warfarin, or if unable to tolerate that agent, aspirin 325 mg/day.8

The Warfarin Versus Aspirin for Reduced Cardiac Ejection Fraction (WARCEF) trial may provide more definitive answers on the best approach for reducing the risk of recurrent stroke in patients with low LVEF. The study will compare warfarin (INR 2.53.0) and aspirin (325 mg/day) in the prevention of all‐cause mortality and all strokes (ischemic and hemorrhagic) in patients with an LVEF of 35% or below but no atrial fibrillation.30 The study has a target enrollment of 2860 patients, who are being recruited at 70 North American and 70 European sites, and it will include patients with recent stroke or TIA.28 The results are anxiously anticipated.

INTRACRANIAL STENOSIS

Stroke patients with symptomatic intracranial atherosclerosis have a high risk of recurrent strokein the range of 10% per yearand this accounts for approximately 8% of ischemic strokes.8, 31, 32 Intracranial stenosis appears to be more common in African Americans and Hispanics than in white patients.31

Recurrent stroke prevention in patients with intracranial stenosis was explored in the Warfarin‐Aspirin Symptomatic Intracranial Disease (WASID) study, a multicenter, double‐blind trial. Patients with angiographically verified 50%99% stenosis of a major intracranial artery who had experienced either a stroke or TIA were randomized to either warfarin (target INR 2.03.0) or high‐dose aspirin (1300 mg/day). The primary end point was ischemic stroke, brain hemorrhage, or death from vascular causes other than stroke.33 Mean follow‐up was 1.8 years, and enrollment was stopped after 569 patients had been randomized because of concerns about the safety of warfarin in this patient population.33 The primary end point occurred in 22.1% of those treated with aspirin and 21.8% of those treated with warfarin.33 There were no significant differences between the 2 treatment groups for any of the prespecified secondary end points, including ischemic stroke in any vascular territory and ischemic stroke in the territory of the stenotic intracranial artery.33

The rate of death was significantly higher in the warfarin group (9.7%) than in the aspirin group (4.3%; P = .02). Patients in the warfarin group had higher rates of death from both vascular and nonvascular causes.33 Major hemorrhage was significantly more common in the warfarin group (8.3%) than in the aspirin group (3.2%; P = .01). The investigators concluded that warfarin should not be used as first‐line prevention of recurrent stroke in patients with intracranial stenosis. However, there was a significant association between an INR less than 2 and increased risk of ischemic stroke and major cardiac events (P < .001) as well as a significant increase in major hemorrhages in patients with INRs greater than 3 (P < .001).33

The failure of many patients in the study to remain within the therapeutic INR casts doubt on these results to some extent, although this may actually mirror a common real‐world scenario. Patients were within the therapeutic INR goal only 63% of the time. Furthermore, a nonstandard high dose of aspirin (1300 mg/day) was used, which also may have affected the results.34 Others looking at this data have suggested that aspirin remains an imperfect therapy, with an unacceptably high risk of ischemic stroke and other vascular events, and that anticoagulation may play a role in the period immediately following ischemic stroke or TIA with transition to antiplatelet therapy.34 This would require additional investigation.34

The current AHA/ASA guidelines recommend that for patients with noncardioembolic ischemic stroke or TIA, antiplatelet agents rather than oral anticoagulants be used to reduce the risk of recurrent stroke (class I, level A). Aspirin (50325 mg/day), the combination of aspirin and extended‐release dipyridamole, and clopidogrel are all acceptable options for initial therapy (class IIa, level A).8 The combination of aspirin and extended‐release dipyridamole is suggested instead of aspirin alone (class IIa, level A), and clopidogrel may be considered instead of aspirin alone (class IIb, level B).8 However, data are insufficient at this point to make evidence‐based recommendations between antiplatelet options other than aspirin.8 In patients with significant intracranial stenosis whose symptoms persist despite medical therapy, including antithrombotics, statins, and antihypertensives, endovascular therapy with angioplasty and/or stent placement is an option, but it remains investigational and its value is uncertain.8

CAROTID STENOSIS

Asymptomatic carotid stenosis greater than 50% has been found in 7% of men and 5% of women older than 65 years.35, 36 Among those with asymptomatic carotid stenosis greater than 50%, there is an annual risk of stroke of up to 3.4%.35 In such patients, the benefit of carotid endarterectomy (CEA) is highly dependent on the surgical risk, and if complication rates exceed 3.0%, benefit is eliminated.35 The AHA/ASA guidelines recommend that patients be given treatment for all identifiable risk factors, including statins for dyslipidemia, antihypertensives for hypertension, and aspirin as an antiplatelet agent. In select patients with high‐grade asymptomatic carotid stenosis, CEA performed by a surgeon with a morbidity/mortality rate below 3% is recommended.35 In asymptomatic patients with greater than 70% carotid stenosis, CEA can be an effective therapy. Trial data indicate that the overall 5‐year risk of any stroke or perioperative death is 11.8% for deferred surgery versus 6.4% for immediate endarterectomy (P < .0001).35, 37 Unfortunately, data on the value of stents or angioplasty compared with CEA in this patient population are limited.35

In patients who have had a recent TIA or stroke, carotid stenosis would be considered symptomatic. In these patients, the benefit of CEA is strongly associated with the degree of stenosis. Data from the Carotid Endarterectomy Trialists' Collaboration and North American Symptomatic Carotid Endarterectomy Trial (NASCET) have shown that in patients with stenosis greater than 70%, CEA reduces the absolute 5‐year risk of ischemic stroke by 16.0% (P < .001), whereas in patients with 50%69% stenosis, the 5‐year absolute risk reduction is 4.6% (P = .04). In those with stenosis of 30%49%, there is no effect, and CEA in patients with less than 30% stenosis increases the risk of stroke.38, 39 In patients with 50%69% stenosis, benefit is achieved only if patients at highest risk are selected.40 Recent data have also questioned the typical 4‐ to 6‐week delay before performing a CEA following a nondisabling stroke. Rothwell et al. found that surgery performed within 2 weeks of such a stroke was not associated with increased operative risk.41 Moreover, benefit from CEA fell rapidly within the first few weeks after a TIA or stroke, particularly in women, perhaps reflecting the high risk of recurrent stroke in the period immediately following an initial event.41

Angioplasty or stents have been investigated as alternatives to CEA, but the evidence to date has been disappointing. The Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS) demonstrated preventive efficacy and major risks similar to those found for CEA after 3 years of follow‐up in 504 patients with carotid stenosis.42 However, a more recent study was stopped prematurely after 527 patients had been enrolled because of a higher incidence of disabling stroke or death at 30 days in the stenting cohort (3.4%) compared with the CEA cohort (1.5%). The 30‐day incidence of any stroke or death was 3.9% after CEA and 9.6% after stenting, yielding a relative risk of 2.5 for stenting.43 The Stent‐Protected Angioplasty Versus Carotid Endarterectomy in Symptomatic Patients (SPACE) trial has also failed to find benefit for carotid stenting and/or angioplasty in comparison with CEA.44

The AHA/ASA guidelines recommend CEA in patients with ipsilateral severe (70%99%) stenosis and a recent TIA or ischemic stroke (within 6 months). Surgery should be performed by a surgeon with a perioperative morbidity/mortality rate less than 6%.8 In patients with 50%69% stenosis, the advisability of CEA depends on patient factors such as age, sex, comorbidities, and severity of symptoms. Surgery should be performed within 2 weeks of an ischemic event. In patients with severe stenosis in whom CEA would be difficult to perform, carotid angioplasty or stenting may be recommended if performed by practitioners with a morbidity/mortality rate less than 4%6%.8 The Seventh ACCP Conference also recommends that patients undergoing CEA receive aspirin 81325 mg/day prior to and following the procedure.7

ATHEROSCLEROSIS OF THE AORTIC ARCH

Atherosclerosis of the aortic arch contributes significantly as an independent factor to risk of embolic stroke.7 Such plaques can be detected using transesophageal echocardiography; those that are thicker than 45 mm, exhibit ulceration, or have mobile components place individuals at higher risk for stroke.7, 45 The stroke risk associated with aortic arch plaques greater than 5 mm is as high as 33% per year.7, 46

However, data from large‐scale randomized clinical trials on the efficacy of therapeutic interventions in this condition are lacking. Two small trials found efficacy for warfarin in patients with mobile thrombi in the thoracic aorta. In one, patients given oral anticoagulants had better outcomes than those treated with antiplatelet agents, and in the other, warfarin proved to be more effective than no treatment.47, 48 A retrospective trial that looked at 519 patients treated with warfarin, antiplatelet agents, or statins found there was a protective effect of statins, with an absolute risk reduction in embolic events, including ischemic stroke, TIA, and peripheral embolization of 17%, and a relative risk reduction in embolic events of 59%. The odds ratio for embolic events was 0.39 for statins, 0.77 for antiplatelet agents, and 1.18 for warfarin.49 The French Study of Aortic Plaque in Stroke found no significant difference in risk of events between those treated with warfarin and those treated with aspirin; however, this study was not designed as a therapeutic trial, and few patients received warfarin, casting doubt on this finding.45

Given the paucity of data, suggestions for treatment of patients with an aortic arch atheromata are difficult. Certainly, statin therapy, which would address general atherosclerotic risk reduction, can be initiated. Warfarin appeared to be more effective than antiplatelet agents in several of the studies; however some have expressed concern about the possibility of anticoagulation increasing the risk of cholesterol embolism in these patients.7

SYMPTOMATIC CORONARY ARTERY DISEASE

For patients with a history of ischemic stroke or TIA who have symptomatic CAD, their condition must be managed for both stroke and CAD risks. In patients with stable or unstable angina and a history of stroke or TIA, similar risks must be managed. The acute treatment of ACS or symptomatic CAD cannot be adequately addressed here; however, it may involve a number of therapeutic modalities, including PCI, ‐blocker therapy, glycoprotein IIb/IIIa inhibitors, anticoagulant therapy, angiotensin‐converting enzyme (ACE) inhibitors, and clopidogrel plus aspirin, depending on the exact nature of the syndrome.5054 The long‐term management and, in particular, prevention of recurrent stroke in the setting of symptomatic CAD are the focus here. As with a patient with a history of CAD and a recent TIA or stroke (as discussed earlier), patients with symptomatic CAD and TIA or stroke must be managed for multiple risk factors. NCEP guidelines recommend aggressive cholesterol lowering with statin therapy. Hypertension must be addressed as well, and long‐term therapy with ‐blockers and ACE inhibitors has been shown to reduce mortality in patients with ACS and is recommended by the AHA/ASA.5355

Once the acute ACS period has resolved, it is reasonable to address the question of the best possible antiplatelet therapy for long‐term stroke prevention. Long‐term use of clopidogrel plus aspirin is not advisable given the increased risk of bleeding events noted in the MATCH and CHARISMA trials.16, 17 At this point, it would be reasonable to start the patient on aspirin 75150 mg/day, which reduces risk of stroke up to 25%,56, 57 aspirin plus extended‐release dipyridamole, which reduces risk by about 37%,57, 58 or clopidogrel 75 mg/day, which reduces the relative risk for stroke alone by 7.3% compared with aspirin.59 In patients who cannot tolerate or are allergic to aspirin, clopidogrel is a reasonable choice.8

ANTIPLATELET FAILURE

Patients who have failed antiplatelet therapythat is, have gone on to have a recurrent strokeare particularly difficult. It is important to remember that any therapeutic intervention only reduces stroke risk; it does not eliminate it. Keeping that in mind, it is essential to reevaluate and reconsider both the original diagnosis and the etiology of the stroke or TIA. A number of diagnostic alternatives should be considered, including sensory seizure and migraine equivalents, as well as other etiologies, such as atrial fibrillation or cerebral amyloid angiopathy. Therapy may have to be adjusted accordingly, but the patient remains at increased risk for stroke recurrence, and thus preventive therapy is critical.

Several key points should be remembered. As outlined previously in this article, if the stroke is still thought to be noncardioembolic in origin, a reduction in the risk of stroke has not been found for those patients receiving warfarin, an increased dose of aspirin, a combination of antiplatelet agents and warfarin, or clopidogrel plus aspirin.8, 16, 31, 60, 61 However, if atrial fibrillation has developed in the patient, the recommendation is warfarin (INR 2.03.0) or, if anticoagulants cannot be taken, aspirin 325 mg/day.8 Risk factors should be reassessed and managed, with agents and lifestyle changes to control hypertension and dyslipidemia. Antiplatelet agents should be continued in patients with noncardioembolic stroke. Acceptable antiplatelet agents include aspirin (50325 mg/day), aspirin plus extended‐release dipyridamole, and clopidogrel. The combination of aspirin plus extended‐release dipyridamole is suggested over aspirin alone. If the patient cannot tolerate or is allergic to aspirin, clopidogrel is a reasonable alternative.8 The decision of which antiplatelet agent to use should be based on the individual patient's risk factor profile.8 The temptation to put patients on anticoagulation therapy because of a wish to do more should be avoided, as this is likely to expose patients to increased risk without known benefit.60, 61

Consider a common case scenarioa patient with a known history of hypertension and TIA presents with a 30‐minute episode of left arm numbness. The patient has been adherent to his prescribed medications, including aspirin 81 mg/day. What is the appropriate approach to acute treatment at this time? This is a common scenario in emergency departmentsnew‐onset TIA while taking aspirin 81 mg/day. There are advocates for several different treatment regimens in these patients: increasing the aspirin dose to 325 mg/day as a new treatment; discontinuing aspirin and initiating clopidogrel 75 mg/day; discontinuing aspirin 81 mg/day and initiating aspirin 325 mg/day plus clopidogrel 75 mg/day; or discontinuing aspirin 81 mg/day and initiating a combination of aspirin 25 mg plus extended‐release dipyridamole 200 mg twice daily. It is clear that patients with the same disease are treated differently in different institutions. What is the appropriate evidence‐based treatment in this case? The answer is clearno evidence supports increasing the dose of aspirin as a new treatment for this case or initiating aspirin 325 mg/day plus clopidogrel 75 mg/day.16, 17 Based on the literature, for a patient who has recently had another cerebral ischemic event while on treatment, it would make sense to consider switching to another agent. Three agents are recommended by the guidelines: aspirin, clopidogrel, and aspirin plus extended‐release dipyridamole. If treatment 1 were to fail, it would not be against the evidence to initiate treatment 2 or 3.

PATIENTS ON WARFARIN

Data from the Warfarin‐Aspirin Recurrent Stroke Study (WARSS), a large‐scale recurrent stroke prevention trial conducted in 2206 patients, demonstrated that there was no survival benefit for noncardioembolic stroke survivors who were treated with warfarin.60, 61 Yet there are patients still taking warfarin to reduce stroke risk who do not have atrial fibrillation. Unless a patient is allergic to or intolerant of antiplatelet agents such as aspirin, clopidogrel, or dipyridamole, they should not be treated with warfarin for noncardioembolic stroke risk.8 The results of other studies of anticoagulation in recurrent stroke prevention, including the European/Australasian Stroke Prevention in Reversible Ischaemia Trial (ESPRIT),62 the Stroke Performance for Reporting the Improvement and Translation (SPIRIT) trial,63 and the WASID study,33 have yet to demonstrate a role for warfarin in prevention of noncardioembolic stroke.

Given these trial results, patients currently on warfarin who do not have a cardioembolic risk factor should be placed on antiplatelet therapy with aspirin, aspirin plus extended‐release dipyridamole, or clopidogrel 35 days after discontinuing warfarin therapy. However, it would be advisable to evaluate these patients for atrial fibrillation, as patients with that risk factor should remain on warfarin.8

SUMMARY

In clinical practice, health care providers often must manage patients with complex profiles. Multiple risk factors and comorbidities complicate treatment of these individuals, and robust clinical data are often lacking as clinical trials rarely include such individuals. Guidelines offer recommendations, but these too are often based on extrapolations from clinical trial data. This is particularly true of patients at risk for ischemic stroke, as the primary underlying causevascular diseasehas systemic implications and comorbidities that often complicate treatment.

In general, antiplatelet therapy should be used to prevent recurrent stroke in patients with TIA or noncardioembolic stroke, whereas anticoagulation therapy should be used in patients with cardioembolic stroke such as that caused by atrial fibrillation. However, therapy must be individualized to account for the patient's full risk profile. Conditions such as dyslipidemia and hypertension must be addressed as well, as these not only give rise to stroke but also to the CAD, coronary heart disease, and ACS that may coexist with stroke. Among patients deemed suitable for antiplatelet therapy, class IIa, level A evidence supports the use of aspirin 50325 mg/day, the combination of aspirin and extended‐release dipyridamole, and clopidogrel for secondary prevention of stroke.8

The risk of recurrent stroke is high following an ischemic stroke or transient ischemic attack (TIA).16 Within the first 90 days following an initial TIA, between 4.8% and 18.3% of individuals will have an ischemic stroke, with many experiencing an ischemic event within the first 27 days.14 The risk of subsequent stroke in a stroke survivor is high as well4.2% at 6 months, 6.5% at 1 year, and 11.8% at 3 years.5 The management of these patients poses substantial challenges for the health care professional. Prevention of secondary stroke, with its risk for greater morbidity and mortality, is a priority. However, depending on the cause of the event, patient comorbidities, and other factors, the most effective therapeutic strategies may differ. For example, cardioembolic strokes, which constitute approximately 20% of ischemic strokes, are treated with anticoagulants, whereas strokes of noncardioembolic origin are usually treated with antiplatelet agents.7, 8 Other risk factors or variables such as recent stent placement or reduced left ventricular ejection fraction (LVEF) may affect therapeutic decisions as well, although in many cases clear data are not available to direct these difficult decisions. Thus, although antiplatelet agents, including aspirin, clopidogrel, and aspirin plus extended‐release dipyridamole, prevent strokes, the choice of agent depends on the individual patient risk profile. A number of challenging patient scenarios are explored in this article with the goal of providing a context for some of the more recent trial data.

RECENT STENT PLACEMENT

In 2004, there were approximately 663,000 percutaneous coronary interventions (PCIs).9 Stenting after PCI is a common procedure and is used in more than 70% of coronary angioplasty procedures. The addition of stenting to the PCI procedure has improved the outcome for patients, reducing the need for revascularization.10 Because restenosis of the area following stent placement is common, drug‐eluting stents are also used to allow slow release of antiproliferative agents such as sirolimus or paclitaxel.11, 12

Studies such as Percutaneous Coronary InterventionClopidogrel in Unstable Angina to Prevent Recurrent Events (PCI‐CURE) and Clopidogrel for Reduction of Events During Observation (CREDO) have supported the use of up to 8 months of clopidogrel plus aspirin following coronary interventions.13, 14 The European Society of Cardiology PCI guidelines state that in regard to PCI procedures, clopidogrel is superior to aspirin. The guidelines recommend 34 weeks of clopidogrel following stenting in patients with stable angina but up to 12 months in patients receiving brachytherapy. Among patients who have received drug‐eluting stents, clopidogrel therapy should be continued for 612 months. In contrast, aspirin therapy (75100 mg/day) should be continued for life in all these patients.10 In patients who have had a nonST segment elevation myocardial infarction (MI) or who have unstable angina, these guidelines recommend the continuation of clopidogrel (75 mg/day) plus aspirin (100 mg/day) for 912 months after a PCI procedure.10

However, although clopidogrel plus aspirin reduces the incidence of major ischemic events in the period immediately following a stenting procedure, some have suggested that long‐term use of clopidogrel is not supported by the evidence.14 It has been proposed that the sustained beneficial effect of clopidogrel given in the immediate postoperative period may account for much of the long‐term benefit, as has been shown to be true of the glycoprotein IIb/IIIa antagonists.14 However, others caution that in the case of drug‐eluting stents, inhibition of endothelialization of the stent struts by the embedded agents makes these stents more susceptible to thrombosis formation, particularly if therapy with clopidogrel plus aspirin is interrupted.12 It is believed that late stent thrombosis, which has a high mortality rate, is more common with drug‐eluting stents than with bare‐metal stents.12, 15 As a result, many cardiologists recommend at least 12 months of dual antiplatelet therapy with aspirin plus clopidogrel for patients who have received drug‐eluting stents.12 However, given the results of the recent Management of Atherothrombosis in High‐risk Patients with Recent Transient Ischemic Attack or Ischemic Stroke (MATCH) and Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trials,16, 17 in particular, the high incidence of bleeding events in the clopidogrel plus aspirin group, there are concerns about longer‐term or lifelong therapy with this combination in a population at risk for recurrent stroke.

What about the patient who has undergone a coronary stent placement in the past 12 months and experiences a subsequent ischemic stroke or TIA? The patient should be continued on clopidogrel plus aspirin for the recommended time, as premature discontinuation of antiplatelet therapy increases the risk of stent thrombosis.18 No data are currently available to support decision making regarding these patients. However, it has been suggested that among patients given drug‐eluting stents, extended use of clopidogrel at 6, 12, and 24 months is associated with reduced risk of death or death/MI.18

LOW EJECTION FRACTION

Patients who have had a stroke or TIA and have underlying left ventricular dysfunction are at increased risk of a cardioembolic stroke.8 The reduction in stroke volume creates a condition of stasis in the ventricle that increases the likelihood of coagulation and thromboembolic events.8, 19 Evidence indicates that the risk of stroke is inversely correlated with LVEF; LVEF of 29%35% carries a cumulative 5‐year stroke risk of 7.8%, and LVEF of 28% or below carries a 5‐year risk of 8.9%.8, 20, 21 Data from the Survival and Ventricular Enlargement (SAVE) study showed an 18% increase in the risk of stroke for every 5% decline in LVEF,19, 21 and the Studies of Left Ventricular Dysfunction (SOLVD) trial found a 58% increase in thromboembolic events for every 10% decrease in LVEF among women (P = .01).19, 22 Among patients with low LVEF who have had a stroke, the 5‐year recurrent stroke rate may be as high as 45%.19, 23

Although it would appear that stroke associated with left ventricular dysfunction and a low LVEF may potentially be cardioembolic in origin, risk reduction for recurrent stroke has not been adequately investigated as a primary end point in clinical trials, particularly in the absence of atrial fibrillation.24 Thus, the question of whether antiplatelet or anticoagulant therapy would be more effective has not yet been answered. However, results of secondary end point analyses in the SOLVD and SAVE trials suggested that patients had a lower risk of sudden death, thromboembolism, and stroke with antiplatelet therapy.21, 2426 In an observational analysis of prospectively collected data on patients enrolled in the SAVE trial, use of aspirin reduced the overall risk of stroke by 66% in patients with an LVEF below 28%.21 Warfarin is the standard of care for stroke prevention in atrial fibrillation, and the 2 conditions often coexist. In those patients, warfarin is the recommended therapy.24

In patients with sinus rhythm and a low LVEF, the choice is less clear. The results of the Warfarin/Aspirin Study in Heart failure (WASH) failed to establish efficacy or safety for aspirin in preventing all‐cause mortality, nonfatal MI, and nonfatal stroke in patients with heart failure. Patients treated with aspirin were significantly more likely to be hospitalized for cardiovascular events, especially worsening heart failure.27 The trial found no significant difference for the composite end point between the 3 treatment groups: aspirin, warfarin, or no antithrombotic treatment. However, this was a small trial, and the findings were far from definitive, as the study was designed primarily to be a feasibility study to aid in the design of a larger outcomes study.24 Because of the inconsistent results and lack of well‐designed studies regarding the benefit of aspirin or anticoagulation for secondary stroke prevention in patients with LVEF in the absence of atrial fibrillation, further study is needed.

More recently, results were presented from the Warfarin and Antiplatelet Therapy in Heart Failure Trial (WATCH), which randomized patients with heart failure, sinus rhythm, and LVEF of 35% or below to either aspirin 162 mg, warfarin (target international normalized ratio [INR] 2.53.0), or clopidogrel.28, 29 Two major comparisons were plannedwarfarin versus aspirin and aspirin versus clopidogrel.28 Whereas warfarin therapy was open‐label because of the need to check blood levels, antiplatelet therapy was given in a double‐blind manner. After a mean follow‐up of 23 months, no significant differences were found for the primary composite end point of all‐cause mortality, nonfatal MI, and nonfatal stroke, which occurred in 20.5% of those on aspirin, 19.8% on warfarin, and 21.8% on clopidogrel. However, for the secondary end point of stroke, there was a strong trend favoring warfarin over aspirin: stroke occurred in 0.7% of patients taking warfarin versus 2.1% of those taking aspirin (P = .06).24, 29 However, the WATCH investigators concluded that the question of warfarin's value for patients with low LVEF and sinus rhythm remained unresolved.29

In the absence of clear data, the American Heart Association (AHA)/American Stroke Association (ASA) guidelines on stroke prevention in this patient population recommend either warfarin (INR 2.03.0) or antiplatelet therapy, including aspirin (50325 mg/day), aspirin plus extended‐release dipyridamole (200 mg twice daily), or clopidogrel (75 mg/day).8 Patients with coexisting atrial fibrillation should be treated with warfarin, or if unable to tolerate that agent, aspirin 325 mg/day.8

The Warfarin Versus Aspirin for Reduced Cardiac Ejection Fraction (WARCEF) trial may provide more definitive answers on the best approach for reducing the risk of recurrent stroke in patients with low LVEF. The study will compare warfarin (INR 2.53.0) and aspirin (325 mg/day) in the prevention of all‐cause mortality and all strokes (ischemic and hemorrhagic) in patients with an LVEF of 35% or below but no atrial fibrillation.30 The study has a target enrollment of 2860 patients, who are being recruited at 70 North American and 70 European sites, and it will include patients with recent stroke or TIA.28 The results are anxiously anticipated.

INTRACRANIAL STENOSIS

Stroke patients with symptomatic intracranial atherosclerosis have a high risk of recurrent strokein the range of 10% per yearand this accounts for approximately 8% of ischemic strokes.8, 31, 32 Intracranial stenosis appears to be more common in African Americans and Hispanics than in white patients.31

Recurrent stroke prevention in patients with intracranial stenosis was explored in the Warfarin‐Aspirin Symptomatic Intracranial Disease (WASID) study, a multicenter, double‐blind trial. Patients with angiographically verified 50%99% stenosis of a major intracranial artery who had experienced either a stroke or TIA were randomized to either warfarin (target INR 2.03.0) or high‐dose aspirin (1300 mg/day). The primary end point was ischemic stroke, brain hemorrhage, or death from vascular causes other than stroke.33 Mean follow‐up was 1.8 years, and enrollment was stopped after 569 patients had been randomized because of concerns about the safety of warfarin in this patient population.33 The primary end point occurred in 22.1% of those treated with aspirin and 21.8% of those treated with warfarin.33 There were no significant differences between the 2 treatment groups for any of the prespecified secondary end points, including ischemic stroke in any vascular territory and ischemic stroke in the territory of the stenotic intracranial artery.33

The rate of death was significantly higher in the warfarin group (9.7%) than in the aspirin group (4.3%; P = .02). Patients in the warfarin group had higher rates of death from both vascular and nonvascular causes.33 Major hemorrhage was significantly more common in the warfarin group (8.3%) than in the aspirin group (3.2%; P = .01). The investigators concluded that warfarin should not be used as first‐line prevention of recurrent stroke in patients with intracranial stenosis. However, there was a significant association between an INR less than 2 and increased risk of ischemic stroke and major cardiac events (P < .001) as well as a significant increase in major hemorrhages in patients with INRs greater than 3 (P < .001).33

The failure of many patients in the study to remain within the therapeutic INR casts doubt on these results to some extent, although this may actually mirror a common real‐world scenario. Patients were within the therapeutic INR goal only 63% of the time. Furthermore, a nonstandard high dose of aspirin (1300 mg/day) was used, which also may have affected the results.34 Others looking at this data have suggested that aspirin remains an imperfect therapy, with an unacceptably high risk of ischemic stroke and other vascular events, and that anticoagulation may play a role in the period immediately following ischemic stroke or TIA with transition to antiplatelet therapy.34 This would require additional investigation.34

The current AHA/ASA guidelines recommend that for patients with noncardioembolic ischemic stroke or TIA, antiplatelet agents rather than oral anticoagulants be used to reduce the risk of recurrent stroke (class I, level A). Aspirin (50325 mg/day), the combination of aspirin and extended‐release dipyridamole, and clopidogrel are all acceptable options for initial therapy (class IIa, level A).8 The combination of aspirin and extended‐release dipyridamole is suggested instead of aspirin alone (class IIa, level A), and clopidogrel may be considered instead of aspirin alone (class IIb, level B).8 However, data are insufficient at this point to make evidence‐based recommendations between antiplatelet options other than aspirin.8 In patients with significant intracranial stenosis whose symptoms persist despite medical therapy, including antithrombotics, statins, and antihypertensives, endovascular therapy with angioplasty and/or stent placement is an option, but it remains investigational and its value is uncertain.8

CAROTID STENOSIS

Asymptomatic carotid stenosis greater than 50% has been found in 7% of men and 5% of women older than 65 years.35, 36 Among those with asymptomatic carotid stenosis greater than 50%, there is an annual risk of stroke of up to 3.4%.35 In such patients, the benefit of carotid endarterectomy (CEA) is highly dependent on the surgical risk, and if complication rates exceed 3.0%, benefit is eliminated.35 The AHA/ASA guidelines recommend that patients be given treatment for all identifiable risk factors, including statins for dyslipidemia, antihypertensives for hypertension, and aspirin as an antiplatelet agent. In select patients with high‐grade asymptomatic carotid stenosis, CEA performed by a surgeon with a morbidity/mortality rate below 3% is recommended.35 In asymptomatic patients with greater than 70% carotid stenosis, CEA can be an effective therapy. Trial data indicate that the overall 5‐year risk of any stroke or perioperative death is 11.8% for deferred surgery versus 6.4% for immediate endarterectomy (P < .0001).35, 37 Unfortunately, data on the value of stents or angioplasty compared with CEA in this patient population are limited.35

In patients who have had a recent TIA or stroke, carotid stenosis would be considered symptomatic. In these patients, the benefit of CEA is strongly associated with the degree of stenosis. Data from the Carotid Endarterectomy Trialists' Collaboration and North American Symptomatic Carotid Endarterectomy Trial (NASCET) have shown that in patients with stenosis greater than 70%, CEA reduces the absolute 5‐year risk of ischemic stroke by 16.0% (P < .001), whereas in patients with 50%69% stenosis, the 5‐year absolute risk reduction is 4.6% (P = .04). In those with stenosis of 30%49%, there is no effect, and CEA in patients with less than 30% stenosis increases the risk of stroke.38, 39 In patients with 50%69% stenosis, benefit is achieved only if patients at highest risk are selected.40 Recent data have also questioned the typical 4‐ to 6‐week delay before performing a CEA following a nondisabling stroke. Rothwell et al. found that surgery performed within 2 weeks of such a stroke was not associated with increased operative risk.41 Moreover, benefit from CEA fell rapidly within the first few weeks after a TIA or stroke, particularly in women, perhaps reflecting the high risk of recurrent stroke in the period immediately following an initial event.41

Angioplasty or stents have been investigated as alternatives to CEA, but the evidence to date has been disappointing. The Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS) demonstrated preventive efficacy and major risks similar to those found for CEA after 3 years of follow‐up in 504 patients with carotid stenosis.42 However, a more recent study was stopped prematurely after 527 patients had been enrolled because of a higher incidence of disabling stroke or death at 30 days in the stenting cohort (3.4%) compared with the CEA cohort (1.5%). The 30‐day incidence of any stroke or death was 3.9% after CEA and 9.6% after stenting, yielding a relative risk of 2.5 for stenting.43 The Stent‐Protected Angioplasty Versus Carotid Endarterectomy in Symptomatic Patients (SPACE) trial has also failed to find benefit for carotid stenting and/or angioplasty in comparison with CEA.44

The AHA/ASA guidelines recommend CEA in patients with ipsilateral severe (70%99%) stenosis and a recent TIA or ischemic stroke (within 6 months). Surgery should be performed by a surgeon with a perioperative morbidity/mortality rate less than 6%.8 In patients with 50%69% stenosis, the advisability of CEA depends on patient factors such as age, sex, comorbidities, and severity of symptoms. Surgery should be performed within 2 weeks of an ischemic event. In patients with severe stenosis in whom CEA would be difficult to perform, carotid angioplasty or stenting may be recommended if performed by practitioners with a morbidity/mortality rate less than 4%6%.8 The Seventh ACCP Conference also recommends that patients undergoing CEA receive aspirin 81325 mg/day prior to and following the procedure.7

ATHEROSCLEROSIS OF THE AORTIC ARCH

Atherosclerosis of the aortic arch contributes significantly as an independent factor to risk of embolic stroke.7 Such plaques can be detected using transesophageal echocardiography; those that are thicker than 45 mm, exhibit ulceration, or have mobile components place individuals at higher risk for stroke.7, 45 The stroke risk associated with aortic arch plaques greater than 5 mm is as high as 33% per year.7, 46

However, data from large‐scale randomized clinical trials on the efficacy of therapeutic interventions in this condition are lacking. Two small trials found efficacy for warfarin in patients with mobile thrombi in the thoracic aorta. In one, patients given oral anticoagulants had better outcomes than those treated with antiplatelet agents, and in the other, warfarin proved to be more effective than no treatment.47, 48 A retrospective trial that looked at 519 patients treated with warfarin, antiplatelet agents, or statins found there was a protective effect of statins, with an absolute risk reduction in embolic events, including ischemic stroke, TIA, and peripheral embolization of 17%, and a relative risk reduction in embolic events of 59%. The odds ratio for embolic events was 0.39 for statins, 0.77 for antiplatelet agents, and 1.18 for warfarin.49 The French Study of Aortic Plaque in Stroke found no significant difference in risk of events between those treated with warfarin and those treated with aspirin; however, this study was not designed as a therapeutic trial, and few patients received warfarin, casting doubt on this finding.45

Given the paucity of data, suggestions for treatment of patients with an aortic arch atheromata are difficult. Certainly, statin therapy, which would address general atherosclerotic risk reduction, can be initiated. Warfarin appeared to be more effective than antiplatelet agents in several of the studies; however some have expressed concern about the possibility of anticoagulation increasing the risk of cholesterol embolism in these patients.7

SYMPTOMATIC CORONARY ARTERY DISEASE

For patients with a history of ischemic stroke or TIA who have symptomatic CAD, their condition must be managed for both stroke and CAD risks. In patients with stable or unstable angina and a history of stroke or TIA, similar risks must be managed. The acute treatment of ACS or symptomatic CAD cannot be adequately addressed here; however, it may involve a number of therapeutic modalities, including PCI, ‐blocker therapy, glycoprotein IIb/IIIa inhibitors, anticoagulant therapy, angiotensin‐converting enzyme (ACE) inhibitors, and clopidogrel plus aspirin, depending on the exact nature of the syndrome.5054 The long‐term management and, in particular, prevention of recurrent stroke in the setting of symptomatic CAD are the focus here. As with a patient with a history of CAD and a recent TIA or stroke (as discussed earlier), patients with symptomatic CAD and TIA or stroke must be managed for multiple risk factors. NCEP guidelines recommend aggressive cholesterol lowering with statin therapy. Hypertension must be addressed as well, and long‐term therapy with ‐blockers and ACE inhibitors has been shown to reduce mortality in patients with ACS and is recommended by the AHA/ASA.5355

Once the acute ACS period has resolved, it is reasonable to address the question of the best possible antiplatelet therapy for long‐term stroke prevention. Long‐term use of clopidogrel plus aspirin is not advisable given the increased risk of bleeding events noted in the MATCH and CHARISMA trials.16, 17 At this point, it would be reasonable to start the patient on aspirin 75150 mg/day, which reduces risk of stroke up to 25%,56, 57 aspirin plus extended‐release dipyridamole, which reduces risk by about 37%,57, 58 or clopidogrel 75 mg/day, which reduces the relative risk for stroke alone by 7.3% compared with aspirin.59 In patients who cannot tolerate or are allergic to aspirin, clopidogrel is a reasonable choice.8

ANTIPLATELET FAILURE

Patients who have failed antiplatelet therapythat is, have gone on to have a recurrent strokeare particularly difficult. It is important to remember that any therapeutic intervention only reduces stroke risk; it does not eliminate it. Keeping that in mind, it is essential to reevaluate and reconsider both the original diagnosis and the etiology of the stroke or TIA. A number of diagnostic alternatives should be considered, including sensory seizure and migraine equivalents, as well as other etiologies, such as atrial fibrillation or cerebral amyloid angiopathy. Therapy may have to be adjusted accordingly, but the patient remains at increased risk for stroke recurrence, and thus preventive therapy is critical.

Several key points should be remembered. As outlined previously in this article, if the stroke is still thought to be noncardioembolic in origin, a reduction in the risk of stroke has not been found for those patients receiving warfarin, an increased dose of aspirin, a combination of antiplatelet agents and warfarin, or clopidogrel plus aspirin.8, 16, 31, 60, 61 However, if atrial fibrillation has developed in the patient, the recommendation is warfarin (INR 2.03.0) or, if anticoagulants cannot be taken, aspirin 325 mg/day.8 Risk factors should be reassessed and managed, with agents and lifestyle changes to control hypertension and dyslipidemia. Antiplatelet agents should be continued in patients with noncardioembolic stroke. Acceptable antiplatelet agents include aspirin (50325 mg/day), aspirin plus extended‐release dipyridamole, and clopidogrel. The combination of aspirin plus extended‐release dipyridamole is suggested over aspirin alone. If the patient cannot tolerate or is allergic to aspirin, clopidogrel is a reasonable alternative.8 The decision of which antiplatelet agent to use should be based on the individual patient's risk factor profile.8 The temptation to put patients on anticoagulation therapy because of a wish to do more should be avoided, as this is likely to expose patients to increased risk without known benefit.60, 61

Consider a common case scenarioa patient with a known history of hypertension and TIA presents with a 30‐minute episode of left arm numbness. The patient has been adherent to his prescribed medications, including aspirin 81 mg/day. What is the appropriate approach to acute treatment at this time? This is a common scenario in emergency departmentsnew‐onset TIA while taking aspirin 81 mg/day. There are advocates for several different treatment regimens in these patients: increasing the aspirin dose to 325 mg/day as a new treatment; discontinuing aspirin and initiating clopidogrel 75 mg/day; discontinuing aspirin 81 mg/day and initiating aspirin 325 mg/day plus clopidogrel 75 mg/day; or discontinuing aspirin 81 mg/day and initiating a combination of aspirin 25 mg plus extended‐release dipyridamole 200 mg twice daily. It is clear that patients with the same disease are treated differently in different institutions. What is the appropriate evidence‐based treatment in this case? The answer is clearno evidence supports increasing the dose of aspirin as a new treatment for this case or initiating aspirin 325 mg/day plus clopidogrel 75 mg/day.16, 17 Based on the literature, for a patient who has recently had another cerebral ischemic event while on treatment, it would make sense to consider switching to another agent. Three agents are recommended by the guidelines: aspirin, clopidogrel, and aspirin plus extended‐release dipyridamole. If treatment 1 were to fail, it would not be against the evidence to initiate treatment 2 or 3.

PATIENTS ON WARFARIN

Data from the Warfarin‐Aspirin Recurrent Stroke Study (WARSS), a large‐scale recurrent stroke prevention trial conducted in 2206 patients, demonstrated that there was no survival benefit for noncardioembolic stroke survivors who were treated with warfarin.60, 61 Yet there are patients still taking warfarin to reduce stroke risk who do not have atrial fibrillation. Unless a patient is allergic to or intolerant of antiplatelet agents such as aspirin, clopidogrel, or dipyridamole, they should not be treated with warfarin for noncardioembolic stroke risk.8 The results of other studies of anticoagulation in recurrent stroke prevention, including the European/Australasian Stroke Prevention in Reversible Ischaemia Trial (ESPRIT),62 the Stroke Performance for Reporting the Improvement and Translation (SPIRIT) trial,63 and the WASID study,33 have yet to demonstrate a role for warfarin in prevention of noncardioembolic stroke.

Given these trial results, patients currently on warfarin who do not have a cardioembolic risk factor should be placed on antiplatelet therapy with aspirin, aspirin plus extended‐release dipyridamole, or clopidogrel 35 days after discontinuing warfarin therapy. However, it would be advisable to evaluate these patients for atrial fibrillation, as patients with that risk factor should remain on warfarin.8

SUMMARY

In clinical practice, health care providers often must manage patients with complex profiles. Multiple risk factors and comorbidities complicate treatment of these individuals, and robust clinical data are often lacking as clinical trials rarely include such individuals. Guidelines offer recommendations, but these too are often based on extrapolations from clinical trial data. This is particularly true of patients at risk for ischemic stroke, as the primary underlying causevascular diseasehas systemic implications and comorbidities that often complicate treatment.

In general, antiplatelet therapy should be used to prevent recurrent stroke in patients with TIA or noncardioembolic stroke, whereas anticoagulation therapy should be used in patients with cardioembolic stroke such as that caused by atrial fibrillation. However, therapy must be individualized to account for the patient's full risk profile. Conditions such as dyslipidemia and hypertension must be addressed as well, as these not only give rise to stroke but also to the CAD, coronary heart disease, and ACS that may coexist with stroke. Among patients deemed suitable for antiplatelet therapy, class IIa, level A evidence supports the use of aspirin 50325 mg/day, the combination of aspirin and extended‐release dipyridamole, and clopidogrel for secondary prevention of stroke.8

Each year in the United States 700,000 individuals experience a stroke500,000 of them for the first time. Despite advances in stroke prevention, this number has increased dramatically over the last quarter century.1 Between 1979 and 2004, the annual number of hospital discharges with stroke as a primary diagnosis swelled to 906,000, a 21% increase over the rate in 1979.1 In the next 1015 years, this number is predicted to double in parallel with a doubling of the number of Americans older than age 65 years. Mortality from stroke is projected to increase faster than the overall US population.2 In addition, the prevalence of diabetes, a major ischemic stroke risk factor, is increasing at an alarming rate.1 A second major risk factor, hypertension, also occurs more frequently in older people and thus is expected to increase in prevalence over the next few decades.1, 3 Blacks, Hispanics, and Mexican Americans, growing segments of the US population, are disproportionately affected by stroke.1

The impact of stroke extends far beyond the initial episode. Stroke is a leading cause of long‐term disability in the United States.1 Total estimated cost for stroke care in 2007 is $62.7 billion. Prevention is the key to reducing the grave personal and societal burden of this condition.

Efforts to prevent the approximately 200,000 recurrent strokes that occur each year are critical. Stroke itself is a harbinger of future stroke, and secondary strokes are frequently more severe and disabling.4 Numerous studies have found that among stroke patients, recurrent stroke is the most likely secondary cardiovascular event, particularly in the first few months following the index event (only in the first 3 months, however; then death from cardiac disease becomes more important; Fig. 1).5, 6 Transient ischemic attack (TIA), once considered a relatively benign event, is now recognized as a significant risk factor for stroke.7, 8 A recent study suggests that 1 in 10 TIA patients will have a stroke in the 90 days after the event, and 24% of those strokes will occur within 48 hours.8 Moreover, improved imaging techniques have revealed that even patients with resolution of symptoms within 1 hour may have evidence of infarction.9, 10 The longer the duration of symptoms, the greater the probability of infarction detectable with magnetic resonance imaging.9, 10 Because the greatest risk of recurrent stroke occurs within hours of the first event, secondary prevention must be initiated as soon as possible after diagnosis.11

Figure 1

MANAGEMENT OF ACUTE STROKE BY HOSPITALISTS

Stroke care is a rapidly evolving field in which expeditious and careful inpatient care significantly affects outcome. Hospitalists are in a unique position to improve acute stroke care and initiate secondary stroke prevention in several ways. First, there is a shortage of neurologists to care for patients with stroke. In one survey of Medicare data from 1991, prior to the widespread presence of hospitalists, only 1 in 9 stroke patients (11%) had a neurologist as the attending physician.12 At that time, there were only 3.25 nonfederal patient care neurologists per 100,000 population. Although the ratio may have improved somewhat in the intervening years (there were an estimated 5.3 self‐reported neurologists per 100,000 population as of 2005),13 the limited number of neurologists combined with the increasing incidence of stroke is expected to reduce the fraction of stroke patients having a neurologist involved in their care. Because neurology practices tend to be concentrated in urban areas, the shortage is likely to affect nonurban areas to a greater degree. The number of hospitalists, currently estimated to be 20,000 in the United States, is projected to reach 30,000 by 2010.14 In the simplest terms, hospitalists are the logical choice to fill the need for physicians to manage inpatient stroke.

Perhaps the most compelling reason for hospitalists to be involved in the care of stroke patients is clinical: patients with stroke frequently have multiple comorbid conditions that affect outcomes and are not within the traditional purview of neurology. A retrospective analysis of data from 1802 patients seen in a geriatric practice revealed that 56% of patients with stroke also had coronary artery disease, and 28% had peripheral arterial disease.15 In addition, the major risk factors for strokediabetes and hypertensionwould be expected to be prevalent in this population. Timely and effective management can improve secondary stroke prevention as well as prevent exacerbation of existing conditions.

A recent report compared outcomes in 44,099 patients following stroke according to physician specialty.16 Although patients treated by neurologists alone had a 10% lower risk of 30‐day mortality compared with those treated by generalists (family practice physicians, general practitioners, or internists) despite having more severe stroke, collaborative care reduced that risk an additional 6%.16 The risk of rehospitalization for infections and aspiration pneumonia within 30 days was 12% lower for those treated by neurologists. However, these patients had a significant, 17% increased relative risk of rehospitalization for coronary heart disease (95% confidence interval [CI], 1.021.34).16

Comanagement of stroke patients by hospitalists and neurologists is likely to become more common over time, as proposed by Likosky and Amin.17 Although studies have not specifically compared outcomes in patients with stroke who have been treated by hospitalists versus other types of physicians, implementation of hospitalist services has been associated with improved short‐term mortality and rehospitalization rates compared with traditional care.1820 Approximately 85% of hospitalists are trained in internal medicine.21 In addition, they have skill sets focusing on the specialized needs of inpatients. As hospitalists assume a greater role in the management of stroke, research into the benefits of collaborative care can be explored.

Finally, hospitalists are ideally positioned to champion the use of standardized protocols for secondary stroke prevention at their institutions. Results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke Registry showed that a minority of acute stroke patients are treated according to established guidelines.22 The 4 prototype registries were in Georgia, Massachusetts, Michigan, and Ohio. The percentage of relevant patient populations that had lipid profiles assessed ranged between 28% and 34%. For smoking‐cessation education, the range was between 17% and 34%. Anticoagulant prescribing for relevant populations at discharge ranged from 64% to 90%, and antithrombotic prescribing ranged from 88% to 98%.22

The use of protocols that initiate secondary prevention of cerebrovascular and cardiovascular events has been demonstrated to improve patient adherence to evidence‐based treatment after discharge.2328 The Preventing Recurrence of Thromboembolic Events Through Coordinated Treatment (PROTECT) program was designed to integrate secondary stroke prevention measures into the standard stroke care provided during acute hospitalization (Table 1).26 Use of appropriate antithrombotic medication was achieved in 100% of cases. Use of statins, angiotensin‐converting enzyme inhibitors or angiotensin receptor blockers, and thiazide diuretics improved significantly during the first year of implementation (P < .001). Patient education in all 4 of the areas established was carried out in 100% of patients prior to discharge.26 Tools for establishing similar hospital‐based secondary prevention programs are presently available from the University of California at Los Angeles PROTECT Program and other programs.

An essential part of any effort to develop standardized treatment procedures must include a plan to minimize any discontinuity of care after discharge. Standardized procedures need to be implemented to ensure communication of discharge summaries to outpatient clinicians in a timely and complete fashion. Only 19% of 226 outpatient physicians responding to a recent survey were satisfied or very satisfied with the timeliness of discharge summaries they received for their patients.29 Approximately one third of respondents reported that most of their patients (60%) were seen for their follow‐up outpatient visit before discharge summaries had been received. Only about one third (32%) of the respondents were satisfied or very satisfied with the summary content. Forty‐one percent believed that at least 1 of their patients hospitalized in the previous 6 months had experienced an adverse event that could have been prevented with improved transfer of discharge information.29

Development of electronic discharge summaries is an obvious alternative to conventional paper versions. This area has received less attention than others that more directly affect patient care. As the primary inpatient physicians, hospitalists can effectively implement improvements in communication among hospital staff and outpatient health care providers.

SUMMARY

This supplement is a call to action for hospitalists based on a roundtable discussion conducted in March 2007. Participants included hospitalists, neurohospitalists, vascular neurologists, and neurointensivists. The objectives of the meeting were to review the clinical data supporting current practice guidelines for secondary prevention of noncardioemboic ischemic stroke, to develop best‐practice recommendations for hospitalist‐based care of stroke inpatients, and finally to recommend improvements in transfer of information to outpatient health care providers.

The consensus of the participants is reported in the following 3 articles. The first, Evidence‐based Medicine: Review of Guidelines and Trials in Prevention of Secondary Stroke, includes an overview of the pathophysiology of stroke and TIA and reviews the clinical data supporting current treatment guidelines. Several case studies illustrating challenging or difficult aspects of secondary stroke prevention are presented in the second article, Secondary Prevention of Ischemic Stroke: Challenging Patient Scenarios. These cases focus on commonly encountered difficulties for which there may not be clear evidence or consensus. In the final article, Systems Approach to Standardization of Care in the Secondary Prevention of Noncardioembolic Ischemic Stroke, the best‐practices recommendations developed at the roundtable are presented. The role of the hospitalist in long‐term prevention strategies and the effective transfer of care to outpatient providers are discussed.

As the hospitalist movement grows, hospital‐based physicians need to identify opportunities to use their unique skills. By taking the lead in improving processes that result in better patient outcomes, hospitalists can ensure that the value of this nascent field will continue to gain recognition in the broader, sometimes skeptical medical community. We sincerely hope that you agree that integrating secondary prevention into inpatient acute stroke care is just such an opportunity. Furthermore, we hope the information we have provided will be useful to you in your hospital‐based practice.

Each year in the United States 700,000 individuals experience a stroke500,000 of them for the first time. Despite advances in stroke prevention, this number has increased dramatically over the last quarter century.1 Between 1979 and 2004, the annual number of hospital discharges with stroke as a primary diagnosis swelled to 906,000, a 21% increase over the rate in 1979.1 In the next 1015 years, this number is predicted to double in parallel with a doubling of the number of Americans older than age 65 years. Mortality from stroke is projected to increase faster than the overall US population.2 In addition, the prevalence of diabetes, a major ischemic stroke risk factor, is increasing at an alarming rate.1 A second major risk factor, hypertension, also occurs more frequently in older people and thus is expected to increase in prevalence over the next few decades.1, 3 Blacks, Hispanics, and Mexican Americans, growing segments of the US population, are disproportionately affected by stroke.1

The impact of stroke extends far beyond the initial episode. Stroke is a leading cause of long‐term disability in the United States.1 Total estimated cost for stroke care in 2007 is $62.7 billion. Prevention is the key to reducing the grave personal and societal burden of this condition.

Efforts to prevent the approximately 200,000 recurrent strokes that occur each year are critical. Stroke itself is a harbinger of future stroke, and secondary strokes are frequently more severe and disabling.4 Numerous studies have found that among stroke patients, recurrent stroke is the most likely secondary cardiovascular event, particularly in the first few months following the index event (only in the first 3 months, however; then death from cardiac disease becomes more important; Fig. 1).5, 6 Transient ischemic attack (TIA), once considered a relatively benign event, is now recognized as a significant risk factor for stroke.7, 8 A recent study suggests that 1 in 10 TIA patients will have a stroke in the 90 days after the event, and 24% of those strokes will occur within 48 hours.8 Moreover, improved imaging techniques have revealed that even patients with resolution of symptoms within 1 hour may have evidence of infarction.9, 10 The longer the duration of symptoms, the greater the probability of infarction detectable with magnetic resonance imaging.9, 10 Because the greatest risk of recurrent stroke occurs within hours of the first event, secondary prevention must be initiated as soon as possible after diagnosis.11

Figure 1

MANAGEMENT OF ACUTE STROKE BY HOSPITALISTS

Stroke care is a rapidly evolving field in which expeditious and careful inpatient care significantly affects outcome. Hospitalists are in a unique position to improve acute stroke care and initiate secondary stroke prevention in several ways. First, there is a shortage of neurologists to care for patients with stroke. In one survey of Medicare data from 1991, prior to the widespread presence of hospitalists, only 1 in 9 stroke patients (11%) had a neurologist as the attending physician.12 At that time, there were only 3.25 nonfederal patient care neurologists per 100,000 population. Although the ratio may have improved somewhat in the intervening years (there were an estimated 5.3 self‐reported neurologists per 100,000 population as of 2005),13 the limited number of neurologists combined with the increasing incidence of stroke is expected to reduce the fraction of stroke patients having a neurologist involved in their care. Because neurology practices tend to be concentrated in urban areas, the shortage is likely to affect nonurban areas to a greater degree. The number of hospitalists, currently estimated to be 20,000 in the United States, is projected to reach 30,000 by 2010.14 In the simplest terms, hospitalists are the logical choice to fill the need for physicians to manage inpatient stroke.

Perhaps the most compelling reason for hospitalists to be involved in the care of stroke patients is clinical: patients with stroke frequently have multiple comorbid conditions that affect outcomes and are not within the traditional purview of neurology. A retrospective analysis of data from 1802 patients seen in a geriatric practice revealed that 56% of patients with stroke also had coronary artery disease, and 28% had peripheral arterial disease.15 In addition, the major risk factors for strokediabetes and hypertensionwould be expected to be prevalent in this population. Timely and effective management can improve secondary stroke prevention as well as prevent exacerbation of existing conditions.

A recent report compared outcomes in 44,099 patients following stroke according to physician specialty.16 Although patients treated by neurologists alone had a 10% lower risk of 30‐day mortality compared with those treated by generalists (family practice physicians, general practitioners, or internists) despite having more severe stroke, collaborative care reduced that risk an additional 6%.16 The risk of rehospitalization for infections and aspiration pneumonia within 30 days was 12% lower for those treated by neurologists. However, these patients had a significant, 17% increased relative risk of rehospitalization for coronary heart disease (95% confidence interval [CI], 1.021.34).16

Comanagement of stroke patients by hospitalists and neurologists is likely to become more common over time, as proposed by Likosky and Amin.17 Although studies have not specifically compared outcomes in patients with stroke who have been treated by hospitalists versus other types of physicians, implementation of hospitalist services has been associated with improved short‐term mortality and rehospitalization rates compared with traditional care.1820 Approximately 85% of hospitalists are trained in internal medicine.21 In addition, they have skill sets focusing on the specialized needs of inpatients. As hospitalists assume a greater role in the management of stroke, research into the benefits of collaborative care can be explored.

Finally, hospitalists are ideally positioned to champion the use of standardized protocols for secondary stroke prevention at their institutions. Results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke Registry showed that a minority of acute stroke patients are treated according to established guidelines.22 The 4 prototype registries were in Georgia, Massachusetts, Michigan, and Ohio. The percentage of relevant patient populations that had lipid profiles assessed ranged between 28% and 34%. For smoking‐cessation education, the range was between 17% and 34%. Anticoagulant prescribing for relevant populations at discharge ranged from 64% to 90%, and antithrombotic prescribing ranged from 88% to 98%.22

The use of protocols that initiate secondary prevention of cerebrovascular and cardiovascular events has been demonstrated to improve patient adherence to evidence‐based treatment after discharge.2328 The Preventing Recurrence of Thromboembolic Events Through Coordinated Treatment (PROTECT) program was designed to integrate secondary stroke prevention measures into the standard stroke care provided during acute hospitalization (Table 1).26 Use of appropriate antithrombotic medication was achieved in 100% of cases. Use of statins, angiotensin‐converting enzyme inhibitors or angiotensin receptor blockers, and thiazide diuretics improved significantly during the first year of implementation (P < .001). Patient education in all 4 of the areas established was carried out in 100% of patients prior to discharge.26 Tools for establishing similar hospital‐based secondary prevention programs are presently available from the University of California at Los Angeles PROTECT Program and other programs.

An essential part of any effort to develop standardized treatment procedures must include a plan to minimize any discontinuity of care after discharge. Standardized procedures need to be implemented to ensure communication of discharge summaries to outpatient clinicians in a timely and complete fashion. Only 19% of 226 outpatient physicians responding to a recent survey were satisfied or very satisfied with the timeliness of discharge summaries they received for their patients.29 Approximately one third of respondents reported that most of their patients (60%) were seen for their follow‐up outpatient visit before discharge summaries had been received. Only about one third (32%) of the respondents were satisfied or very satisfied with the summary content. Forty‐one percent believed that at least 1 of their patients hospitalized in the previous 6 months had experienced an adverse event that could have been prevented with improved transfer of discharge information.29

Development of electronic discharge summaries is an obvious alternative to conventional paper versions. This area has received less attention than others that more directly affect patient care. As the primary inpatient physicians, hospitalists can effectively implement improvements in communication among hospital staff and outpatient health care providers.

SUMMARY

This supplement is a call to action for hospitalists based on a roundtable discussion conducted in March 2007. Participants included hospitalists, neurohospitalists, vascular neurologists, and neurointensivists. The objectives of the meeting were to review the clinical data supporting current practice guidelines for secondary prevention of noncardioemboic ischemic stroke, to develop best‐practice recommendations for hospitalist‐based care of stroke inpatients, and finally to recommend improvements in transfer of information to outpatient health care providers.

The consensus of the participants is reported in the following 3 articles. The first, Evidence‐based Medicine: Review of Guidelines and Trials in Prevention of Secondary Stroke, includes an overview of the pathophysiology of stroke and TIA and reviews the clinical data supporting current treatment guidelines. Several case studies illustrating challenging or difficult aspects of secondary stroke prevention are presented in the second article, Secondary Prevention of Ischemic Stroke: Challenging Patient Scenarios. These cases focus on commonly encountered difficulties for which there may not be clear evidence or consensus. In the final article, Systems Approach to Standardization of Care in the Secondary Prevention of Noncardioembolic Ischemic Stroke, the best‐practices recommendations developed at the roundtable are presented. The role of the hospitalist in long‐term prevention strategies and the effective transfer of care to outpatient providers are discussed.

As the hospitalist movement grows, hospital‐based physicians need to identify opportunities to use their unique skills. By taking the lead in improving processes that result in better patient outcomes, hospitalists can ensure that the value of this nascent field will continue to gain recognition in the broader, sometimes skeptical medical community. We sincerely hope that you agree that integrating secondary prevention into inpatient acute stroke care is just such an opportunity. Furthermore, we hope the information we have provided will be useful to you in your hospital‐based practice.

Each year in the United States 700,000 individuals experience a stroke500,000 of them for the first time. Despite advances in stroke prevention, this number has increased dramatically over the last quarter century.1 Between 1979 and 2004, the annual number of hospital discharges with stroke as a primary diagnosis swelled to 906,000, a 21% increase over the rate in 1979.1 In the next 1015 years, this number is predicted to double in parallel with a doubling of the number of Americans older than age 65 years. Mortality from stroke is projected to increase faster than the overall US population.2 In addition, the prevalence of diabetes, a major ischemic stroke risk factor, is increasing at an alarming rate.1 A second major risk factor, hypertension, also occurs more frequently in older people and thus is expected to increase in prevalence over the next few decades.1, 3 Blacks, Hispanics, and Mexican Americans, growing segments of the US population, are disproportionately affected by stroke.1

The impact of stroke extends far beyond the initial episode. Stroke is a leading cause of long‐term disability in the United States.1 Total estimated cost for stroke care in 2007 is $62.7 billion. Prevention is the key to reducing the grave personal and societal burden of this condition.

Efforts to prevent the approximately 200,000 recurrent strokes that occur each year are critical. Stroke itself is a harbinger of future stroke, and secondary strokes are frequently more severe and disabling.4 Numerous studies have found that among stroke patients, recurrent stroke is the most likely secondary cardiovascular event, particularly in the first few months following the index event (only in the first 3 months, however; then death from cardiac disease becomes more important; Fig. 1).5, 6 Transient ischemic attack (TIA), once considered a relatively benign event, is now recognized as a significant risk factor for stroke.7, 8 A recent study suggests that 1 in 10 TIA patients will have a stroke in the 90 days after the event, and 24% of those strokes will occur within 48 hours.8 Moreover, improved imaging techniques have revealed that even patients with resolution of symptoms within 1 hour may have evidence of infarction.9, 10 The longer the duration of symptoms, the greater the probability of infarction detectable with magnetic resonance imaging.9, 10 Because the greatest risk of recurrent stroke occurs within hours of the first event, secondary prevention must be initiated as soon as possible after diagnosis.11

Figure 1

MANAGEMENT OF ACUTE STROKE BY HOSPITALISTS

Stroke care is a rapidly evolving field in which expeditious and careful inpatient care significantly affects outcome. Hospitalists are in a unique position to improve acute stroke care and initiate secondary stroke prevention in several ways. First, there is a shortage of neurologists to care for patients with stroke. In one survey of Medicare data from 1991, prior to the widespread presence of hospitalists, only 1 in 9 stroke patients (11%) had a neurologist as the attending physician.12 At that time, there were only 3.25 nonfederal patient care neurologists per 100,000 population. Although the ratio may have improved somewhat in the intervening years (there were an estimated 5.3 self‐reported neurologists per 100,000 population as of 2005),13 the limited number of neurologists combined with the increasing incidence of stroke is expected to reduce the fraction of stroke patients having a neurologist involved in their care. Because neurology practices tend to be concentrated in urban areas, the shortage is likely to affect nonurban areas to a greater degree. The number of hospitalists, currently estimated to be 20,000 in the United States, is projected to reach 30,000 by 2010.14 In the simplest terms, hospitalists are the logical choice to fill the need for physicians to manage inpatient stroke.

Perhaps the most compelling reason for hospitalists to be involved in the care of stroke patients is clinical: patients with stroke frequently have multiple comorbid conditions that affect outcomes and are not within the traditional purview of neurology. A retrospective analysis of data from 1802 patients seen in a geriatric practice revealed that 56% of patients with stroke also had coronary artery disease, and 28% had peripheral arterial disease.15 In addition, the major risk factors for strokediabetes and hypertensionwould be expected to be prevalent in this population. Timely and effective management can improve secondary stroke prevention as well as prevent exacerbation of existing conditions.

A recent report compared outcomes in 44,099 patients following stroke according to physician specialty.16 Although patients treated by neurologists alone had a 10% lower risk of 30‐day mortality compared with those treated by generalists (family practice physicians, general practitioners, or internists) despite having more severe stroke, collaborative care reduced that risk an additional 6%.16 The risk of rehospitalization for infections and aspiration pneumonia within 30 days was 12% lower for those treated by neurologists. However, these patients had a significant, 17% increased relative risk of rehospitalization for coronary heart disease (95% confidence interval [CI], 1.021.34).16

Comanagement of stroke patients by hospitalists and neurologists is likely to become more common over time, as proposed by Likosky and Amin.17 Although studies have not specifically compared outcomes in patients with stroke who have been treated by hospitalists versus other types of physicians, implementation of hospitalist services has been associated with improved short‐term mortality and rehospitalization rates compared with traditional care.1820 Approximately 85% of hospitalists are trained in internal medicine.21 In addition, they have skill sets focusing on the specialized needs of inpatients. As hospitalists assume a greater role in the management of stroke, research into the benefits of collaborative care can be explored.

Finally, hospitalists are ideally positioned to champion the use of standardized protocols for secondary stroke prevention at their institutions. Results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke Registry showed that a minority of acute stroke patients are treated according to established guidelines.22 The 4 prototype registries were in Georgia, Massachusetts, Michigan, and Ohio. The percentage of relevant patient populations that had lipid profiles assessed ranged between 28% and 34%. For smoking‐cessation education, the range was between 17% and 34%. Anticoagulant prescribing for relevant populations at discharge ranged from 64% to 90%, and antithrombotic prescribing ranged from 88% to 98%.22

The use of protocols that initiate secondary prevention of cerebrovascular and cardiovascular events has been demonstrated to improve patient adherence to evidence‐based treatment after discharge.2328 The Preventing Recurrence of Thromboembolic Events Through Coordinated Treatment (PROTECT) program was designed to integrate secondary stroke prevention measures into the standard stroke care provided during acute hospitalization (Table 1).26 Use of appropriate antithrombotic medication was achieved in 100% of cases. Use of statins, angiotensin‐converting enzyme inhibitors or angiotensin receptor blockers, and thiazide diuretics improved significantly during the first year of implementation (P < .001). Patient education in all 4 of the areas established was carried out in 100% of patients prior to discharge.26 Tools for establishing similar hospital‐based secondary prevention programs are presently available from the University of California at Los Angeles PROTECT Program and other programs.

An essential part of any effort to develop standardized treatment procedures must include a plan to minimize any discontinuity of care after discharge. Standardized procedures need to be implemented to ensure communication of discharge summaries to outpatient clinicians in a timely and complete fashion. Only 19% of 226 outpatient physicians responding to a recent survey were satisfied or very satisfied with the timeliness of discharge summaries they received for their patients.29 Approximately one third of respondents reported that most of their patients (60%) were seen for their follow‐up outpatient visit before discharge summaries had been received. Only about one third (32%) of the respondents were satisfied or very satisfied with the summary content. Forty‐one percent believed that at least 1 of their patients hospitalized in the previous 6 months had experienced an adverse event that could have been prevented with improved transfer of discharge information.29

Development of electronic discharge summaries is an obvious alternative to conventional paper versions. This area has received less attention than others that more directly affect patient care. As the primary inpatient physicians, hospitalists can effectively implement improvements in communication among hospital staff and outpatient health care providers.

SUMMARY

This supplement is a call to action for hospitalists based on a roundtable discussion conducted in March 2007. Participants included hospitalists, neurohospitalists, vascular neurologists, and neurointensivists. The objectives of the meeting were to review the clinical data supporting current practice guidelines for secondary prevention of noncardioemboic ischemic stroke, to develop best‐practice recommendations for hospitalist‐based care of stroke inpatients, and finally to recommend improvements in transfer of information to outpatient health care providers.

The consensus of the participants is reported in the following 3 articles. The first, Evidence‐based Medicine: Review of Guidelines and Trials in Prevention of Secondary Stroke, includes an overview of the pathophysiology of stroke and TIA and reviews the clinical data supporting current treatment guidelines. Several case studies illustrating challenging or difficult aspects of secondary stroke prevention are presented in the second article, Secondary Prevention of Ischemic Stroke: Challenging Patient Scenarios. These cases focus on commonly encountered difficulties for which there may not be clear evidence or consensus. In the final article, Systems Approach to Standardization of Care in the Secondary Prevention of Noncardioembolic Ischemic Stroke, the best‐practices recommendations developed at the roundtable are presented. The role of the hospitalist in long‐term prevention strategies and the effective transfer of care to outpatient providers are discussed.

As the hospitalist movement grows, hospital‐based physicians need to identify opportunities to use their unique skills. By taking the lead in improving processes that result in better patient outcomes, hospitalists can ensure that the value of this nascent field will continue to gain recognition in the broader, sometimes skeptical medical community. We sincerely hope that you agree that integrating secondary prevention into inpatient acute stroke care is just such an opportunity. Furthermore, we hope the information we have provided will be useful to you in your hospital‐based practice.