EFNS TASK FORCE
lead to an increase in the health burden of stroke with (American Academy of Neurology, 1996). The WHO/Eu-
consequent increases in economic costs. Lifetime costs ropean Stroke Council “Helsingborg Declaration” (World
of first-ever stroke are estimated at between US$40,000 Health Organization/European Stroke Council, 1996) ad-
in the Netherlands and US$80,000 in Sweden, of which dressed the need to develop specific acute care and reha-
hospitalcosts accountfor45% in the first year after a stroke bilitation facilities in an organized fashion in order to re-
(Asplund et al., 1993; Bergman et al. , 1995); it is esti- duce mortality and disability. The present guidelines are
mated that hospital costs will increase increase by 1.5% intended to focus on the role of neurological services and
per year (Bergman et al., 1995). Across Europe, there- neurologists in the acute care of people with stroke, and
fore, with its ageing population, there is a pressing need supplement already existing reports and guidelines.
to cope with this increase, and make stroke prevention
and treatment a priority to reduce the growing health bur- NEUROLOGISTS MUST COPE WITH STROKE
den and lessen its socio-economic impact. EMERGENCIES ON A DAILY BASIS
Improving acute stroke care along established and evi-
dence-based principles is a substantial undertaking. The Neurologists have played a major part in experimental and
World Health Organization (WHO) Monitoring in Car- clinical research into therapy of cerebrovasculardiseases.
diovascular Disease (MONICA) project (Thorvaldsen et Harrison et al. (197 1) first suggested that aspirin might
al., 1995) has demonstrated a large variation between reduce the frequency of transient ischaemicattacks. These
countries in case fatality rates (the proportion of fatalities observations opened the way for the first of many large
occurring within 28 days after onset of acute stroke), rang- randomized trials of aspirin in both secondary prevention
ing from 15% in northern countries to 50% in some East- (Canadian Cooperative Study Croup, 1978) and, more
ern European states. The implications of these findings recently, acute intervention trials in ischaemic stroke (In-
are that the quality of acute stroke care varies between ternational Stroke Trial Collaborative Group, 1997; Chi-
countries and that an improvement in initial diagnosis, nese Acute Stroke Trail Collaborative Group, 1997).
treatment and rehabilitation programmes will reduce case Data from both animal and human stroke research sug-
fatality rates. gest that the earlierintervention takes place the more likely
it is that the outcome will be favourable. Cerebral tissue
EFNS TASK FORCE ON STROKE AND EXISTING may remain viable for up to 18-24 h after acute ischae-
GUIDELINES mia, even if cerebral blood flow is below 22 ml/100 g per
min, provided there is a high level of oxygen extraction
National neurological societieshave become increasingly and a high cerebral metabolic rate for oxygen (Baron et
aware of the need to address the problem of acute stroke al., 1995). Thrombolyis is now recommended in North
care through consensus statements and training pro- America if undertaken within 3 h of the onset of stroke,
grammes. In particular, there is a need to set up special- and further clinical trials testing the efficacy of throm-
ized treatment facilities for the acute care of patients with bolysis are under way with a time window of up to 6 h.
stroke. In 1995 the European Federation of Neurological Neuroprotectiontrials presently test a therapeutictime win-
Societies (EFNS) made stroke treatment and prevention a dow up to 12 h.
major policy issue. A Task Force on Neurological Acute Therefore, neurologists involved in the acute care of
Stroke Care was instructed to develop guidelines for the stroke must now operate within organizational structures
implementation of stroke management policies for use by that allow them to react quickly, as therapeutic measures
neurologists in all European countries with the intention undertaken outside the appropriate time windows are less
that these guidelines could then be modified according to likely to be effective and may be harmful. The fact that
local and national requirements. neurologistsmust now cope with stroke emergencies on a
Other guidelines have focused on general issues of daily basis represents a major turning point in clinical
stroke care, public awareness and education programmes neurology.
and treatment, rehabilitation and organizational issues. An
ad hoc consensus group (European Ad Hoc Consensus TREATMENT DELAYS BOTH OUTSIDE AND
Group, 1996, 1997) has compiled all available evidence INSIDE THE HOSPITAL CAN BE AVOIDED
about early intervention in acute stroke and laid down the
principles of intensive care. A report from the American Various factors are responsible for delay in patient refer-
Heart Association has developed guidelines for the man- ral to hospital. These include a poor awareness of the stroke
agement of patients with acute ischaemic stroke (Adams by the victim or family, reluctance to seek emergency or
et al., 1994). This has been supplemented by a practice medical help, incorrect diagnosisby the paramedical serv-
advisory document on thrombolytic therapy for acute is- ice and rating stroke as a non-emergency by medical per-
chaemic stroke by the American Academy of Neurology sonnel and the family physician. In Germany, only 5% of
436 European Journal of Neurology . Vo1 4 . 1997
NEUROLOGICALACUTE STROKE CARE
the population are aware of the warning signs of stroke and CT criteria proved to be correct in 62% of patients
compared with 50% who know about myocardial ischae- compared with a final diagnosis 3 months later for all
mia (A. Kottmaier, unpublished data). Jørgenson et al. stroke types (Madden et al., 1995).Even then, 15% could
(1996) showed that patients who had previously suffered not be categorized. In another study, lacunar stroke was
a transient ischaemic attack underwent shorter admission correctly distinguished from non-lacunar stroke in 56%.
times when they experienced a stroke in comparison to of cases within 12h after onset in 517patients presenting
patients with first-ever strokes not known to’have had a with first-ever stroke (Toni et al., 1995).The effect of this
transient ischaemic attack. Ambulance staff are known to on prognosis was demonstrated by Libman et al. (1995)
make an incorrect diagnosis in 50% of patients compared who showed that pure motor hemiparesis tended to im-
with 25% by paramedics (Kothari et al., 1995). prove rapidly over 10 days. These studies show that even
Because of these factors, traditional settings for the for expert neurologists, the early diagnosis of stroke
treatment of patients with acute stroke may not be appro- subtype is difficult on clinical grounds alone but may be-
priate. Neurologists should be involved in the public edu- come more reliable when the results of further examina-
cation and training of paramedical staff, stress the impor- tions, in addition to the initial CT, are readily available.
tance of awareness about stroke, the early recognition of
symptoms and the need for emergency hospitalization. NEUROLOGICAL STROKE UNITS ARE NOW
Within the hospital, factors delaying early intervention WIDELY ACCEPTED
include admission polices that require placement of pa- The use of neurological stroke units within the first week
tients on general medical wards, lack of access to early has yet to be evaluated in formal trials. Most of the evi-
brain-imaging facilities,the rating of stroke as non-urgent dence for the effectiveness of organized stroke care has
by hospital staff,non-existenttreatment facilities for stroke come from stroke rehabilitation.
at the point of admission and unavailability of a neurolo-
gist in the emergency room. This may lead to delays in Level I evidence
the early recognition of stroke, prolonged ‘door to drug’ Randomized studies have shown that stroke units are ef-
times, failure to use new treatment modalities, including fective in reducing mortality compared with treating pa-
thrombolysis, and the introductionof early secondary pre- tients on general medical ward. A meta-analysisof all tri-
vention measure. Conditions other than stroke that pres- als of stroke units compared with care in a general medical
ent stroke-likesyndromes,such as hypoglycaemiaor sub- setting showed a 21% reduction in mortality after 12
dural haematoma,may be life-threatening,and non ischae- months (Langhorne et al., 1993).Individualtrials include
mic stroke may account for up to 25% of acute stroke a prospective randomized study of acute stroke patients
admissions(Bratina et al., 1995).Delay in imaging is com- in which there was an improvement in functionaloutcome
mon (Anderson et al., 1995):despite a median arrivaltime and a reduction in the need for long-term care in patients
of 4.3 h for ischaemic stroke and 2 h for intracerebral or who randomly allocated to a non- intensive stroke unit
subarachnoid haemorrhage, computed tomography (CT) compared with a general medical ward (Strand et al.,
was delayed by a median of 66 h for the former compared 1985). In this study, 15% of those randomized to stroke
with a few hours by the intervention of the emergency unit care and 39% of those on a general ward remained
stroke team (Bratina et al., 1995). hospitalized. Functional outcome was better, as measured
by the Activities of Daily Living scale, including meas-
EARLY ACCURATE DIAGNOSIS CAN BE urements in the domainsof walking, personal hygiene and
ACHIEVED dressing. The precise mechanism for this remains unclear,
though the authors suggest that teamwork and early onset
Different stroke subtypes have different aetiologies, dif- of rehabilitation may be important. Kaste et al. (1995)
fering frequencies with which complications and associ- showed that patients randomly allocated to care in the neu-
ated comorbidities occur, differing rates of early recur- rology department left hospital earlier, went home directly
rence and varying rates and patterns of deterioration and more often, and were more commonly independent after
prognosis; therefore accurate diagnosis is required for 1year than those treated on medical wards. While the fa-
appropriate management. tality rate was no different,the authors suggested that bet-
A history and physical examination may assist in the ter organizationof care played a part in the improved out-
classificationof the stroke subtype and allow priorities to come.
be determinedfor additional investigations and treatments. Demographic variables are well matched in most stud-
Subtypes include stroke due to arterial atherothrombosis, ies, but data about the time of admission and the onset
cardiac embolism, small-vessel occlusion and and intensity of rehabilitation measures are usually lack-
crypotogenicand uncertain aetiologies. In a recent study, ing in these studies, so that no general conclusions can be
early categorization of the initial stroke type on clinical drawn about the efficacy of acute stroke units.
European Journal of Neurology .Vo1 4 . 1997 437
EFNS TASK FORCE
Level II evidence with internists and general practitioners. This mirrors
Two studies have compared the efficacy of stroke reha- Medicare costs for myocardial infarction as cardiologists
bilitation units with historical controls treated on general have gradually taken over the care of this group of pa-
medical wards. Jørgenson et al. (1995) studied 936 acute tients.
stroke patients managed in a stroke unit of 61 beds and There are no evidence-based guidelines in the follow-
compared them with 305 patients managed in general ing areas, though clinical trials are under way in some:
wards at another hospital. Both groups were similar in the management of swallowing disorders and the preven-
terms of age and time of admissionto the unit. Case fatal- tion of aspiration using either a nasogastric tube or via
ity rates, percentage discharged to a nursing home, length percutaneous gastrostomy (Holas et al., 1994; Smithard
of stay for all patients and length of stay for those aged et al., 1996), nutritional supplements after a stroke, the
over 70years were all lower among those treated in a stroke use of intracranial monitoring and other aspects of inten-
unit. The authors (Jørgenson et al., 1995) attributed their sive care, including optimal control of hyperpyrexia
success to a standardevaluationprogramme, intensive ob- (Grotta et al., 1995; El-Ad et al., 1996;Reith et al., 1996;
servation and goal-oriented treatment in the first few days Schwab et al., 1996;European Ad Hoc Consensus Group,
after admission. 1997).
Bath et al. (1996) provided some evidence that imme- Ethical issues also need to be considered,and a rational
diate admission to a stroke unit might be efficacious. Pa- approach to ‘do not resuscitate’ orders is required, espe-
tients admitted immediately to an eight-bed stroke unit cially to assist in the management of patients who are
and transferred after an average of 7 days to postacute locked in or in vegetative states (Alecandrow et al., 1995,
treatment facilitieswere compared with historical controls 1996).
treated on a general medical or geriatric ward. The pa- Further unanswered questions about acute stroke care
tients treated in a special unit showed a reduced time to include questions about appropriatediagnostic techniques
admission from the emergency unit, earlier CT, more ca- (Mohr et al., 1995; Schneider et al., 1996), pharmaco-
rotid sonography, a greater likelihood of being given as- therapy for cerebral ischaemia (Marshall and Mohr, 1993)
pirin on discharge and a reduced in-patient hospital stay. and effectivemeans of preventing complications (Daven-
Bath et al. (1996) suggest that a better outcome might be port et al., 1966).
related to an improved diagnosis, directed rehabilitation, None of the studies reported above give a detailed analy-
fewer early complications and better early secondary pre- sis of the elements of stroke care that might be essential in
vention measures. a neurological acute stroke unit and detailed descriptions
of all the components remain to be undertaken.
Level III evidence
In a survey in 18 Spanish hospitals, the time from the on- RECOMMENDATIONS
set of a stroke or a transient ischaemic attack to contact Two models of care are proposed. One involves the use of
with a neurologistwas compared. Those admitted or evalu- dedicated facilities for acute neurological care within a
ated early had an improved outcome as measured by an hospital,the other a mobile neurological acute stroke team
impairment scale and a reduced length of stay (Davalos et when dedicated facilities are not available.
al., 1995).Intensive management of hypertensionand hy- The primary aim of either model is:
perglycaemia in the acute phase was thought to have con- accurate diagnosis to minimize complications and
tributed to this. Some evidence from North America sug- recurrence;
gests that neurologists who treat stroke are more likely to initiation of acute stroke treatment.
be knowledgeable about the causes of stroke that others, Acute care starts with the patient as soon as symptoms
and their patients have a better outcome than those treated are experienced. Patients require education about the sig-
by internists or general practitioners. Medicare evaluated nificance of their symptoms, while paramedical staff and
20% of their stroke patients treated in 1991 and found general practitioners need to know how to accessthe stroke
that patients treated by neurologists had a significantly team or stroke unit. All admissions trails should therefore
lower 90-day mortality (Mitchell et al., 1996); however be clearly visible and available to the community prehos-
these patients were not matched for stroke severity or pita1 services. Within the hospital setting neurologists
comorbid disease. The same data showed that neurolo- should have early access to patients with stroke and full
gists used more imaging procedures, including CT, mag- access to appropriate investigational facilities.
netic resonance imaging (MRI), angiography and more
prothrombin tests for anticoagulant control. They also Neurological acute stroke unit
made more use of rehabilitation facilities and made less It is important to have a stroke unit within a defined set-
use of nursing facilities and nursing homes. However, the ting. These units mostly consist of a small number of beds
neurologistsincurred high hospitalization costs compared which are exclusively and immediately available for all
438 European Journal of Neurology . Vo1 4 . 1997
NEUROLOGICAL ACUTE STROKE CARE
TABLE I. Investigative facilities required to be readily accessible
Acute stroke (24 h) for acute neurologicalstroke units
Neurosonologicalinvestigations (extracranial and intracranial
vessels, colour-coded Duplexsonography, transcranial
Hospital Echocardiography (transthoracic)
Laboratory examinations (including coagulation parameters)
General Acute Blood pressure, blood gases, blood glucose, body temperature
Medical/Geriatic Neurological Emergeny
Ward Stroke Department
TABLE II. Additional investigatory facilities recommended to be
available for patients in acute neurological stroke units
Magnetic resonance imaging/Magnetic resonance angiography
Computed tomography angiography
Stroke Rehabilitation Neurological Regional Transcranial Doppler sonography (embolus detection
Unit Ward Hospital monitoring)
FIG. 1. Flow chart showing organization of an acute neurological Monitoring of central breathing disturbances
admissions with stroke-like symptoms. Many neurologi- Acute stroke units have the primary aim of initiating
cal departments treating acute patients already have moni- stroke treatment on an emergency basis and of clarifying
toring devices available for their patients as well as neu- the causes of the stroke in order to prevent further relapses
rologists with a specialized interest and training to treat and complications. All other aspects are secondary.
acute strokes. With an increasing tendency to apply early Prehospital structures must be adapted to this primary aim.
and very early interventions, especially thrombolysis, it Within the hospital their patients must have access to all
is also evident that neurological acute stroke units will be investigations on an emergency basis. There should be a
the specialized setting for any kind of treatment which ready availability of CT, arteriography, echocardiography,
has to be performed within the first few hours after the Dopplersonography including transcranial Doppler
onset of a stroke. Initial management on a general ward sonography and transcranial Doppler monitoring, as well
will not be easily justified for treatments that have a haz- as CT angiography,MRI and MRA if situated within reach
ardous potential. (Tables 1 and 2).
A concept for neurological acute stroke care should
follow an integrated concept for treatment and include the Neurological acute stroke team
prehospital emergency setting, rapid transfer to centers Many hospitals do not yet have acute stroke units. These
with a neurological stroke unit, emergency treatment fol- are often large general hospitals with a neurological de-
lowing admission, initial neurological and neuroradiolog- partment that consists of a small number of beds provid-
ical assessment, additional investigations, secondary and ing a ward consultation service. In this type of traditional
tertiary prophylaxis, continuous neuromonitoring and ther- setting, neurologists are recommended to install mobile
apy during the acute phase, and consecutive transfer, pref- stroke teams that are rapidly available on an emergency
erably to a non-intensive-care stroke rehabilitation unit basis throughout the hospital (Fig. 2). These neurologists
(Fig. 1). should have the primary competence for all acute strokes
This acute stroke unit has the primary task of perform- as soon as they are admitted. If there is an emergency room
ing very early evaluation and management following the or emergency department for all hospital admissions, a
early days after a stroke. Other additional elements are neurologist should be easily and rapidly available for those
optional and vary from country to country. In some coun- admissions and have a chance to take a history from the
tries or regions, stroke units have been established to in- paramedics and make a first assessment. They should fol-
clude early rehabilitation as well. This is a compatible low the patients to intensive care, to coronary care, to gen-
addition to an acute stroke unit. Stroke units that exclu- eral medical or geriatric wards. After initial management,
sively focus on rehabilitation in the postacute phase fol- neurologists should be able to start with neurorehabilita-
low a different task and their efforts cannot easily be com- tion as soon as possible. Physiotherapists, occupational
pared with a neurological acute stroke unit. therapists and speech therapists should therefore also be-
European Journal of Neurology . Vo14 . 1997 439
EFNS TASK FORCE
phy should be made available within an organizational
Acute stroke structure which defines the accessibility to those teams
according to the emergency status of the patient.
Within the first few hours and days the patient will be
Paramedics monitored continuously. It is essential that a portable
Emergency physicians monitor for electrocardiography (ECG) and pulse-
oxymetry is available when the patient is moved to ex-
Medical Emergency aminations outside the acute stroke unit (Tables 1 and 2).
Ward Ward Blood pressure, ECG lood glucose, peripheral oxygen
Neurological Intensive Care saturation, body temperature and fluid balance should be
monitored continuously while the patient is in the unit. A
Unit Coronary Care
short neurological stroke scale repeated at regular inter-
Ward Unit vals in helpful to detect progression or deterioration. It is
essential to observe changes in consciousness and to rec-
ognize secondary or tertiary complication.
Stroke Rehabilitation Neurological Regional A neurological acute stroke unit should consist of four to
Unit Ward Hospital
eight beds (ideal size six beds) and aim at treating acute
FIG. 2. Flow chart showing organization of an acute neurological strokes for 2-7 days (average 4-5 days) depending on the
mobile stroke team. severity, complicationsand case-mix. With a regional in-
cidence of 200 first-ever strokes per 100000 inhabitants
long to the mobile stroke team. After some time, a suc- per year and a hospitalization rate of 70-80%, one acute
cessful mobile stroke team is often able to install an acute stroke unit may serve approximately 200,000-300,000
stroke unit with beds and neuromonitoring equipment. In inhabitants. This implies that 350-500 strokes are treated
order to justify this, the work of this team should be care- per year, including recurrent strokes. If the average pa-
fully and separately documented from the beginning, to tient stay can be reduced to 2-4 days, a stroke unit can
allow an evaluation of the clinical successes and the costs serve a larger population of approximately 300,000-
of the interventions. In addition, a checklist of interven- 400,000 inhabitant, thus treating up to 800 strokes per
tions and investigations is helpful in recording the work year.
on an individual patient. A written clinical pathway for
acute strokes is even better than a mere checklist because ORGANIZATIONAL RECOMMENDATIONS
it makes the process of clinical decision-making more
transparent. Interdisciplinary cooperation with other med- Diagnosticand therapeutic measures are standardized and
ical specialists is essential for the general management of are updated at regularintervals. Strokeprotocols and pres-
stroke. ent clinical pathways are important for a continuous as-
sessment of the quality of performance of the stroke unit.
Prehospital and acute inhospital management The first assessment includes the decision on whether the
A neurological acute stroke unit must startoutside the hos- patient has to be intubated and ventilated and therefore
pital: Emergency physicians, paramedics and referring transferred to an intensive.care unit. If no intensive care
doctors as well as the general population within the dis- unit is easily available some strokeunits might decide to
trict should know that this acute stroke unit exists and is start artificial or assisted ventilation before transferring
operational at all hours. A neurologist should be available the patient. The majority of patients, however, do not re-
24 h a day on an emergency basis in order to attend to all quire ventilation and can be referred on after stabilization,
admissions with a stroke syndrome, to perform a rapid preferably to a non-intensive-care stroke rehabilitation
neurological investigation and to check vital and cardio- unit. Physiotherapy, occupational therapy and speech
vascular parameters. He will take a short patient history therapy can be started in order to guarantee seamless care.
and decide on the location and cause of the infarct. Inev-
itably, a CT scan and a laboratory screening will be need- REFERENCES
ed on an emergency basis as well as further investigations.
Adams HP, Brott TG Crowell RM, Furlan AJ, Gomez, CR, Grotta J,
Doppler investigations of the extra- and intracranial ves- Helgason, CM, Marler JR, Woolson RF, Zivin JA, Feinberg W
sels, echocardiographic investigations of the heart and and Mayberg M (1994 ) Guidelines for the management of pa-
aortic arch and the option to perform cerebral angiogra- tients with acute ischemic stroke. Stroke, 25,1901-1914.
440 European Journal of Neurology . Vo1 4 . 1997
Alexandrow AV, Bladin CF, Meslin EM and Norris JW (1995) Jørgensen HS, Nakayama H, Raaschou HO, Larsen K, Hübbe P
Do-not-resuscitateorders in acute stroke.Neurology, 45, 634- and Olsen TS (1995) The effect of a stroke unit: reductions
640. in mortality, discharge rate to nursing home, length of hospi-
Alexandrow AV, Pullicino PM, MedlinEM and Norris JW (1995) tal stay,and cost.A community-basedstudy.Stroke, 26, 1178-
Agreement on disease specific criteria for do-not-resuscitate 1182.
orders in acute stroke. Stroke, 27, 232-237. Jørgensen HS, Nakayama H, Reith J, Raaschou HO and Olsen
Anderson NE, Broad JB and Bonita R (1995) Delays in hospital TS (1996) Factors delaying hospital admission in acute stroke:
admission and investigation in acute stroke. British Medical the Copenhagen Stroke Study. Neurology, 47, 383-387.
Journal, 311, 162. KasteM, Palomäki H and Sarna S (1995)Where and how should
Asplund K (1996) Stroke in Europe: widening gap between East elderly stroke patients be treated? Stroke, 26, 249-253.
and West. Cerebrovascular Disease, 6, 3-6. Khaw K (1996) Epidemiology of stroke. Journal of Neurology,
Asplund K, Marke L, Terent A, Gustafsson S and Wester PO Neurosurgery and Psychiatry, 61, 333-338.
(1993) Costs and gains in stroke prevention: European per- Kothari R, Barsan W, Brott T, BroderickJ and Asjbrook S (1995)
spective. Cerebrovascular Disease, 3 (suppl l), 34-42. Frequency and accuracy of prehospital diagnosis of stroke.
Baron JC, von Kummer R and del Zoppo GJ (1995) Treatment Stroke, 26, 937-941.
of acute ischemic stroke: challenging the concept of a rigid Langhorne PO, Williams B, Gilchrist W and Howie K (1993)
and universal time window. Stroke, 26, 2219-2221. Do stroke units save lives? Lancet, 342, 395-398.
Bath PMW, Soo J, Butterworth RJ and Herr JE (1996) Do acute Libman RB, Sacco RL, Shi T, Tatemichi TK and Mohr JP (1992)
strokeunits improve care? Cerebrovascular Disease, 6, 346- Neurologic improvement in pure motor hemiparesis: impli-
349. cations for clinical trials. Neurology, 42, 1713-1716.
Bergman L, Van der Meulen JHP, Limburg M and Habbema Madden KP, Karanjia PN, Adams HP, ClarkeWR and the TOAST
JDF (1995) Costs of medical care after first-ever stroke in investigations (1995) Accuracy of initial stroke subtype di-
the Netherlands. Stroke, 26, 1830-1836. agnosis in the TOAST study. Neurology, 45, 1975-1979.
Bowen J and Yaste C (1994) Effect of a stroke protocol on hos- Marshall RS and Mohr JP (1993) Current management of
pital costs of stroke patients. Neurology, 44, 1961-1964. ischemic stroke. Journal of Neurology, Neurosurgery and
Bratina P, Greenberg L, Pasteur W and Grotta JC (1995) Cur- Psychiatry, 56, 6-16.
rent emergency department management of stroke in Hou- Mitchell JB, Ballard DJ, Whisnant JP, Ammering CJ, Samsa GP
ston, Texas. Stroke, 26, 409-414. and Matchar DB (1996) What role do neurologists play in
Canadian Cooperative Study Group (1978) A randomized trial determiningthe costs and outcomes of stroke patients? Stroke,
of aspirin and sulfapyrazon in threatening stroke. New Eng- 27, 1937-1943.
land Journal of Medicine, 299, 53-55. Mohr JP, Biller J, Hilal SK, Yum WTC, Tatemichi TK, Hedges
Chinese Acute Stroke Trial Collaborative Group (1997) CAST: S, Tall E, Nguyen H, Mun I, Adams HP, Grimsman K and
a randomized placebo-controlled trial of early aspirin use in Marler JR (1995) Magnetic resonance versus computed to-
20,000 patients with acuteischemicstroke. Lancet 349, 1641- mographic imaging in acute stroke. Stroke, 26, 807-812.
1649. Reith J, Jørgensen HS, Pedersen PM, Nakayama H, Raaschov
Davalos A, Castillo J and Martinez-VilaE (1995) Delay in neu- HO, Jeppesen LL and Olsen TS (1 996) Body temperature in
rological attention and stroke outcome. Stroke, 26, 2233- acute stroke: relation to stroke severity, infarcts size, mortal-
2237. ity, and outcome. Lancet, 347, 422-425.
Davenport RJ, Dennis MS, Wellwood I and Warlow CP (1996) American Academy of Neurology (1996) Practice advisory:
Complications after acute stroke. Stroke, 27,415-420. Thrombolytic therapy for acute ischemic stroke. Summary
El-Ad B, Bornstein NM, Fuchs P and Korczyn AD (1996) Me- statement. A report of the Quality Standards Subcommittee.
chanical ventilation in stroke patients: is it worthwhile?Neu- Neurology, 47, 835-8 39.
rology, 47, 657-659. Schneider LB, Libman RB and Kanner R (1996) Utility of re-
EuropeanAd Hoc Consensus Group (1 996) European strategies peat brain imaging in stroke. American Journal of Neurora-
for early intervention in stroke. Cerebrovascular Disease, diology, 17, 1259-1263.
6, 315-324. Schwab S, Aschoff A, SprangerM, Albert F and Hacke W (1996)
European Ad Hoc Consensus Group (1997) Optimizing inten- The value of intracranial pressure monitoring in acute hemi-
sive care in stroke: a European perspective. Cerebrovascular spheric stroke. Neurology, 47, 393-398.
Disease, 7, 113-128. Smithard DG, O’Neill PA, Park C, Morris J, Wyatt R, England
Fogelholm R, Murros K, Rissanen A and Ilmavirta M (1996) R and Martin DF (1995) Complications and outcome after
Factors delaying hospital admission after acute stroke. Stroke, acute stroke: does dysphagia matter? Stroke, 27, 1200-1204.
27, 398-400. Strand T, Asplund K, Eriksson S, Hagg E, Lithner F and Wester
Grotta J, Pasteur W, Khwaja G, Hamel T, Fisher M and Ramirez PO (1985) A non-intensive stroke unit reduces functional
A (1995) Elective intubation for neurological deterioration disability and the need for long term hospitalization. Stroke,
after stroke. Neurology, 45, 640-644. 16, 29-34.
Harrison MJG, Marshall J, Meadows JC and Ross Russell RW Thorvaldsen PM Asplund K, Kuulasmaa K, Rajakangas A-M
(1971) Effect of aspirin in amaurosis fugax. Lancet, ii, 743- and Schroll M (1995) Stroke incidence, case fatality, and
744. mortality in the WHO MONICA project. Stroke, 26, 361-
Holas MA, DePippo KL and Reding MJ (1994) Aspiration and 367.
relative risk of medical complications following stroke. Ar- Toni D, Fiorelli M, De Michele M, Bastianello S, Sachetti ML,
chives of Neurology, 51, 1051-1053. Montinaro E, Zanette EM and Argentino C (1995) Clinical
International Stroke Trial Collaborative Group (1997) The In- and prognostic correlates of stroke subtype misdiagnosis
ternational Stroke Trial (IST): a randomized trial of aspirin, within 12 hours from onset. Stroke, 26, 1837-1840.
subcutaneousheparin, both or neither among 19 435 patients World Health Organization/European Stroke Council (1997) Pan
with acute ischaemic stroke. Lancet 349, 1569-1581. European consensus meeting on stroke management. (in press).
European Journal of Neurology . Vol 4 . 1997 441