Vous êtes sur la page 1sur 10

Original Paper

Received: April 2, 2003

Eur Neurol 2003;50:207214
Accepted: June 26, 2003
DOI: 10.1159/000073861

Impact of Emergency Room

Neurologists on Patient Management
and Outcome
Thierry Moulin a Denis Sablot a Elisabeth Vidry a Faouzi Belahsen a
Eric Berger a Patrick Lemounaud b Laurent Tatu a Fabrice Vuillier a
Anne Cosson a Eugeniu Revenco a Gilles Capellier b Lucien Rumbach a
a Department of Neurology and b Emergency Care Unit, University Hospital, Besanon, France

Key Words (33.1%), epilepsy (20%), loss of consciousness (9%),

Emergency room care W Epilepsy W Neurological headaches (9%), confusion (5.4%), peripheral nervous
examination W Stroke outcome system disorders (4.4%), vertigo (4.2%), cognitive dys-
functions (4%), gait disorders (3.2%) and miscellaneous
(7.1%). Overall, false positive or negative diagnoses
Abstract were produced by the ER in 37.3 and 36.6% of ER admis-
The frequency and impact of in-patient assessment by a sions, respectively. A complete change of diagnosis by
neurologist in the emergency room (ER) setting remain the neurologist was found in 52.5% of patients. Of the
largely underestimated. The objective of our study was patients undergoing a neurological examination, 18.4%
to analyse the impact of neurologist in-patient manage- were able to go home, 31.8% were admitted to the stroke
ment. Methods: Over a period of 12 months, we prospec- unit, 32.4% to the general neurology unit and 17.4% to
tively recorded the demographics of patients requiring other departments. Conclusion: Our study stresses the
examination in the ER, the ER teams tentative neurologi- need for a neurologist in the ER, both in quantitative
cal diagnosis, the neurology teams final diagnosis and terms and for the benefit of patient management.
patient outcomes. The time interval between admission, Copyright 2003 S. Karger AG, Basel

call for a neurologist and the assessment by the neurolo-

gist were recorded. Results: Assessments by neurolo-
gists were performed in 14.7% (1,679/11,421) of all pa- Introduction
tients admitted to the ER. The mean time between
admission and examination was 32 (B 36) min, irrespec- While out-patient management for chronic neurologi-
tive of the day of the week, and dependent on the tenta- cal disorders is now well established for most of the main
tive diagnosis: shorter for stroke and status epilepticus neurological diseases, the impact of a neurological in-
(p ! 0.05) and longer for confusion and vertigo (p ! patient assessment on diagnosis, patient management and
0.05). The initial causes for examination were: stroke treatment in the emergency room (ER) remains largely

2003 S. Karger AG, Basel Thierry Moulin, MD, PhD

ABC 00143022/03/05040207$19.50/0 Service de Neurologie
Fax + 41 61 306 12 34 CHU Jean Minjoz
E-Mail karger@karger.ch Accessible online at: FR25030 Besanon (France)
www.karger.com www.karger.com/ene Tel. +33 3 81 66 82 36, Fax +33 3 81 66 84 70, E-Mail thierry.moulin@univ-fcomte.fr
underestimated [1, 2]. Neurological emergencies are de- or revised by the clinical findings and, if necessary, by neuroimaging
fined by certain criteria, including rapid onset, immediate or neuroelectrophysiological examinations (CT, MRI, EEG, EMG,
lumbar puncture). These neurological examinations could not have
poor prognosis, and the requirement for prompt resuscita-
been previously conducted by the ER team. For the tentative diagno-
tion. These criteria apply to standard neurological emer- sis, the analyses of these neurological examinations by the ER team
gencies, i.e. generalised status epilepticus (SE) and respi- were always taken into account when they were performed before the
ratory failure due to neurological diseases. However, re- neurological assessment. A formal synthesis of all available data was
cent advances in neuroimaging techniques and new ther- done in collaboration with the senior neurologist on duty and
approved by the emergency team before a patient was discharged
apeutic imperatives make it necessary to establish prog-
from the ER.
nostic factors as soon as possible in order to select the From January 1st to December 31st, 1997, we conducted a pro-
right treatment for any given patient. Furthermore, the spective study of the neurological examinations performed in re-
concept of emergency neurology has evolved to allow bet- sponse to calls from the ER. Only patients for who assessment by a
ter preservation of neurological functions by saving as neurologist was requested by the ER were included in this study; ER
practitioners were naive of this study. Whereas we did not pre-define
much cerebral tissue as possible. The role of the neurolo-
the lists of neurological symptoms and signs, the procedures for neu-
gist in the ER is therefore crucial in defining, confirming rology patient management (exams and treatment) followed com-
and implementing patient management. Stroke diagnosis monly used flow charts and diagnostic algorithms. For each patient,
has already been shown to exemplify the necessity for spe- we analysed the demographic data, the time interval between admis-
cialised management [3]. Other recent reports emphasis- sion to the ER and the call to the neurologist (emergency reaction
time) and between the emergency call and the assessment by a neu-
ing the specific role of the neurologist in modifying the
rologist (neurological response time), tentative and final diagnoses,
tentative diagnosis in neurological emergencies highlight and patient outcome. Analyses of timing, type, and amount of gener-
the apparent complexity of a neurological examination al (i.e., ECG, blood tests, etc.) or neurological examinations were not
compared to a cardiological investigation [4]. specifically performed. The tentative and final diagnoses were di-
The aim of our study was to evaluate the need for neu- vided into groups according to symptom or syndrome (e.g. headache,
vertigo, loss of consciousness, focal deficits, gait disorders) or, when
rologists in an ER and to determine the impact their pres-
possible, into nosologic groups (e.g. stroke, epilepsy, multiple sclero-
ence has on patient management for every specific neuro- sis). For the nosologic groups, classic neurological criteria were used,
logical symptom and sign. such as clinical findings and neuroimaging for stroke, tumour, etc.
Final diagnosis was determined again by the neurology team in daily
case review meetings during which the impact of the neurologists
Patients and Methods act(s) was evaluated.
The neurological impact was analysed in two complementary
The University Hospital of Besanon (1,200 beds) serves as a pri- ways. Firstly, it constitutes what we called the general neurological
mary care hospital for a population of about 200,000 people. The ER contributive effect. It was a general and subjective evaluation of neu-
is the only unit to manage medical and surgical emergencies in rologists actions in the ER, considered as (1) just confirmative when
adults. The Neurology Department is divided into different units, no modification in diagnosis or management was proposed to the ER
including a stroke unit. With the progressive development of this physicians; (2) additional when the tentative diagnosis was con-
stroke unit since its creation in 1987, the neurology team 4 resi- firmed in the same nosologic group, allowed a more accurate analysis
dents and 4 senior physicians has structured itself to meet emergen- of neurological examinations, additional patient management or
cy calls 24 h a day: there is always a resident on duty and a senior even when neurologists specifically asked for new neurological exam-
practitioner on call. The shifts for duty are 06.3018.30 h and 18.30 inations, and (3) complete when the tentative diagnosis was totally
06.30 h. The emergency department has 10 senior physicians certi- changed, e.g. syndromes or symptoms in a clear nosologic group
fied for ER care and 6 residents trained to identify vital failures. The (from dysarthria, confusion to stroke or from syncope to seizure, ...)
normal admission procedure is to evaluate the degree of vital failure or in a different group (i.e. from epilepsy to stroke, from transient
in every patient, to prescribe the appropriate examinations and ther- ischemic attack to ischemic stroke), or when it resulted in active ther-
apy according to the diagnostic algorithms and flow charts usually apeutic measures that potentially modified patients outcome, such
applied in emergency medicine, and to ensure optimum dispatch- as antiepileptic drugs, anticoagulants or decision for surgery. Sec-
ing. ondly, with the same amount of information (clinical and neurologi-
For each neurological problem detected in a patient admitted to cal examinations), we grouped the quality of the ER tentative diagno-
the ER, the neurology resident on duty was called and informed of sis into 3 categories: correct diagnosis, false positive or false negative
the ERs tentative diagnosis, the kind and intensity of symptoms or diagnosis. All data were prospectively coded using a pre-specified
signs (i.e., onset and course, as well as general demographics). These form after each period of duty.
calls were systematically made after patients had been examined by Statistical analyses were performed using the statistic software
an ER resident and an ER practitioner. The ER physicians were also SPSS. Data were considered significantly different if p ! 0.05.
asked to give severity criteria based on instruments such as the Glas-
gow Scale (for loss of consciousness) and pre-specified guidelines
(complete or partial) for motor deficits. Following examination by
the resident and/or senior neurologist, the diagnosis was confirmed

208 Eur Neurol 2003;50:207214 Moulin et al.

Fig. 1. Comparison between the different
response times, periods of the day and days
of the week. a Significant differences for the
different types of response times: shorter
neurological response times after 18.30 h
(a 27.2 vs. 36.1 min, p ! 0.01) and longer
emergency response times after 18.30 h
(b 67.5 vs. 81.1 min, p ! 0.01). b Constantly
and significantly shorter neurological re-
sponse times after 18.30 h throughout the
week. * p ! 0.05.

Results sciousness (87.5 min, p ! 0.05), vertigo (90.5 min, p !

0.01) and confusion (110.5 min, p ! 0.01). The neurologi-
Evaluation of Neurological Acts in the Emergency cal response time was 32.2 B 36.1 min irrespective of the
Room day of the week. Significant differences in reaction times
A neurologist was called to examine 1,679 of the were observed between the different periods of duty (be-
11,421 patients admitted to the ER (14.7%). Fifty-two tween 06.30 and 18.30 h and after 18.30 h), as shown in
percent of the examined patients were male, the mean age figure 1.
was 56.9 B 21 years.
Almost 10% (9.4%) of the neurological examinations Contribution of the Neurologist in the Emergency
took place at night (00.006.30 h), 18.3% in the morning Room
(6.3012.00 h), 38% in the afternoon (12.3018.30 h) and Global Evaluation
34.3% in the evening (18.3024 h). The emergency reac- The main tentative diagnoses are shown in table 1.
tion time (73.4 B 70.4 min) depended on the tentative Examination by a neurologist led to a positive contribu-
diagnosis; it was shorter for stroke (66.9 min, p ! 0.03) tion in 87.6% of the patients over all (additional in 35.1%
and SE (68.1 min, p ! 0.05), and longer for loss of con- and complete in 52.5%). The positive effect of the neurol-

Impact of Emergency Room Neurologists Eur Neurol 2003;50:207214 209

on Patient Management and Outcome
Table 1. Distribution of the tentative and final diagnoses and results of the neurological contributive effect according
to the different final diagnoses

Tentative Final Global neurological

diagnoses diagnoses contributive effect, %
(n = 1,679), % (n = 1,679), % confirmative/additional/complete

Stroke 556 (33.1) 567 (33.8) 13.6/48.8/37.6

Epileptic seizures 336 (20) 349 (20.8) 8.3/69.9/21.8
Headaches 151 (9) 117 (7) 8.5/63.2/28.2
Loss of consciousness 151 (9) 94 (5.7) 4.3/58.5/37.2
Confusion 90 (5.4) 95 (5.7) 1.1/26.3/72.6
Peripheral neurological disorders 74 (4.4) 72 (4.3) 56.9/11.1/45.8
Dizziness-vertigo 71 (4.3) 21(1.3) 0/95.2/4.8
Cognitive dysfunctions 67 (4) 77 (4.6) 32.5/11.7/55.8
Gait disorders 39 (2.2) 37 (2.2) 21.6/54.0/24.4
Multiple sclerosis 39 ( 2.3) 43 (2.6) 25.6/58.1/16.3
Neurological coma 27 (1.6) 5 (0.3) 0/20.0/80.0
Infectious diseases 32 (1.9) 24 (1.4) 29.2/4.1/66.8
Psychiatric disorders 21 (1.3) 43 (2.6) 12.6/11.6/76.8
Subdural hematoma 10 (0.7) 22 (1.3) 0/13.6/86.4
Brain tumour 1 (0.1) 36 (2.1) 0/14.3/85.7
No neurological disturbances 0 77 (4.6) 0/0/100
Miscellaneous 2 (0.1) 0

ogist was strongest when stroke, confusion or coma were been made prior to examination by the neurologist in 124
suspected (86.4, 98.9, 100%, respectively), as shown in of the 556 (22.3%) patients with this tentative diagnosis.
table 1. It was poorest when tentative ER diagnoses were The main final diagnoses for these 124 patients were: epi-
very precise, as was the case with migraine in 20% (1 of 5 lepsy in 20.1% (25 patients), tumour in 12.9% (n = 16),
affected patients), global transient amnesia in 28% (4/14 cognitive dysfunction in 14.5% (n = 18), and confusion in
patients), dementia in 26.3% (5/19 patients) or SE in 50% 10.4% (n = 13). A false negative diagnosis of stroke had
(3/6 patients). been made in 23.8% (135 patients). The tentative diag-
False positive or false negative diagnoses were pro- noses in these cases were: vertigo in 22.2% (30 patients),
duced by the ER team in 37.3 and 36.6% of all patients, cognitive dysfunction in 16.2% (n = 22), peripheral ner-
respectively; and their tentative diagnosis was correct in vous system disorder in 14.1% (n = 19), epilepsy in 12.5%
only 26.1%. The different percentages of false negative, (n = 17), confusion in 11.1% (n = 15) and isolated head-
false positive or correct diagnoses for specific symptoms, aches in 8.9% (n = 12).
syndromes or nosologic groups are shown in figure 2. Epileptic Seizures. Epileptic seizure was the final diag-
Only 64.2% (1,078/1,679) of the patients were admit- nosis in 349 patients. Sixty-three patients (18.7%) had
ted to a neurology unit following assessment by a neurolo- had a false positive tentative diagnosis of epilepsy, of
gist: 31.8% to the stroke unit and 32.4% to the general which 44.4% (28 patients) had a final diagnosis of confu-
neurology unit. Of the remaining 35.8% (601/1,679 pa- sion, 19% (n = 12) syncope, 11.1% (n = 7) stroke and
tients), 17.4% were directed to other departments, includ- 11.1% (n = 7) a tumour. Seventy-six patients (21.8%) had
ing 5% to the intensive care unit and 2.5% to neurosurge- had a false negative diagnosis of epilepsy. The tentative
ry, while 18.4% were able to go home. Of the patients dis- diagnosis was syncope in 50% (n = 38), stroke in 32.9%
charged (309/1,679), the distribution of final diagnoses is (n = 25) and confusion in 10.5% (n = 8). SE was observed
shown in figure 3. in 5.5% of epileptic patients (n = 19). Unfortunately, the
tentative diagnosis was ascertained in only 3 of them
Specific Insights (15.8%). The main false negative diagnosis, correspond-
Stroke. In 567 patients the final diagnosis was stroke or ing to 84.2% of SE, was poor evaluation of seizure in
TIA. A false positive tentative diagnosis of stroke had 68.8% (n = 11 patients), confusion in 18.7% (n = 3), par-

210 Eur Neurol 2003;50:207214 Moulin et al.

Fig. 2. Comparisons between the percentage
of false negative or false positive diagnoses
and the correct tentative diagnoses for each
category of signs, symptoms and syndromes
(figures within the diagram are absolute

kinsonian tremor in 6.3% (n = 1) and subdural haemato- and clinically identified (false negative patients). The ten-
ma in 6.3% (n = 1). tative diagnoses for these patients were: confusion in 2
Headaches. Headaches motivated the assessment by a patients, migraine in 2, stroke in 1 and coma in 1.
neurologist in 151 patients (9%). The main distribution of Confusion. Ninety patients were tentatively diagnosed
the tentative diagnosis was: headaches with no further with confusion. The neurologist confirmed this diagnosis
details in 62.2% (n = 94), subarachnoidal haemorrhage in only 28.9% (26/90 patients). The other tentative diag-
(SAH) in 12.5% (n = 19), migraine or cluster headaches in noses (false positive) were: degenerative disorders in 20%
3.9% (n = 6) and miscellaneous (intracranial hyperten- (dementia in 11 patients, transient global amnesia in 3,
sion, orthostatic, psychological or tension headaches) in parkinsonism in 4), stroke in 16.7% (15 patients), epilep-
21.1% (n = 32). Of the 19 patients suspected to have SAH, tic seizure or SE in 10% (5 and 3 patients, respectively),
the diagnosis was evident in only 47.4% (9/19 patients). SAH or subdural haematoma in 5.6% (5 patients), psy-
In the remaining 52.6% (10/19 patients), the diagnosis of chiatric disorders in 6.7% (n = 6), normal neurological
SAH was not confirmed (false positive patients). Seven of examination in 7.8% (n = 7) and miscellaneous in 5.5%
these patients had migraine, 1 meningitis, 1 was in meta- (cerebral tumour in 1 patient, metabolic coma in 1,
bolic coma and 1 had stroke. On the other hand, 6 of the encephalitis in 3). Among the 95 patients with a final
15 patients (overall) with true SAH (40%) were not clearly diagnosis of confusion, the main aetiologies were: meta-

Impact of Emergency Room Neurologists Eur Neurol 2003;50:207214 211

on Patient Management and Outcome

Fig. 3. Black bars represent the percentage

distribution of the final diagnosis (by symp-
tom or syndrome and based on the neurolog-
ical examination) of the 309 patients dis-
charged directly from the ER. Grey bars
represent the percentage of discharged pa-
tients with a given symptom or syndrome in
relation to the total number of patients with
the same symptom or syndrome.

bolic disorders in 10.5% (10/95 patients), acute alcohol- already been detected. This does not imply that all neuro-
ism in 5.2% (5/95), hypoglycaemia in 11.6% (11/95), ther- logical patients admitted to the ER were assessed by a
apeutic intoxication in 21% (20/95), infection or fever in neurologist, although this is the usual procedure in our
21% (20/95) and dementia in 30.5% (29/95). hospital to minimise risk. Moreover, ER physicians were
not aware of the study so that they would not change their
usual calling behaviour. Additionally, our study took a
Discussion pragmatic approach in order to analyse the potential and
different impacts of neurological problems in an ER. A
There is little data on the contribution of the neurolo- neurologist clearly needs to have specific training in neu-
gist in the ER despite the fact that neurological emergen- rological emergencies [6, 7].
cies represent approximately 15% of admissions to the Our study found that neurological intervention in the
ER [5]. Most of the studies concern a specific category of ER led to a complete contribution in 52.5% of the cases.
neurological emergency, such as stroke, epilepsy, or head- This global contribution might well appear to be overesti-
ache [14]. The emergency unit at Besanon University mated if the role of the neurologist was limited only to the
Hospital is the only one for adults in the city and as such neurological examination and did not include placing
must deal with all neurological conditions. Some 15% examination results in an overall neurological line of rea-
(14.7%) of emergency admissions are related to a neuro- soning. Moreover, the benefits of these examinations
logical dysfunction. It is to be noted that there may be a clearly increase after a patients assessment by a neurolo-
certain bias, as neurologists were undoubtedly called to gist, and not only in terms of a positive contribution to
see some patients in whom a neurological problem had technical investigations. This is especially true for correla-

212 Eur Neurol 2003;50:207214 Moulin et al.

tions between clinical examinations and EEGs [8] and The assessment by a neurologist in the ER is the first
between clinical examinations and CT scans [9]. While step. Neurological monitoring then has to be pursued in
the neurologist often does not influence decision-making the designated neurology unit before the neurologist takes
in neuro-traumatology, his specific role appears essential charge of implementing the appropriate follow-up. Im-
in all neurological pathologies for which clinical deduc- proved mortality and morbidity rates epitomise the effi-
tion is fundamentally relevant [6, 10]. Thus our auto-eval- ciency and benefits of stroke patient management in a
uation during the final synthesis was obviously subjective. stroke unit [20]. In our series, only one third of the
The neurological impact in the ER would have been better patients admitted to our stroke unit had a consistent diag-
analysed by a direct comparison between our ER perfor- nosis and clear indications for patient management. Cor-
mance and no neurological interventions in the ER. On rect early diagnosis also improves the cost/benefit ratio of
the other hand, we unfortunately did neither determine the stroke unit, as shown in the Lille experience, where up
the degree of impact for each and every act, especially to 15% of the patients with no stroke were mistakenly
when neurological examinations were needed, nor their admitted to this unit [4]. Other studies have reported sim-
individual benefits. However, the high rates of false nega- ilar errors in 1.519% of cases [16, 21]. In our study, a
tive and positive diagnoses (37.3 and 36.6%, respective- false positive diagnosis of stroke or TIA was suspected in
ly), whatever the different groups of signs or syndromes, 22.3% of our cases. Inversely, the diagnosis was not sus-
also highlight the specificity of the neurological evalua- pected in 23.8%. In such circumstances, when the exper-
tion. It was also interesting to note that the so-called tenta- tise required for correct diagnoses and immediate access
tive diagnoses of the ER team were in part only signs and to the best-suited treatment, such as thrombolysis, are not
symptoms and not a proper diagnosis including all ele- applied, the chance for improvement can be missed.
ments in a formal synthesis. Finally, the contribution of Moreover, efficient and experienced clinical examination
the neurologist to patient management is difficult to for early CT scan signs allows for better therapeutic deci-
determine because it can also depend on the neurological sions about anti-thrombotic drugs [22, 23].
training, experience and competence of each ER physi- Epileptic seizures also show the contribution of the
cian. They, too, emphasise the need to implement collabo- neurologist in the ER. Neurological evaluation allows
rative procedures between neurologists and the ER team patients to be discharged quickly, as was the case for
using upgraded algorithms and patient management, as 25.9% of all our discharged patients. In addition, the neu-
described in neurology guidelines [11]. rologist should be able to improve the management of epi-
The fact that examinations took place during the night leptic patients, especially those with SE, which was cor-
(18.3006.30 h) for 43.7% of patients, regardless of the rectly diagnosed in only 15.8% of our patients [8]. Head-
day of the week, stresses the need for a neurologist 24 h a aches are frequently clarified by emergency neurological
day. The fact that the mean interval between the time examination, in a proportion varying from 9.3 to 16% in
neurologists were called and their first act in the ER was recent reports close to the rate found in our series (9%)
about 30 min highlights the importance of their availabili- [24, 25]. While migraine and tension headaches have a
ty. However, the reaction time before this call (1 h) benign prognosis, SAH clearly requires careful attention
depended largely on the tentative diagnosis. Both of these with respect to specific management. Interestingly, the
intervals indicate the loss of time that needs to be recov- 7.5/100,000 incidence rate found in our series seems to
ered, especially in the case of thrombolytic therapy for correspond to rates published previously in European
stroke patients (which should be initiated within the first countries (411%) [26]. Surprisingly, our results showed
3 h [5, 12, 13]) or in patients with SE [14]. On the other that a clear clinical diagnosis for SAH was not the rule
hand, the assessment by a neurologist also made it possi- (only 47.4% of our patients), contrary to the high rate
ble to redirect more than one third of the patients admit- (87%) reported by Ferro et al. [27]. On the other hand, the
ted to the ER (17.4% were sent to another unit and 18.4% number of false negative patients (40%) does indeed high-
discharged). This effect, already demonstrated in the light the utility of neurological specificity. Acute confu-
management of an ageing population [15], stroke [16], sion is a non-specific syndrome that is probably over-used
headaches [17], confusion [18], and non-convulsive SE for neurological disturbances [18]. Our study confirms
[19], does seem particularly efficient when seen in the per- this with a low rate of correct diagnoses (23%). Important
spective of reducing length of hospital stay and improving diagnoses, such as epileptic seizures (41%), non-convul-
cost efficiency and the quality of healthcare. sive SE (10%), stroke (16.7%) and, surprisingly, SAH
(5.6%) have too frequently been diagnosed [28]. All of

Impact of Emergency Room Neurologists Eur Neurol 2003;50:207214 213

on Patient Management and Outcome
these examples emphasise the important prognostic con- In conclusion, our study stresses the need for a neurolo-
sequences and the potential medico-legal implications. gist in the ER setting, both in quantitative terms and for
The neurologist does have a major role in the ER, as has the benefit of patient management and neurological fol-
been underlined by guidelines in the management of neu- low-up. This also emphasises the importance of specific
rological emergencies [29, 30]. training to reinforce the links between ER and neurology:
the ER physician needs to be trained in neurology and the
neurologist needs emergency training.


1 Hopkins A, Menken N, Defriese G: A record of 11 Bertram M, Schwarz S, Hacke W: Acute and 22 Moulin T, Cattin F,Crepin-Leblond T, Tatu L,
patient encounters in neurological practise in critical care in neurology. Eur Neurol 1997;38: Chavot D, Piotin M,Viel JF, Rumbach L, Bon-
the United Kindom. J Neurol Neurosurg Psy- 155166. neville JF: Early CT-scan signs in acute middle
chiatry 1989;52:436438. 12 Adams HP, Brott TG, Furlan AJ, Gomez CR, cerebral artery infarction: Predictive value for
2 Papatetropoulos T, Tsibre E, Pelekoudas V: Grotta J, Helgason CM, Kwiatkowski T, Lyden subsequent infarct location. Neurology 1996;
The neurological content of general practice. J PD, Marler JR, Torner J, Feinberg W, Mayberg 47:366375.
Neurol Neurosurg Psychiatry 1989;52:434 M, Thies W: Guidelines for thrombolitic thera- 23 Toni D, Fiorelli M, Bastianello S, Sacchetti
435. py for acute stroke: A supplement to the guide- ML, Sette G, Argentino C, Montinaro E, Boz-
3 Ferro JM, Pinto AN, Falcao I, Rodrigues G, lines for the management of patients with acute zao L: Hemorrhagic transformation of brain
Ferreira J, Falcao F, Azevedo E, Canhao P, ischemic stroke. A statement for healthcare infarct: Predictability in the first five hours
Melo TP, Rosas MJ, Olivieira V, Salgado AV: professionals from a special writing group of from stroke onset and influence on clinical out-
Diagnosis of stroke by the nonneurologist: A the stroke council, American Heart Associa- come. Neurology 1996;46:341345.
validation study. Stroke 1998;29:11061109. tion. Stroke 1996;27:17111718. 24 Bigal ME, Bordini CA, Speciali JG: Etiology
4 Anonymous: The Lille Stroke Program. Mis- 13 The national institute for neurological disor- and distribution of headaches in two Brazilian
diagnosis in 1,250 consecutive patients admit- ders and stroke rt-PA stroke study group. Tis- primary care units. Headache 2000;40:241
ted to an acute stroke unit. Cerebrovasc Dis sue plasminogen activator for acute ischemic 247.
1997;7:284288. stroke. N Engl J Med 1995;333:15811587. 25 Newmann LC, Lipton RB: Emergency depart-
5 Adams HP, Brott TG, Crowell RM, Furlan AJ, 14 De Lorenzo RJ, Hauser WA, Towne AR, Boggs ment evaluation of headache. Neurol Clin
Gomez CR, Grotta J, Helgason CM, Marler JG, Pellock JM, Penberthy L, Garnett L, Fort- 1998;16:285303.
JR, Woolson RF, Zivin JA, Feinberg W, May- ner CA, Ko D: A prospective, population-based 26 Sudlow CL, Warlow CP: Comparable studies
berg M: Guidelines for the management of epidemiologic study of status epilepticus in of the incidence of stroke and its pathological
patients with acute ischemic stroke. A state- Richmond, Virginia. Neurology 1996;46: types: Results from an international collabora-
ment for healthcare professionals from a spe- 10291035. tion. International Stroke Incidence Collabora-
cial writing group of the stroke council, Ameri- 15 Camicioli RM, Kaye JA, Brummel-Smith K: tion. Stroke 1997;28:491499.
can Heart Association. Stroke 1994;25:1901 Recognition of neurologic diseases in geriatric 27 Ferro JM, Lopes J, Melo TP, Oliveira V, Cres-
1914. inpatients. Act Neurol Scand 1998;97:265 po M, Campos JG, Trindade A, Antunes JL:
6 Chimowitz MI, Logician EL, Caplan LR: The 270. Investigation into the causes of delayed diagno-
accuracy of bedside neurological diagnoses. 16 Kothari RU, Brott T, Broderick JP, Hamilton sis of subarachnoid hemorrhage. Cerebrovasc
Ann Neurol 1990;28:7885. CA: Emergency physicians. Accuracy in the Dis 1991;1:161164.
7 DEsposito M: Profile of the neurology residen- diagnosis of stroke. Stroke 1995;26:2238 28 Reijneveld JC, Wermer M, Boonman Z, van
cy. Arch Neurol 1995;52:11231126. 2241. Gijn J, Rinkel GJ: Acute confusional state as
8 Jallon P, Goumaz M, Haenggelli C, Morabia A: 17 Luda E, Comitangelo R, Sicuro L: The symp- presenting feature in aneurysmal subarachnoid
Incidence of first epileptic seizures in the can- tom of headache in emergency departments. hemorrhage: Frequency and characteristics. J
ton of Geneva, Switzerland. Epilepsia 1997;38: The experience of a neurology emergency de- Neurol 2000;247:112116.
547552. partment. Ital J Neurol Sci 1995;16:295301. 29 Practice advisory. Thrombolytic therapy for
9 Schriger DL, Kalafut M, Starkman S, Krueger 18 Calandre L, Esteban J, Bermejo F: Acute con- acute ischemic stroke-summary statement. Re-
M, Saver JL: Cranial computed tomography fusional syndrome of unknown cause. Prospec- port of the quality standards subcommittee of
interpretation in acute stroke: Physician accu- tive study in the emergency room (abstract). the American Academy of Neurology. Neurolo-
racy in determining eligibility for thrombolytic Neurologia 1990;5:196199. gy 1996;47:835839.
therapy. JAMA 1998;279:12931297. 19 Kaplan PW: Nonconvulsive status epilepticus 30 The European ad hoc consensus group: Euro-
10 Patterson VH, Esmonde TFG: Comparison of in the emergency room. Epilepsia 1996;37: pean strategies for early intervention in stroke:
the handling of neurological outpatient refer- 643650. A report of an ad hoc consensus group meeting.
rals by general physicians and a neurologist. J 20 Langhorne P, Williams BO, Gilchrist W, Cerebrovasc Dis 1996;6:315324.
Neurol Neurosurg Psychiatry 1993;56:830 Howie K: Do stroke units save lives? Lancet
833. 1993;342:395398.
21 Ferro JM, Falcao I, Rodrigues G, Canhao P,
Melo TP, Oliveira V, Pinto AN, Crespo M, Sal-
gado AV: Diagnosis of transient ischemic at-
tack by the nonneurologist: A validation study.
Stroke 1996;27:22252229.

214 Eur Neurol 2003;50:207214 Moulin et al.

Copyright: S. Karger AG, Basel 2003. Reproduced with the permission of S. Karger AG, Basel. Further
reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright