Published Ahead of Print on April 21, 2017 as 10.1212/WNL.
0000000000003940
Tobias Bobinger, MD
Bernd Kallmünzer, MD
Markus Kopp, MD
Natalia Kurka, MD
Martin Arnold, MD
Stefan Heider, MD
Stefan Schwab, MD
Martin Köhrmann, MD
Correspondence to
Dr. Bobinger:
Tobias.Bobinger@uk-erlangen.de
Supplemental data
at Neurology.org
ª 2017 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Diagnostic value of prehospital ECG in
acute stroke patients
ABSTRACT
Objective: To investigate the diagnostic yield of prehospital ECG monitoring provided by emer-
gency medical services in the case of suspected stroke.
Methods: Consecutive patients with acute stroke admitted to our tertiary stroke center via emer-
gency medical services and with available prehospital ECG were prospectively included during
a 12-month study period. We assessed prehospital ECG recordings and compared the results
to regular 12-lead ECG on admission and after continuous ECG monitoring at the stroke unit.
Results: Overall, 259 patients with prehospital ECG recording were included in the study (90.3%
ischemic stroke, 9.7% intracerebral hemorrhage). Atrial fibrillation (AF) was detected in 25.1% of
patients, second-degree or greater atrioventricular block in 5.4%, significant ST-segment eleva-
tion in 5.0%, and ventricular ectopy in 9.7%. In 18 patients, a diagnosis of new-onset AF with
direct clinical consequences for the evaluation and secondary prevention of stroke was estab-
lished by the prehospital recordings. In 2 patients, the AF episodes were limited to the prehospital
period and were not detected by ECG on admission or during subsequent monitoring at the stroke
unit. Of 126 patients (48.6%) with relevant abnormalities in the prehospital ECG, 16.7%
received medical antiarrhythmic therapy during transport to the hospital, and 6.4% were trans-
ferred to a cardiology unit within the first 24 hours in the hospital.
Conclusions: In a selected cohort of patients with stroke, the in-field recordings of the ECG de-
tected a relevant rate of cardiac arrhythmia. The results can add to the in-hospital evaluation
and should be considered in prehospital care of acute stroke. Neurology® 2017;88:1–5
GLOSSARY
AF 5 atrial fibrillation; ASSERT 5 Asymptomatic Atrial Fibrillation and Stroke Evaluation Trial; CEM 5 continuous monitoring;
CI 5 confidence interval; EMS 5 emergency medical services; IQR 5 interquartile range.
ECG abnormalities are highly prevalent in patients with acute stroke. They may represent car-
diac comorbidity, can be the cause of the present stroke mainly in cases with atrial fibrillation
(AF), but also can be a consequence of autonomic dysfunction due to the ischemic lesion it-
self.1 Severe cardiac arrhythmia may harm patients by hemodynamic instability and can even
lead to sudden cardiac death.2–4 Therefore, cardiac monitoring plays an important role in acute
stroke care. Current stroke guidelines recommend the initiation of cardiac monitoring by
EMS personnel during prehospital evaluation, but transport to the hospital should not be
delayed.5 While prehospital ECG monitoring in patients with acute myocardial infarction is
embedded in all guidelines and widely accepted,6,7 data on its value in patients with stroke are
sparse.8 Because AF accounts for up to 30% of all strokes, monitoring in acute stroke patients
is essential to detect subclinical episodes of AF besides recognizing and addressing life-
threatening arrhythmias.9,10 Detection of subclinical episodes of AF in a clinical setting is
challenging, and extended cardiac monitoring detects higher rates of paroxysmal AF.11,12
Therefore, a prehospital 12-lead ECG may be another tool in the detection of AF. The aim
of the current study was to determine the diagnostic value of prehospital ECG in acute stroke
patients.
From the Departments of Neurology (T.B., B.K., M. Kopp, N.K., S.H., S.S., M. Köhrmann) and Cardiology (M.A.), Universitätsklinikum
Erlangen; and Department of Neurology (M. Köhrmann), Universitätsklinikum Essen, Germany.
Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.
© 2017 American Academy of Neurology 1
 METHODS Study design and setting. Consecutive patients
with acute ischemic or hemorrhagic stroke admitted by EMS to
the Neurological emergency room at the University-Hospital
Erlangen, Germany, with prehospital ECG recording were pro-
spectively included in this study. The study period covered Jan-
uary to December 2013. Clinical data and medical information
were obtained from our institutional prospective database. Stroke
severity was assessed with the NIH Stroke Scale. Functional status
was evaluated with the modified Rankin Scale.
Standard protocol approvals, registrations, and patients
consents. Data of all stroke patients treated at the Department of
Neurology, University-Hospital Erlangen, Germany, were
entered into a prospective database, which was approved by the
local ethics committee.
Analysis of ECG and statistical analysis. Prehospital ECGs
and baseline admission ECGs (recorded with a sensitivity of 10
mm/mV and a paper speed of 50 mm/s in all stroke patients as
standard of care) were evaluated by 2 experienced investigators.
ECGs were analyzed for the following parameters, according to
published guidelines13: accuracy and quality of the ECG, heart
rate, axis deviation, rhythm disturbances (AF, atrial flutter, sinus
tachycardia, pacemaker rhythm), atrioventricular block, ST-
segment depression or elevation, and T-wave abnormalities.
Depression of the ST segment of .1 mm and elevation of 1 mm
in extremity and precordial leads were considered significant
abnormalities. Continuous ECG monitoring in our stroke unit
was performed with standard telemetric monitoring with
a structured algorithm used to assess for episodes of AF and other
higher-grade arrhythmias, as recently published.14 Data were
processed with the Office 2013 (Microsoft Corp, Redmond, WA)
software package. Statistical analysis was performed with SPSS
version 21.0 (SPSS Inc, Chicago, IL). For the baseline data,
normality of distribution was tested with the Kolmogorov-
Table 1 Baseline characteristics of the study population
Study group (n 5 259)
Age, y 76 (67–82)
Female, n (%) 115 (44.4)
Pre-mRS (0–6) 1 (0–2)
Ischemic stroke, n (%) 234 (90.3)
Intracerebral hemorrhage, n (%) 25 (9.7)
Treated with thrombolysis, n (%) 77 (29.7)
Time from stroke onset to admission to hospital, min 177 (67–389)
Transferred patients from other hospitals, n (%) 26 (10.0)
NIHSS score on admission (0–42) 5 (2–10)
Arterial hypertension, n (%) 230 (88.8)
Diabetes mellitus, n (%) 75 (29.0)
Current nicotine abuse, n (%) 38 (14.7)
History of AF, n (%) 73 (28.2)
AF detected during stay in hospital, n (%) 31 (12.0)
Anticoagulation use before stroke, n (%) 37 (14.3)
Anticoagulation use/recommendation after discharge, n (%) 85 (32.8)
Structural heart disease, n (%) 106 (40.9)
Abbreviations: AF 5 atrial fibrillation; mRS 5 modified Rankin Scale; NIHSS 5 NIH Stroke
Scale.
2 Neurology 88 May 16, 2017
ª 2017 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Smirnov test. Normally distributed data were presented as
means and SDs; otherwise, medians and interquartile ranges
(IQRs) were provided. For group comparisons between baseline
characteristics of patients with and without prehospital ECG
recording, the x2 test or Mann-Whitney U test was used as
applicable. For visualizing of AF detection, illustration via Venn
diagrams was applied with the Venn Diagram Plotter (version
1.5.5228, Pacific Northwest National Laboratory, Richland,
WA, http://omics.pnl.gov).15
RESULTS Analysis of the prehospital ECG and char-
acteristics of the study population. During our study
period, 259 patients with prehospital ECG were
included in this study. In 0.8% of the patients, the
prehospital ECG recording was of insufficient quality
for full evaluation. Full 12-lead prehospital ECG was
available in 78% of the patients; in 22%, fewer leads
were documented by EMS personal. Prehospital
ECG was recorded within a median of 189 minutes
(IQR 86–576 minutes) after stroke onset. The mean
age of the study cohort with prehospital ECG was 76
years (67–82 years); 90.3% (n 5 234) presented with
ischemic stroke, and 9.7% (n 5 25) presented with
hemorrhagic stroke. Further baseline characteristics
are presented in table 1. During the study period, 822
stroke patients were admitted to our hospital without
prehospital ECG. When the baseline characteristics
of patients with prehospital ECG (n 5 259) were
compared to those of patients without prehospital
ECG recording (n 5 822), patients with prehospital
ECG were significantly older (76 years [IQR 67–82
years] with prehospital ECG vs 72 years [IQR 60–80
years] without prehospital ECG, p , 0.001), showed
a higher rate of structural heart disease (106 [40.9%]
with prehospital ECG vs 207 [25.2%] without
prehospital ECG, p , 0.001), and had a worse
neurologic state on admission (NIH Stroke Scale on
admission: 5 [IQR 2–10] with prehospital ECG vs 4
[IQR 1–10] without prehospital ECG, p 5 0.035)
(table e-1 at Neurology.org).
Prevalence of AF. In our study population (n 5 259),
AF was previously known in 28.2% (n 5 73) of
patients. In patients without a history of AF (n 5
186), new AF was found in 31 patients (16.7%,
95% confidence interval [CI] 11.8–23.0) during
the hospital stay. Prehospital ECG showed AF in
18 patients (9.7%, 95% CI 6.0–15.1), and in 2 pa-
tients, detection of AF was limited to the prehospital
ECG recording and did not reoccur on 12-lead ECG
recording on hospital admission and during contin-
uous monitoring (CEM). Detection of AF by the
different methods is illustrated in figure 1 (Venn
diagram).
In addition to the 31 patients with newly diag-
nosed AF, 54 of 73 patients (74%, 95% CI 62.2–
83.2) with previously diagnosed AF showed at least
one episode of AF during in-hospital ECG
 Figure 1 Venn diagram of newly detected AF (n 5 31) stratified by different
methods
Atrial fibrillation (AF) in prehospital ECG (PE-ECG, n 5 18), AF in ECG on admission to
hospital (AE-ECG, n 5 15), and AF in continuous monitoring (CEM) in our stroke unit (n 5
29). The graph shows that many episodes are detected only during CEM; however, in 2
patients, detection of AF was limited to prehospital ECG.
diagnostics (figure e-1). Figure 2 illustrates detection
yields for different ECG diagnostics in this patient
cohort with at least one AF episode during monitor-
ing (n 5 85) (figure 2). In 3 cases, detection of AF
was limited to prehospital ECG; in 1 case, to ECG on
admission; and in 16 cases, to CEM. However, the
majority of cases (n 5 50) showed AF in all 3 detec-
tion methods.
Prevalence of other cardiac arrhythmias. The rate of
ECG abnormalities in prehospital ECG is presented
in table 2. A second- or third-degree atrioventricular
block was detected in 5.4% (95% CI 3.1–9.2),
significant ST-segment elevation in 5.0% (95% CI
Figure 2 Venn diagram of detection of AF in patients with detected AF during
the hospital stay
Atrial fibrillation (AF) in prehospital ECG (PE-ECG, n 5 65), AF in ECG on admission to
hospital (AE-ECG, n 5 56), and AF in continuous monitoring (CEM) in our stroke unit (n 5
79). The graph illustrates that CEM and ECG on admission are the most successful methods
to detect AF; however, in 3 cases, detection was limited to prehospital ECG.
ª 2017 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Table 2 ECG characteristics of prehospital ECG
Study group (n 5 259)
Insufficient quality/artifacts, n (%) 2 (0.8)
12-Lead ECG, n (%) 202
(78.0)
Sinus rhythm, n (%) 180
(69.5)
Atrial fibrillation, n (%) 65 (25.1)
Pacemaker rhythm, n (%) 14 (5.4)
Second- or third-degree atrioventricular block, 14 (5.4)
n (%)
Significant ST-segment elevation, n (%) 13 (5.0)
Significant premature ventricular complexes, 25 (9.7)
n (%)
2.8–8.6), and a pacemaker rhythm in 5.4% (95% CI
3.1–9.2) of the ECGs. Because full 12-lead pre-
hospital ECG was available in only 78% of the cases
(n 5 202), analysis of bundle-branch block was
limited to these cases. Of the ECGs, 8.9% (n 5 18)
showed signs of a complete right bundle-branch block,
and 2.0% (n 5 4) showed signs of a complete left
bundle-branch block. Thus, in 48.6% (n 5 126) of
patients, prehospital ECG showed relevant abnormal-
ities. Of these patients, 16.67% (n 5 21) received
antiarrhythmic medications during the prehospital
phase, 6.35% (n 5 8) were transferred to a cardiology
unit during the first 24 hours, and 1.6% (n 5 2)
needed cardiac resuscitation during the first 24 hours.
DISCUSSION Prehospital ECG recording evolved as
a widely accepted standard of care in patients with
suspected myocardial infarction.6,16–18 While cardiac
monitoring is also recommended in acute stroke pa-
tients by the American Stroke Association and the
European Stroke Organisation, data on the clinical
value of prehospital ECG are still lacking.5,19
Even though it is advised to start ECG recording
as early as possible, a prehospital ECG was available
in only a quarter of patients admitted to our hospital
via EMS. In a small number of cases, ECGs were of
insufficient quality or restricted to several leads
because of technical errors or the incapability of the
older ECG devices to perform a full 12-lead record-
ing. Accurate prehospital 12-lead ECG recording
may be difficult, especially in agitated or severely
aphasic patients.20 As expected, analysis of baseline
characteristics revealed that ECG was more fre-
quently performed and handed over to emergency
room personnel for patients who had a higher rate
of cardiac morbidities and a worse state of health.
Detection of paroxysmal AF remains a challenging
task in stroke care. Several approaches have been
offered to increase the number of detected patients
with underlying AF.11,12,21 Moreover, the relationship
Neurology 88 May 16, 2017 3
 between paroxysmal AF and embolism is not clearly
understood. Recently, published data from the
Asymptomatic Atrial Fibrillation and Stroke Evalua-
tion Trial (ASSERT) gained major attention. Just
a minority of patients with implanted pacemaker
showed episodes of AF in the month before their
stroke episode or even during the ischemic event.10,22
This fact supports the thesis that AF may simply be
a general risk marker for stroke. More complex mech-
anisms such as alterations in the atrial endothelium
may lead to systemic embolism.22 In patients without
a pacemaker or loop recorder, prehospital ECG
recording is the first ECG after the onset of stroke.
Actually, in 2 patients without a history of AF, AF
was limited to prehospital ECG recording. In patients
with prehospital sinus rhythm, ECG on admission
and CEM detected AF in 13 patients. In summary,
prehospital ECG adds value to detect AF beside base-
line ECG and CEM in the stroke unit.
As shown recently, serious cardiac arrhythmias are
most likely to occur within the first 24 hours after
stroke unit admission.3 This may be explained at least
in part by stroke-associated autonomic dysfunction
and thus the arrhythmogenic nature of stroke itself.
Several studies revealed that abnormalities on ECG
are a predictor of early mortality after stroke.23–25 In
our study, a relevant proportion of patients showed
advanced atrioventricular block, significant ST-
segment elevation, or premature ventricular complexes.
Neglecting these ECG abnormalities may result in
serious acute cardiac adverse events, and therefore, car-
diac monitoring should start as early as possible.3 Our
results of serious ECG abnormalities are in concor-
dance with other studies published on ECG changes
after stroke and emphasize the importance of preho-
spital ECG monitoring in stroke patients.24,26
The study population represents a typical popula-
tion in a stroke unit. However, it is limited by the
single-center design and a still relatively small number
of patients. In addition, the results may have been
biased because prehospital ECG may have been re-
corded more often in patients presenting with comor-
bidities, and therefore, only this subset is included in
the study. This may lead to overestimation of abnor-
malities seen in our cohort compared to an unselected
stroke population.
Acute stroke patients are endangered by cardiac ar-
rhythmias and ECG abnormalities; therefore, cardiac
monitoring should start as early as possible. This study
demonstrated that prehospital ECG recording detects
a relevant rate of abnormalities. Moreover, prehospital
ECG recording can be a helpful tool in AF detection.
AUTHOR CONTRIBUTIONS
Tobias Bobinger: study concept and design, acquisition of data, analysis
and interpretation of data, statistical analysis, drafting of manuscript.
4 Neurology 88 May 16, 2017
ª 2017 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Bernd Kallmünzer: study concept and design, acquisition of data, statis-
tical analysis, critical revision of manuscript. Markus Kopp: study concept
and design, acquisition of data. Natalia Kurka: study concept and design,
critical revision of manuscript. Martin Arnold: acquisition of data, critical
revision of manuscript for intellectual content. Stefan Heider: critical
revision of manuscript, acquisition of data. Stefan Schwab: study concept
and design, critical revision of manuscript for intellectual content. Martin
Köhrmann: study concept and design, drafting of the manuscript, critical
revision of manuscript for intellectual content.
ACKNOWLEDGMENT
The authors thank Pacific Northwest National Laboratory (operated by
Battelle Memorial Institute, Omics.PNL.gov) for distribution of Venn
Diagram plotter version 1.5.52.
STUDY FUNDING
No targeted funding reported.
DISCLOSURE
The authors report no disclosures relevant to the manuscript. Go to
Neurology.org for full disclosures.
Received October 20, 2016. Accepted in final form February 10, 2017.
REFERENCES
1. Dorrance AM, Fink G. Effects of stroke on the autonomic
nervous system. Compr Physiol 2015;5:1241–1263.
2. Klingelhofer J, Sander D. Cardiovascular consequences of
clinical stroke. Baillieres Clin Neurol 1997;6:309–335.
3. Kallmunzer B, Breuer L, Kahl N, et al. Serious cardiac
arrhythmias after stroke: incidence, time course, and
predictors–a systematic, prospective analysis. Stroke
2012;43:2892–2897.
4. Soros P, Hachinski V. Cardiovascular and neurological
causes of sudden death after ischaemic stroke. Lancet Neu-
rol 2012;11:179–188.
5. Jauch EC, Saver JL, Adams HP Jr, et al. Guidelines for
the early management of patients with acute ischemic
stroke: a guideline for healthcare professionals from the
American Heart Association/American Stroke Association.
Stroke 2013;44:870–947.
6. Cannon AR, Lin L, Lytle B, Peterson ED, Cairns CB,
Glickman SW. Use of prehospital 12-lead electrocardiog-
raphy and treatment times among ST-elevation myocardial
infarction patients with atypical symptoms. Acad Emerg
Med 2014;21:892–898.
7. Brainard AH, Raynovich W, Tandberg D, Bedrick EJ.
The prehospital 12-lead electrocardiogram’s effect on time
to initiation of reperfusion therapy: a systematic review
and meta-analysis of existing literature. Am J Emerg
Med 2005;23:351–356.
8. Munro SF, Cooke D, Kiln-Barfoot V, Quinn T. The use
and impact of 12-lead electrocardiograms in acute stroke
patients: a systematic review. Eur Heart J Acute Cardio-
vasc Care Epub 2015 Dec 4.
9. Rizos T, Guntner J, Jenetzky E, et al. Continuous stroke
unit electrocardiographic monitoring versus 24-hour
Holter electrocardiography for detection of paroxysmal
atrial fibrillation after stroke. Stroke 2012;43:2689–2694.
10. Healey JS, Connolly SJ, Gold MR, et al. Subclinical atrial
fibrillation and the risk of stroke. N Engl J Med 2012;366:
120–129.
11. Kallmunzer B, Bobinger T, Kahl N, et al. Peripheral pulse
measurement after ischemic stroke: a feasibility study.
Neurology 2014;83:598–603.
 12. Higgins P, MacFarlane PW, Dawson J, McInnes GT,
Langhorne P, Lees KR. Noninvasive cardiac event
monitoring to detect atrial fibrillation after ischemic
stroke: a randomized, controlled trial. Stroke 2013;44:
2525–2531.
13. O’Keefe JH. The Complete Guide to ECGs, 3rd ed. Royal
Oak: Physicians’ Press; 2008:5–11.
14. Kallmunzer B, Breuer L, Hering C, et al. A structured
reading algorithm improves telemetric detection of atrial
fibrillation after acute ischemic stroke. Stroke 2012;43:
994–999.
15. Kestler HA, Muller A, Gress TM, Buchholz M. General-
ized Venn diagrams: a new method of visualizing complex
genetic set relations. Bioinformatics 2005;21:1592–1595.
16. Ioannidis JP, Salem D, Chew PW, Lau J. Accuracy and
clinical effect of out-of-hospital electrocardiography in the
diagnosis of acute cardiac ischemia: a meta-analysis. Ann
Emerg Med 2001;37:461–470.
17. Drew BJ, Sommargren CE, Schindler DM, Benedict K,
Zegre-Hemsey J, Glancy JP. A simple strategy improves
prehospital electrocardiogram utilization and hospital
treatment for patients with acute coronary syndrome (from
the ST SMART Study). Am J Cardiol 2011;107:347–352.
18. Ferguson JD, Brady WJ, Perron AD, et al. The prehospital
12-lead electrocardiogram: impact on management of the
out-of-hospital acute coronary syndrome patient. Am J
Emerg Med 2003;21:136–142.
ª 2017 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
19. European Stroke Organisation Executive Committee,
ESO Writing Committee. Guidelines for management of
ischaemic stroke and transient ischaemic attack 2008. Cer-
ebrovasc Dis 2008;25:457–507.
20. Yperzeele L, Van Hooff RJ, De Smedt A, et al. Prehospital
stroke care: limitations of current interventions and focus
on new developments. Cerebrovasc Dis 2014;38:1–9.
21. Sanna T, Diener HC, Passman RS, et al. Cryptogenic
stroke and underlying atrial fibrillation. N Engl J Med
2014;370:2478–2486.
22. Brambatti M, Connolly SJ, Gold MR, et al. Temporal
relationship between subclinical atrial fibrillation and
embolic events. Circulation 2014;129:2094–2099.
23. Villa A, Bacchetta A, Milani O, Omboni E. QT interval
prolongation as predictor of early mortality in acute
ischemic stroke patients. Am J Emerg Med 2001;19:
332–333.
24. Christensen H, Fogh Christensen A, Boysen G. Abnor-
malities on ECG and telemetry predict stroke outcome at
3 months. J Neurol Sci 2005;234:99–103.
25. Katsanos AH, Korantzopoulos P, Tsivgoulis G, Kyritsis
AP, Kosmidou M, Giannopoulos S. Electrocardiographic
abnormalities and cardiac arrhythmias in structural brain
lesions. Int J Cardiol 2013;167:328–334.
26. Khechinashvili G, Asplund K. Electrocardiographic
changes in patients with acute stroke: a systematic review.
Cerebrovasc Dis 2002;14:67–76.
Neurology 88 May 16, 2017 5
 Diagnostic value of prehospital ECG in acute stroke patients
Tobias Bobinger, Bernd Kallmünzer, Markus Kopp, et al.
Neurology published online April 21, 2017
DOI 10.1212/WNL.0000000000003940

This information is current as of April 21, 2017

Updated Information &
Services
Supplementary Material
Subspecialty Collections
Permissions & Licensing
Reprints
including high resolution figures, can be found at:
http://www.neurology.org/content/early/2017/04/20/WNL.0000000000
003940.full.html
Supplementary material can be found at:
http://www.neurology.org/content/suppl/2017/04/21/WNL.000000000
0003940.DC1
This article, along with others on similar topics, appears in the
following collection(s):
All Cerebrovascular disease/Stroke
http://www.neurology.org//cgi/collection/all_cerebrovascular_disease_
stroke
Cardiac
http://www.neurology.org//cgi/collection/cardiac
Critical care
http://www.neurology.org//cgi/collection/critical_care
Information about reproducing this article in parts (figures,tables) or in
its entirety can be found online at:
http://www.neurology.org/misc/about.xhtml#permissions
Information about ordering reprints can be found online:
http://www.neurology.org/misc/addir.xhtml#reprintsus

Neurology ® is the official journal of the American Academy of Neurology. Published continuously since
1951, it is now a weekly with 48 issues per year. Copyright © 2017 American Academy of Neurology. All
rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.