Neglecting the Difference

Does Right or Left Matter in Stroke Outcome After Thrombolysis?

Silvia Di Legge, MD, PhD; Gustavo Saposnik, MD, MSc; Yongchai Nilanont, MD;
Vladimir Hachinski, MD, DSc, FRCPC

Background and Purpose—Patients with right hemispheric strokes (RHSs) present later to an emergency department, have
a lower chance to receive intravenous recombinant tissue plasminogen activator (IV rt-PA), and have worse clinical
outcomes than do patients with left hemispheric strokes (LHSs). We analyzed outcomes after IV rt-PA with respect to
the side of the affected hemisphere.
Methods—A prospective cohort of acute stroke patients was treated with IV rt-PA at the London Health Sciences Centre
(December 1998 to March 2003). Differences between patients with RHS and LHS were identified by univariate
analysis. Logistic-regression analysis was used to determine a subset of variables independently associated with major
neurological improvement at 24 hours and good outcome at 3 months after treatment.
Results—Of 219 stroke patients who received IV rt-PA, 165 had hemispheric strokes (68 RHSs and 97 LHSs). Patients
with RHSs were less hypertensive (P⫽0.001) and had lower pretreatment National Institutes of Health Stroke Scale
(NIHSS) scores (P⫽0.005). LHS (odds ratio [OR], 2.29; 95% CI, 1.14 to 4.59; P⫽0.019), age (OR, 0.96; 95% CI, 0.93
to 0.99; P⫽0.012), and pretreatment NIHSS (OR, 0.83; 95% CI, 0.78 to 0.89; P⬍0.0001) were independent predictors
of 3-month outcome. Female sex (OR, 3; 95% CI, 1.53 to 5.90; P⫽0.001) and LHS (OR, 2.07; 95% CI, 1.05 to 4.08;
P⫽0.03) were independent predictors of major neurological improvement at 24 hours after IV rt-PA.
Conclusions—Despite higher pretreatment NIHSS, patients with LHSs have a 2-fold increased chance of a good outcome
3 months after rt-PA treatment compared with patients with RHSs. This gain can be clinically detected at 24 hours after
treatment. These results need to be coupled with neuroimaging and hemodynamic characteristics known to influence
stroke outcome. (Stroke. 2006;37:2066-2069.)
Key Words: acute stroke 䡲 outcome 䡲 thrombolysis

I nteresting and consistent data are emerging on barriers

preventing the successful administration of thrombolytic
therapy to patients who experience an acute ischemic stroke.
In particular, these barriers seem to be applicable to patients
with right hemispheric strokes (RHSs). Two studies have
shown that patients with RHSs present later to an emergency
department1,2 and are less frequently diagnosed as having had
a stroke.1 Our group previously observed that patients with
RHS are 45% less likely to receive intravenous recombinant
tissue plasminogen activator (IV rt-PA) compared with pa-
tients with left hemispheric strokes (LHSs).2 Neuroimaging
studies have shown that the National Institutes of Health Stroke
Scale (NIHSS) is underpowered for assessing the severity of
right hemisphere damage in the acute phase of stroke.3,4 As a
consequence, under current guidelines, patients with RHS are at
risk of not being properly assessed or treated with IV
thrombolysis. Outcome studies have also shown that patients
with RHS have a worse functional recovery compared with LHS
patients.5,6 Whether the administration of IV rt-PA can reverse
this pattern has not been specifically investigated. Our aim was
to investigate whether the affected hemisphere is (1) predictor of
neurological improvement at 24 hours and (2) a determinant of
good outcome at 3 months after IV administration of rt-PA.
Methods
Data from consecutive stroke patients treated with IV rt-PA at the
London Health Sciences Centre, University Campus in London,
Ontario, were collected prospectively. The hospital category, size of
the population served, and catchment area have been described.7 For
all patients, the time of symptom onset was defined by the time they
were “last seen to be well.” The onset-to-door, onset-to— computed
tomography (CT) scan of the brain, and onset-to-needle time were
obtained from nursing records. Data were entered into a database by
stroke Fellows, all of whom were certified in administering the
NIHSS. Collected information included demographics (age, sex,
handedness), evaluation and treatment time, admission and 24-hour

Received November 23, 2005; final revision received March 17, 2006; accepted March 27, 2006.
From the Stroke Unit (S.D.L.), Stroke Unit, Università di Tor Vergata, Rome, Italy; the Department of Clinical Neurological Sciences (G.S., V.H.),
London Health Science Centre, University of Western Ontario, London, Ontario, Canada; the Division of Neurology (Y.N.), Siriraj Hospital, Mahidol
University Bangkok, Thailand; and the Stroke Research Program (G.S.), Department of Medicine, St. Michael’s Hospital, University of Toronto, Ontario,
Canada.
Correspondence to Silvia Di Legge, MD, PhD, Stroke Unit, University of Tor Vergata, Viale Oxford 81-00133 Rome, Italy. E-mail
silvia.dilegge@uniroma1.it
© 2006 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org DOI: 10.1161/01.STR.0000229899.66019.62

2066
Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016
 Di Legge et al Outcome Predictors in Hemispheric Stroke Patients 2067

NIHSS scores, side of the affected brain area, and 3-month outcome. TABLE 1. Differences Between Patients With RHS and LHS
Handedness was recorded as part of the routine neurological exam-
ination. History of coronary artery disease, atrial fibrillation, heart RHS, LHS,
failure, transient ischemic attack or stroke, hypertension, diabetes Characteristic n⫽68 n⫽97 P
mellitus, hypercholesterolemia, smoking, and alcohol consumption Age (⫾SD), y 70.5 (⫾11) 72.3 (⫾13) 0.25
was obtained through self-report and supported by hospital records
or lists of medications taken and defined according to Canadian Age ⬍60 n (%) 52 (76) 80 (82) 0.43
standards. Excessive alcohol consumption was defined as any daily Sex, male, n (%) 30 (44) 52 (54) 0.27
intake ⬎50 g.8 During admission, all patients underwent a stroke Handedness, right, n (%) 63 (93) 92 (95) 0.74
diagnostic work-up, consisting of blood cell count, metabolic panel,
ECG, transthoracic echocardiogram, and carotid ultrasound. Mag- Median NIHSS at baseline 12 15 0.0047
netic resonance imaging (MRI), magnetic resonance and CT angiog- Mean glucose at 7.6 (⫾2.9) 7.7 (⫾2.7) 0.81
raphy, transesophageal echocardiography, and conventional angiog- admission, mmol/L (⫾SD)
raphy were preformed when appropriate. Stroke subtypes were recorded
History of vascular risk factors, n (%)
as lacunar and nonlacunar infarcts. Administration of IV rt-PA followed
the National Institute of Neurological Disorders and Stroke protocol,9 Hypertension 33 (49) 71 (74) 0.001
except for the exclusion of patients with involvement of more than one Diabetes 16 (24) 21 (22) 0.85
third of the middle cerebral artery territory on the baseline CT scan. The
Coronary artery disease 16 (24) 35 (37) 0.09
acute stroke protocol adopted in our institution has been de-
scribed.2,7,10,11 We retrospectively analyzed the distribution of all vari- Hypercholesterolemia 23 (34) 37 (41) 0.51
ables by both graphic and analytical methods (frequency distribution by Smoking 10 (15) 20 (21) 0.41
quartiles or quintiles). When the relation between a continuous variable
and the primary outcome was linear, it was kept continuous. When the Atrial fibrillation 13 (19) 22 (23) 0.70
relation suggested a cut-off, the variable was categorized. When there Excessive alcohol intake 5 (8) 9 (10) 0.75
was no clear relation, we used clinical criteria to analyze the variable. Prior transient ischemic attack 10 (15) 11 (11) 0.65
Age, baseline NIHSS, time from stroke onset to treatment, and total
rt-PA dose were considered continuous. A major neurological improve- Previous medications, n (%)
ment (MNI) was defined as an NIHSS score equal to 0 or 1 at 24 hours Aspirin 22 (34) 44 (46) 0.14
or an improvement ⱖ8 points compared with baseline, according to the Angiotensin-converting enzyme 17 (26) 31 (34) 0.30
definition of Brown and colleagues.12 Three-month outcome was
inhibitors
analyzed by the modified Rankin Scale (mRS).13
Statins 13 (20) 24 (26) 0.45
Statistical Analysis CT findings
Differences between patients were identified by univariate analysis. Hemorrhagic transformation, n (%) 6 (9) 12 (12) 0.62
Logistic-regression analysis was used to determine a subset of variables
independently associated with good outcome at the 3-month follow-up Stroke subtype
(ie, mRS 0 to 1) and MNI at 24 hours after treatment. Discrimination of Lacunar, n (%) 8 (12) 9 (9) 0.61
the model was assessed by the area under the receiver operating rt-PA–related times, mean
characteristic curve, and calibration was assessed with a goodness-of-fit (⫾SD), min
test. Statistical analysis was performed with a commercially available
software package (STATA, version 7.0, Stata Corp LP). Probability Onset-to–rt-PA 158 (30) 156 (32) 0.65
values ⬍0.05 were considered significant. Onset-to-door 88.5 (49) 81.7 (44) 0.45
Door-to-CT 49.7 (31) 49.1 (29) 0.90
Results rt-PA total dose, mg 66.9 (16) 67.3 (15) 0.87
Two-hundred nineteen ischemic stroke patients received rt-PA at
the London Health Sciences Centre between January 1999 and NIHSS at baseline was the NIHSS score on admission to the emergency
department.
March 2003. Demographic and clinical characteristics of this
cohort have been presented.11 They were 111 (51%) males, and
Logistic-regression analysis for MNI at 24 hours after the
the mean (SD) age was 71.5 (12) years, with 178 (81%) of the
IV rt-PA infusion and infarct side showed that female sex
patients being ⬎60 years old. London was the city of onset in
(OR, 3.00; 95% CI, 1.53 to 5.90; P⫽0.001) and LHS (OR,
129 (59%) patients, their median NIHSS (SD) was 13 (6), and 2.07; 95% CI, 1.05 to 4.08; P⫽0.03) were independent
195 (89%) had a nonlacunar stroke. From this initial cohort,
patients with brainstem or cerebellar infarcts, unknown side of TABLE 2. Logistic-Regression Analysis for 3-Month Outcome
the affected hemisphere, and lack of follow-up data were and Infarct Side, Adjusted by Age, Sex, Hypertension, and
excluded. One hundred sixty-five patients were eligible for this NIHSS at Baseline
analysis. Of them, 97 (59%) patients had LHS and 68 (41%)
mRS 0 –1 at 3 months OR P 95% CI
patients had RHS. Patients with LHS were more hypertensive
(P⫽0.001) and had higher NIHSS (P⫽0.005) (Table 1). LHS 2.29 0.019 1.14 4.59
Logistic-regression analysis for good outcome at 3 months Age 0.96 0.012 0.93 0.99
(mRS 0 –1) and infarct side showed that LHS independently Female sex 1.25 0.509 0.63 2.48
predicted good outcome at 3 months (odds ratio [OR], 2.29; NIHSS at baseline 0.83 ⬍0.0001 0.78 0.89
95% CI, 1.14 to 4.59), after adjusting for age, sex, hyperten- Hypertension 1.18 0.623 0.59 2.37
sion, and baseline NIHSS. Age (OR, 0.96; 95%, CI 0.93 to NIHSS at baseline was the NIHSS score on admission to the emergency
0.99; P⫽0.012) and pretreatment NIHSS (OR, 0.83; 95% CI, department.
0.78 to 0.89; P⬍0.0001) were also negative predictors of Goodness of fit test: P⫽0.30.
good outcome at 3 months (Table 2). Area under the receiver operating characteristic curve⫽0.79.

Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016

2068 Stroke August 2006
TABLE 3. Logistic-Regression Analysis for MNI and Infarct
Side, Adjusted by Age, Sex, Hypertension, and NIHSS
at Baseline
MNI OR P 95% CI
LHS 2.07 0.034 1.05
Age 0.98 0.312 0.959
Female sex 3.00 0.001 1.53
NIHSS at baseline 0.97 0.298 0.92
Hypertension 1.72 0.119 0.86
NIHSS at baseline was the NIHSS score on admission to the emergency
department.
Goodness of fit test: P⫽0.39.
Area under the receiver operating characteristic curve⫽0.68.
predictors of MNI after adjusting for age, hypertension, and
baseline NIHSS (Table 3).
Discussion
In this cohort of patients with acute ischemic strokes, we
observed that patients with LHS had significantly higher NIHSS
scores on admission and were more frequently hypertensive
compared with RHS patients. Older age and higher NIHSS
scores on admission were found to be independent predictors of
poor outcome at 3 months, whereas patients with LHS had a
higher chance of good outcome at 3 months compared with RHS
patients. Female sex and LHS also predicted MNI at 24 hours
after IV rt-PA administration.
A negative effect of age on patient outcome after IV
thrombolysis has been described,12 whereas data on the impact
of initial stroke severity as measured by the NIHSS on patient
outcome are controversial.11,12,14 We discovered that, after cor-
rection for initial stroke severity, patients with LHS had a 2-fold
increased likelihood of a good outcome at 3 months after IV
thrombolysis than did patients with RHS. Left hemispheric
damage and female sex also conferred a higher chance of MNI
at 24 hours, which is an indicator of recanalization15,16 and a
predictor of excellent outcome at 3 months.11,12,17 Although
women have more severe strokes and worse outcomes than
men,18 women seem to benefit more from IV thrombolysis than
men.11,19 Right hemispheric damage has been also associated
with poor outcome.5,6 Owing to a combination of underdiagno-
sis, delayed arrival to an emergency department, and consoli-
dated use of stroke severity scales biased toward left hemisphere
damage, patients with RHS are less likely to be treated with
rt-PA compared with those with LHS.1,2 However, we previ-
ously observed that the presence of visuospatial neglect on
admission confers a 2-fold increased probability of receiving IV
rt-PA,2 meaning that patients with RHS are adequately treated if
symptoms are severe enough to be recognized.
The reasons why patients with RHS should benefit less than
those with LHS from IV thrombolysis are not clear. One
possibility is that the available stroke assessment tools may be
inherently poor at correctly measuring RHS function, and this
limitation may have confounded our findings. For instance, a
1-point NIHSS improvement in the language item (ie, from
severe to moderate aphasia) can reflect changes in other items,
eg, the level of consciousness questions and commands. Con-
versely, the clinical spectrum of the neglect syndrome can be
Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016
defined only within a narrow score, and for the most severe
neglect, no fixed rules are applied to other items such as gaze
preference, visual field, and level of consciousness. Neurological
improvement might be “magnified” by the structure of the
4.08
NIHSS for patients with damage in the left hemisphere. Alter-
natively, physiological and cellular differences between the
1.01
brain hemispheres might differently influence blood flow and
5.90
thus, neuronal functional recovery after an ischemic injury. The
1.02 notion of “competition” between the hemispheres and the
3.40 negative influence of activity in the intact hemisphere are
emerging from functional MRI studies as important principles
for understanding recovery.20 In our cohort, patients with LHS
were more frequently hypertensive before the stroke compared
with patients with RHS. Although this might be explained by the
small sample size, cerebral asymmetry in sympathetic and
cardiac autonomic control has been demonstrated,21,22 and a
possible role of these variables on patient outcome deserves
further investigation.
Factors other than those reported might account for the better
outcome we observed in the left hemispheric group. Differences
between the 2 groups in terms of stroke subtype, site of arterial
occlusion, or the presence of a diffusion-perfusion mismatch on
MRI might have influenced our results. Although no differences
were found in stroke etiology between the 2 groups, the site of
occlusion and the presence of mismatch were not evaluated
before IV rt-PA administration, with the exception of a very few
selected cases. Furthermore, early ischemic changes on baseline
CT were not entered into our rt-PA database in a standardized
fashion. On the basis of the information available in this study,
we cannot provide definite evidence that the 2 groups were well
matched for those hemodynamic and neuroimaging characteris-
tics (ie, site of arterial occlusion, presence of diffusion-perfusion
mismatch) known to strongly influence outcome. Unfortunately,
previous studies aimed at investigating predictors of outcome
after IV thrombolysis in patients with documented middle
cerebral artery occlusion or aimed at correlating stroke subtypes
with different patterns of perfusion-diffusion MRI mismatch
have not looked at hemispherical differences.23–25
Summary
We speculate that some patients with RHS may be less likely
to benefit from IV rt-PA administration than patients with
LHS, even if the former are promptly recognized and ade-
quately treated. The bias introduced by the NIHSS in evalu-
ating stroke severity in patients with RHS can hamper proper
evaluation of the clinical gain in these patients: language
improvement is reflected by changes in several items on the
NIHSS, whereas the score for neglect only allows a narrow
range of changes. The observed results need to be confirmed
by studies coupling clinical information with those hemody-
namic and MRI characteristics that are strong determinants of
patient outcome.
Acknowledgments
The authors thank the Stroke fellows and the Stroke research nurses
at the LHSC for their help in assessing patients and collecting data.
Disclosures
None.
 Di Legge et al Outcome Predictors in Hemispheric Stroke Patients
References
1. Foerch C, Misselwitz B, Sitzer M, Berger K, Steinmetz H, Neumann-
Haefelin T; for the Arbeitsgruppe Schlaganfall Hessen. Difference in
recognition of right and left hemispheric stroke. Lancet. 2005;366:
392–393.
2. Di Legge S, Fang J, Saposnik G, Hachinski V. The impact of lesion side
on acute stroke treatment. Neurology. 2005;65:81– 86.
3. Fink JN, Selin MH, Kumar S, Silver B, Linfante I, Caplan LR, Schlaugh
G. Is the association of National Institutes of Health Stroke Scale scores
and acute magnetic resonance imaging stroke volume equal for patients
with right- and left-hemispheric stroke? Stroke. 2002;33:954 –958.
4. Lyden P, Claesson L, Havstad S, Ashwood T, Lu M. Factor analysis of
the National Institutes of Health Stroke Scale in patients with large
strokes. Arch Neurol. 2004;61:1677–1680.
5. Ween JE, Alexander MP, D’Esposito M, Roberts M. Factors predictive of
stroke outcome in a rehabilitation setting. Neurology. 1996;47:388 –392.
6. Aszalos Z, Barsi P, Vitrai J, Nagy Z. Lateralization as a factor in the
prognosis of middle cerebral artery territorial infarct. Eur Neurol. 2002;
48:141–145.
7. Merino JG, Silver B, Wong E, Foell B, Demaerschalk B, Tamayo A,
Poncha F, Hachinski V. Extending tissue plasminogen activator use to
community and rural stroke patients. Stroke. 2002;33:141–146.
8. National Institutes of Health. The Framingham Study: an epidemiological
investigation of cardiovascular disease. Washington, DC: US Dept of
Health, Education, and Welfare; 1971;section 27:1– 42.
9. Tissue plasminogen activator for acute ischemic stroke: the National
Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group.
N Engl J Med. 1995;333:1581–1587.
10. Saposnik G, Young B, Silver B, Di Legge S, Webster F, Beletsky V, Jain
V, Nilanont Y, Hachinski V. Lack of improvement in patients with acute
stroke after treatment with thrombolytic therapy: predictors and asso-
ciation with outcome. JAMA. 2004;292:1839 –1844.
11. Saposnik G, Di Legge S, Webster F, Hachinski V. Predictors of major
neurological improvement after thrombolytic therapy for acute stroke.
Neurology. 2005;65:1169 –1174.
12. Brown DL, Johnston KC, Wagner DP, Haley EC. Predicting major
neurological improvement with intravenous recombinant tissue plasmin-
ogen activator treatment of stroke. JAMA. 2004;35:147–150.
13. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Inter-
observer agreement for the assessment of handicap in stroke patients.
Stroke. 1998;19:604 – 607.
14. Kwiatkowski T, Libman R, Tilley BC, Lewandowski C, Grotta JC, Lyden
P, Levine SR, Brott T. National Institute of Neurological Disorders and
Stroke Recombinant Tissue Plasminogen Activator Stroke Study Group.
Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016
2069
The impact of imbalances in baseline stroke severity on outcome in the
National Institute of Neurological Disorders and Stroke Recombinant
Tissue Plasminogen Activator Stroke Study. Ann Emerg Med. 2005;45:
377–384.
15. Demchuk AM, Burgin WS, Christou I, Felberg RA, Barber PA, Hill MD,
Alexandrov AV. Thrombolysis in brain ischemia (TIBI) transcranial
Doppler flow grades predict clinical severity, early recovery, and mor-
tality in patients treated with intravenous tissue plasminogen activator.
Stroke. 2001;32:89 –93.
16. Felberg RA, Okon NJ, El-Mitwalli A, Burgin WS, Grotta JC, Alexandrov
AV. Early dramatic recovery during intravenous tissue plasminogen ac-
tivator infusion: clinical pattern and outcome in acute middle cerebral
artery stroke. Stroke. 2002;33:1301–1307.
17. Marler JR, Tilley BC, Lu M, Brott TG, Lyden PC, Grotta JC, Broderick JP,
Levine SR, Frankel MP, Horowitz SH, Haley EC Jr, Lewandowski CA,
Kwiatkowski TP. Early stroke treatment associated with better outcome: the
NINDS rt-PA stroke study. Neurology. 2000;55:1649–1655.
18. Glader EL, Stegmayr B, Norrving B, Terent A, Hulter-Asberg K, Wester
PO, Asplund K, Riks-Stroke Collaboration. Sex differences in man-
agement and outcome after stroke: a Swedish national perspective.
Stroke. 2003;34:1970 –1975.
19. Kent DM, Price LL, Ringleb P, Hill MD, Selker HP. Sex-based dif-
ferences in response to recombinant tissue plasminogen activator in acute
ischemic stroke: a pooled analysis of randomized clinical trials. Stroke.
2005;36:62– 65.
20. Corbetta M, Kincade MJ, Lewis C, Snyder AZ, Sapir A. Neural basis and
recovery of spatial attention deficits in spatial neglect. Nat Neurosci.
2005;8:1603–1610.
21. Yoon BW, Morillo CA, Cechetto DF, Hachinski V. Cerebral hemispheric
lateralization in cardiac autonomic control. Arch Neurol. 1997;54:
741–744.
22. Helms AK, Torbey MT, Hossein Almassi G. Determinants of mortality in
stroke patients with right brain damage. Curr Atheroscler Rep. 2005;7:
289 –295.
23. Restrepo L, Jacobs MA, Barker PB, Beauchamp NJ, Skolasky RL,
Keswani SC, Wityk RJ. Etiology of perfusion-diffusion magnetic res-
onance imaging mismatch patterns. J Neuroimaging. 2005;15:254 –260.
24. Rubiera M, Alvarez-Sabin J, Ribo M, Montaner J, Santamarina E, Arenillas
JF, Huertas R, Delgado P, Purroy F, Molina CA. Predictors of early arterial
reocclusion after tissue plasminogen activator-induced recanalization in acute
ischemic stroke. Stroke. 2005;36:1452–1456.
25. Fischer U, Arnold M, Nedeltchev K, Brekenfeld C, Ballinari P, Remonda
L, Schroth G, Mattle HP. NIHSS score and arteriographic findings in
acute ischemic stroke. Stroke. 2005;36:2121–2125.
 Neglecting the Difference: Does Right or Left Matter in Stroke Outcome After
Thrombolysis?
Silvia Di Legge, Gustavo Saposnik, Yongchai Nilanont and Vladimir Hachinski

Stroke. 2006;37:2066-2069; originally published online June 22, 2006;

doi: 10.1161/01.STR.0000229899.66019.62
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2006 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/37/8/2066

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Stroke is online at:

http://stroke.ahajournals.org//subscriptions/

Downloaded from http://stroke.ahajournals.org/ by guest on June 24, 2016