Professional Documents
Culture Documents
Background and Purpose—The importance of perioperative cognitive decline has long been debated. We recently
demonstrated a significant correlation between perioperative cognitive decline and long-term cognitive dysfunction.
Despite this association, some still question the importance of these changes in cognitive function to the quality of life
of patients and their families. The purpose of our investigation was to determine the association between cognitive
dysfunction and long-term quality of life after cardiac surgery.
Methods—After institutional review board approval and patient informed consent, 261 patients undergoing cardiac surgery
with cardiopulmonary bypass were enrolled and followed for 5 years. Cognitive function was measured with a battery
of tests at baseline, discharge, and 6 weeks and 5 years postoperatively. Quality of life was assessed with well-validated,
standardized assessments at the 5-year end point.
Results—Our results demonstrate significant correlations between cognitive function and quality of life in patients after
cardiac surgery. Lower 5-year overall cognitive function scores were associated with lower general health and a less
productive working status. Multivariable logistic and linear regression controlling for age, sex, education, and diabetes
confirmed this strong association in the majority of areas of quality of life.
Conclusions—Five years after cardiac surgery, there is a strong relationship between neurocognitive functioning and
quality of life. This has important social and financial implications for preoperative evaluation and postoperative care
of patients undergoing cardiac surgery. (Stroke. 2001;32:2874-2881.)
Key Words: cardiac surgical procedures 䡲 cognitive disorders 䡲 quality of life
Received December 28, 2000; final revision received June 28, 2001; accepted September 5, 2001.
From the Department of Anesthesiology (M.F.N., H.P.G., J.P.M., W.D.W., J.G.R.), Department of Surgery (K.L.), Department of Medicine, Division
of Neurology (D.T.L.), Department of Medicine, Division of Cardiology (D.B.M.), and Department of Psychiatry and Behavioral Sciences (J.A.B.), Duke
University Medical Center, Durham, NC.
A complete list of the participants in the Neurologic Outcome Research Group and the CARE Investigators of the Duke Heart Center appears in
Appendix 1 and Appendix 2.
Reprint requests to Mark F. Newman, MD, Department of Anesthesiology, Box 3094, DUMC, Duke University Medical Center, Durham NC 27710.
E-mail newma005@mc.duke.edu
© 2001 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org
2874
Downloaded from http://stroke.ahajournals.org/ by guest on June 25, 2016
Newman et al Neurocognitive Function and Quality of Life After CABG 2875
sleep problems), sexual (eg, decreased pain and worry, functional status (minimum 0, maximum 58.2). A higher weighted
increased desire and energy), social (eg, increased participa- score is better.
(2) The Medical Outcomes Study 36-Item Short Form Health
tion in social activities and hobbies), and work activity (return Survey (SF-36).21,22 The SF-36 was designed to measure general
to work, increased job performance).12–14 Furthermore, 1 year health status. The results are expressed in terms of 8 subscores and
after surgery, patients report increased life satisfaction and 2 summary scores: the Physical Component Summary (PCS) and the
improvements in emotional well-being and family life.15 Mental Component Summary (MCS). SF-36 items and scales are
However, these studies have not included detailed serial scored so that a high score indicates a better health state. Summary
scores have been standardized to the US general population (mean
psychometric testing to assess the role of cognitive function. score of 50⫾10) to allow easier norm-based interpretation.
To determine the clinical importance of cognitive dysfunc- (3) Center for Epidemiological Studies Depression Scale (CES-
tion after cardiac surgery, we examined these changes in the D).23 The CES-D is a 20-item self-report measure designed to
context of other aspects of quality of life in patients with measure symptoms of depression. Subjects rate the degree to which
they have experienced a range of symptoms of depression on items
cardiovascular disease undergoing treatment. The purpose of such as “I had crying spells” and “I felt lonely.” Scores range from
our investigation was to determine the impact of neurocog- 0 to 60, with higher scores indicating greater depressive symptoms.
nitive dysfunction on quality of life after cardiac surgery. Scores ⬎16 are typically considered indicative of clinically signifi-
cant depression.
(4) Spielberger State and Trait Anxiety Inventory (STAI).24 The
Subjects and Methods STAI consists of two 20-item scales that measure state (current) and
Patient Enrollment trait (chronic) anxiety. Representative items include statements such
After institutional review board approval and written informed as “I feel nervous” and “I feel worried.” These items are rated on a
consent were obtained, 261 patients undergoing elective cardiac 4-point scale of how well they describe the patient’s current or
surgery were enrolled in this investigation. Patients with a history of typical mood from “not at all” to “very much so.” Scores range from
symptomatic cerebrovascular disease (with residual deficit), psychi- 20 to 80, with higher scores indicating greater anxiety.
atric illness, renal disease (serum creatinine concentration ⬎2.0 (5) Mini-Mental State Examination (MMSE).25 This test is de-
mg/dL), active liver disease, less than a seventh-grade education, or signed to grossly assess executive cognitive functioning. It assesses
an inability to read were excluded from enrollment. orientation, memory, calculations, reading and writing capacity,
visuospatial ability, and language. Patients are quantitatively mea-
sured on those functions; a perfect score is 30 points, a score of ⬍25
Measurement of Neurocognitive Function suggests impairment, and a score of ⬍20 indicates definite
A brief neurocognitive test battery was administered before surgery impairment.25,26
(baseline), the day before discharge (approximately 7 days after (6) Perceived Social Support Scale.27 Twelve items include how
coronary artery bypass grafting [CABG]), and at 6 weeks, 6 months, strongly subjects agree there is “a special person who is around when
and 5 years postoperatively (results of neurocognitive outcome were I am in need” and “my family really tries to help me.” Choices range
previously reported).3 Assessments were performed individually by from “very strongly disagree” to “very strongly agree.” Items are
experienced psychometricians using a well-validated battery that summed for a range of 12 to 84, with a high score indicating a better
included the following: (1) The Short Story module of the Randt status.
Memory Test requires subjects to recall the details of a short story (7) Working Status. This is a single-item instrument with 9
immediately after it was read to them (immediate) and after a possible choices: 1, full time; 2, part time; 3, homemaker; 4,
30-minute delay (delay). Scoring is based on both the “verbatim” and long-term sick leave; 5, short-term sick leave; 6, retired; 7, disabled;
“gist” of the response on immediate and delayed testing (4 variable 8, unemployed; 9, other.
scores).16 (2) The Digit Span subtest of the Wechsler Adult Intelli-
gence Scale–Revised (WAIS-R) requires subjects to repeat a series Patient Management
of digits that have been orally presented to them both forward and, Anesthetic management with midazolam, fentanyl, isoflurane, and
in an independent test, in reverse order (2 variable scores).17 (3) The vecuronium has been previously described.28 The perfusion appara-
Benton Revised Visual Retention Test requires subjects to reproduce tus consisted of a Cobe CML oxygenator (COBE Chem Labs), a
from memory a series of geometric shapes after a 10-second Sarns 7000 max pump (Sarns Inc), and a Pall SP 3840 arterial line
exposure (1 variable score).18 (4) The Digit Symbol subtest of the filter (Pall Biomedical Products Co). Nonpulsatile perfusion of 2 to
WAIS-R is a paper-and-pencil task that requires subjects to repro- 2.4 L/min per square meter of body surface area was maintained
duce, within 90 seconds, as many coded symbols as possible in blank throughout the study periods. The pump was primed with crystalloid
boxes beneath randomly generated digits according to a coding solution designed to achieve a hematocrit of ⱖ0.18% during extra-
scheme for pairing digits with symbols (1 variable score).17 (5) The corporeal circulation. Packed red blood cells were added when
Trail-Making Test (Trails B) requires subjects to connect, by necessary to achieve the desired hematocrit. All patients underwent
drawing a line, a series of numbers and letters in sequence (ie, cardiopulmonary bypass (CPB) through an ascending aortic cannula.
1-A-2-B) as quickly as possible (1 variable score).19 Arterial carbon dioxide tension was maintained throughout CPB at
35 to 40 mm Hg (uncorrected for temperature), and partial pressure
Measurement of Quality of Life of oxygen (PaO2) was maintained at 150 to 250 mm Hg.
Quality of life instruments were administered individually by a
trained psychometrician who was blinded to the patient’s neurocog- Statistical Methods
nitive test results at the 5-year end point. Most of the questionnaires All assessment instruments were scored according to the specific
were self-administered; however, the measures were read to the validated algorithms described by the test developers. In particular, the
patient if he or she was feeling ill or was unable to read. The DASI uses a weighted sum, and the SF-36 uses weights and transfor-
following quality of life measures were used: mations to arrive at its 8 scales and 2 summary components.21
(1) The Duke Activity Status Index (DASI).20 The DASI is a To assess neurocognitive decline over time while minimizing the
12-item scale of functional capacity that has been found to correlate potential for redundancy in the neurocognitive measures, a factor
well with objective measures of maximal exercise capacity. Items analysis with orthogonal rotation was first performed on the 9
reflect activities of personal care, ambulation, household tasks, individual baseline neurocognitive test scores. This analysis included
sexual function, and recreational activities. Activities reported to be scores from the entire baseline population of 261 patients. Factor
done “with no difficulty” receive scores weighted higher for more analysis was used as a variable reduction technique to reduce the
taxing activities, which are summed for a quantitative measure of larger number of correlated scores to a smaller number of uncorre-
lated variables to be used in the final analysis. The factor loadings demonstrate a significant relationship over and above the effects of
(weights) of each test on each factor were used to construct covariables. Because there was no single way to bring together our
comparable domain scores at the 5-year follow-up time periods on multiple assessment of quality of life, we have presented all of the
the basis of patients’ test scores for that time period. In this manner, univariate and multivariate probability values for comparison.
the domains were identified at baseline and remained consistent at
follow-up.3 Results
Factor analysis on 9 baseline neuropsychological test scores Two hundred sixty-one patients initially enrolled in the study
suggests that 4 factors account for 86% of the variance present in our
and underwent baseline neurocognitive testing. Of these, a
test battery at baseline. The 4 factors coherently represent the
cognitive domains of (1) verbal memory and language comprehen- total of 172 patients were available for 5-year follow-up and
sion (short-term and delayed); (2) abstraction and visuospatial completed both neurocognitive functioning and quality of life
orientation; (3) attention, psychomotor processing speed, and con- assessments (reasons for withdrawal of the 89 patients were
centration; and (4) visual memory. previously described).3 The demographics and background
A change score for each of the factors was calculated by clinical information of this patient population are listed in
subtracting baseline factor scores from the follow-up factor scores.
Categorically, a cognitive deficit was defined as a decline of 1 SD or Table 1. Demographic parameters of the entire population
more in performance on any 1 of the 4 domains. To quantify overall (baseline) and those available for 5-year follow-up were
cognitive function and the degree of learning (practice from repeated similar, with exceptions as noted.
exposure to the testing procedures) or cognitive decline across all
domains, a composite cognitive index was calculated as the sum of TABLE 2. Neurocognitive Function Scores at Baseline
the 4 domain scores to yield a single, continuous measure of
and 5 Years
cognitive function. This summary measure, accounting for improve-
ment as well as decline, was used to represent overall cognitive Baseline 5 Years
functioning as the predictor of interest in our models. Test/Score (n⫽261) (n⫽172)*
The association between quality of life and long-term cognitive
dysfunction was investigated univariately with each of the measures Digit Symbol† 38.78⫾13.71 38.00⫾14.94
of quality of life with the use of our continuous measure of overall Benton Revised Visual Retention 5.14⫾2.25 5.02⫾2.17
cognitive function. Continuous quality of life outcome measures
Randt Delayed Gist‡ 5.01⫾2.14 5.63⫾2.31
were tested with Pearson correlation tests and confirmed with
Spearman correlation tests to guard against nonnormal distributions. Randt Delayed Verbatim‡ 6.23⫾3.26 7.30⫾3.51
For analysis, the 6 responses to the “Work Status” item were grouped Randt Immediate Gist‡ 5.73⫾1.83 6.49⫾1.90
into a dichotomous response of “Yes” if work status was full time,
Randt Immediate Verbatim‡ 7.88⫾3.25 9.21⫾3.34
part time, or homemaker, and “No” if work status was reported as
retired, disabled, or unemployed. Logistic regression was then used Digit Span†
to test the effect of cognitive function on the dichotomous working Repeat Forward 5.43⫾2.26 5.30⫾2.19
outcome. Quality of life outcomes that showed a significant univar-
iate relationship with cognitive function at ␣⫽0.05 were subse- Repeat Backward 7.25⫾2.29 7.00⫾2.29
quently tested in multivariable linear or logistic regression models to Trail Making Test (Trails B) 142.24⫾73.56 184.54⫾146.69
account for the possible competing effects of age, sex, years of Composite cognitive index 0.183⫾1.944 ⫺0.133⫾2.321
education, diabetes, and the 2-way interactions of each of these with
cognitive function. In the multivariate models, the best predictive Values are mean⫾SD.
model without cognitive function was developed first, starting with *Number of patients still participating in trial (excludes withdrawals and
all covariables listed and removing nonsignificant terms iteratively patients lost to follow-up and death).
until only significant terms remained. The effect of cognitive †Subtest of WAIS-R.
function was then added to this “best” model to test whether it would ‡Short Story module of Randt Memory Test.
TABLE 3. Multivariable Adjusted Effects of Total 5-Year Cognitive Function Score (Factor Sum) on Quality
of Life Measures*
Quality of Life Spearman
Score, Correlation Multivariable Total
Quality of Life Outcome Mean⫾SD Coefficient R P† P‡ Covariates§ Model R 2
DASI weighted sum (higher⫽better) 25.5⫾16.8 0.289 0.0002 0.0085 Age, sex, DM 0.298
Cognitive Difficulties Scale (higher⫽worse) 43.7⫾27.1 ⫺0.308 ⬍0.0001 0.0046 Age, sex, DM, education 0.199
SF-36 Standardized PCS 41.0⫾11.3 0.228 0.0038 0.0170 Sex 0.105
Physical Functioning (10 items) 64.6⫾27.4 0.259 0.0009 0.0010 Sex, DM 0.216
Role—Physical (4 items) 61.8⫾40.2 0.173 0.0296 0.0058 Sex, DM 0.089
Pain Index (2 items) 67.6⫾26.7 0.102 0.1961
General Health Perceptions (5 items) 57.0⫾23.0 0.178 0.0231 0.0037 Age, sex 0.081
SF-36 Standardized MCS 52.7⫾10.1 0.024 0.7600 䡠䡠䡠 䡠䡠䡠 䡠䡠䡠
Vitality (4 items) 53.9⫾23.3 0.089 0.2607 䡠䡠䡠 䡠䡠䡠 䡠䡠䡠
Social Functioning (2 items) 80.2⫾25.8 0.057 0.4745 䡠䡠䡠 䡠䡠䡠 䡠䡠䡠
Role–Emotional (3 items) 78.6⫾33.7 0.146 0.0660 䡠䡠䡠 䡠䡠䡠 䡠䡠䡠
Mental Health Index (5 items) 78.4⫾19.0 0.085 0.2825 䡠䡠䡠 䡠䡠䡠 䡠䡠䡠
CES-D: Depression 10.2⫾9.9 ⫺0.254 0.0012 0.0006 Age, sex, DM 0.177
STAI: State Anxiety 33.0⫾11.8 ⫺0.026 0.7401
STAI: Trait Anxiety 35.0⫾11.5 ⫺0.179 0.0231 0.0071 Age, sex, DM, education 0.206
Perceived Social Support Scale 73.8⫾10.4 ⫺0.137 0.0827 䡠䡠䡠 䡠䡠䡠 䡠䡠䡠
MMSE 27.1⫾3.4 0.617 ⬍0.0001 0.0001 Age, sex 0.455
Working Status (32.4%⫽Yes) Odds ratio⫽1.4 ⬍0.0001㛳 0.0060 Age, sex C⫽0.786¶
*All models allowed for potential effects of age in years, sex, years of education, diabetes mellitus (DM), and preoperative left ventricular ejection fraction.
†Correlation with total 5-year cognitive function score.
‡Effect of cognitive function score in multivariable regression. All models were linear regression on continuous outcome measures except for
working status, which was a logistic regression on the binary outcome of working productively (yes, no).
§Covariates remaining significant in the final model.
㛳P-value from logistic regression with 5-year cognitive function score as single predictor (odds ratio for multiple testing⫽1.3).
¶Logistic model C index.
Editorial Comment
Cognitive Function and Quality of Life
Cognitive dysfunction after major surgery is not uncommon. CABG. Self-perceived health status/quality of life indices are
Cognitive dysfunction is not the same as delirium, encepha- useful as broad outcome measures of the impact of disease
lopathy, or an altered state of consciousness. Rather, it refers and interventions,11 and associations have been found be-
to a condition in which memory and intellectual abilities tween cognitive function and quality of life.12,13 The study by
seem impaired when the patient appears to have otherwise Newman et al represents an important addition to this
recovered from the immediate effects of surgery. It is a major literature by demonstrating significant correlations between
concern for the patient, family, and physician when a patient quality of life and cognitive function 5 years after CABG.
is found not to be intellectually the same on awakening after Lower cognitive status was associated with lower perceived
surgery as before. general health and less-productive work status. Unfortu-
The precise pathophysiological mechanisms for postoper- nately, however, the conclusions are limited by the lack of
ative cognitive decline are unknown, but are probably mul- baseline quality of life data, which makes comparisons with
tifactorial.1 Elderly patients undergoing major cardiac surgery postsurgical values impossible. Without this comparison, it is
(eg, coronary artery bypass grafting [CABG] and thoracic not possible to evaluate whether there has been a change in
vascular surgery)2 and noncardiac surgery (eg, orthopedic and quality of life after CABG.
abdominal)3 are at the greatest risk for postoperative cogni- In this study, 5-year quality of life measures (16 continuous
tive decline. Other individual patient features that increase the variables) were correlated with 5 cognitive indices, including
risk of postoperative cognitive dysfunction include previous the baseline cognitive index score (sum of 4 cognitive domain
cerebrovascular disease, undetected cognitive impairment or scores), change in composite cognitive index at 5 years, and
dementia, and cardiovascular risk factors such as hyperten- an absolute 5-year cognitive index score. All 3 composite
sion, diabetes, and peripheral vascular disease.4 Intraopera- indices correlated with the 5-year quality of life measures.
tive risk factors include surgical technique (eg, duration of The association between the 5-year cognitive index and
cardiopulmonary bypass and duration of aortic cross- quality of life at 5 years was reported to be more robust than
clamping), hypotension, manipulation of the diseased aorta, the association of quality of life with baseline cognitive
and the effects of general anesthesia and hypothermia. function or change in function in the univariate analyses.
Atherothromboembolic phenomena (microemboli) and hyp- It is surprising that baseline cognitive status would corre-
oxia with watershed area injury secondary to hypoperfusion late with 5-year quality of life measures. This raises the
are also possible etiological mechanisms.5–7 possibility that some of the patients did not have normal
Cognitive changes may be obvious when there are gross cognitive abilities before the surgery. It is also possible that
deficits in learning, memory, attention, or concentration. The the correlations represent spuriously significant results given
decline can also be subtle, with problems in initiative and the number of comparisons made. The latter explanation is
planning. These changes can persist well beyond the imme- likely because when the multiple cognitive indices were
diate postoperative period, when the effects of anesthesia and placed in the multivariate model, only the 5-year cognitive
analgesia directly affecting cognitive functions have clearly index was associated with quality of life.
worn off. Most mental status changes improve but can persist CABG is a common surgical procedure in the United
for months and years.2,8,9 States. As with other major surgeries, there is a risk of
In a recent report, Newman and colleagues8 found the cognitive changes that can persist for months and years.
incidence of cognitive decline after CABG to be 53% at Newman and colleagues eloquently demonstrated this in a
discharge, 36% at 6 weeks, 24% at 6 months, and 42% at 5 recent study.8 The intent of the current study to investigate
years. Cognitive decline was defined as a drop of 1 or more quality of life in patients who have undergone CABG is
standard deviations from baseline scores on tests in at least 1 important and laudable. However, due to the absence of
of 4 domains of cognitive function identified by factor baseline quality of life data, the conclusion should be limited
analysis. Older age, lower level of education, and evidence of to the association of 5-year cognitive status and 5-year quality
cognitive decline at discharge were found to be significant of life.
predictors of long-term cognitive dysfunction. This suggests
that injury during surgery caused in some way cognitive Joan M. Swearer, PhD, Guest Editor
deficits 5 years later. In their accompanying editorial, Selnes Department of Neurology
and McKhann10 offer alternative explanations including the University of Massachusetts Medical School
possibility that patients who undergo CABG are not cogni- Worcester, Massachusetts
tively normal because of comorbid cerebrovascular disease.
1. Selnes OA, Goldsborough MA, Borowicz LM Jr, Enger C, Quaskey SA,
In the preceding article, Newman and colleagues have
McKhann GM. Determinants of cognitive change after coronary artery
expanded their earlier study to investigate the relationship bypass surgery: a multifactorial problem. Ann Thorac Surg. 1999;67:
between cognitive function and quality of life 5 years after 1669 –1676.
2. Borowicz LM, Goldsborough MA, Selnes OA, McKhann GM. Neuro- Group and the Cardiothoracic Anesthesiology Research Endeavors Investi-
psychologic change after cardiac surgery: a critical review. J Cardio- gators. Longitudinal assessment of neurocognitive function after coronary-
thorac Vasc Anesth. 1996;10:105–112. artery bypass surgery. N Engl J Med. 2001;344:395–402.
3. Moller JT, Cluitmans P, Rasmussen LS, Houx P, Rasmussen H, Canet J, 9. Selnes OA, Royall RM, Grega MA, Borowicz LM Jr, Quaskey S,
Rabbitt P, Jolles J, Larsen K, Hanning CD, Langeron O, Johnson T, McKhann GM. Cognitive changes 5 years after coronary artery bypass
Lauven PM, Kristensen PA, Biedler A, van Beem H, Fraidakis O, Sil- grafting: is there evidence of late decline? Arch Neurol. 2001; 58:
verstein JH, Beneken JE, Gravenstein JS. Long-term postoperative cog- 598 – 604.
nitive dysfunction in the elderly: ISPOCD1 study. Lancet. 1998;351: 10. Selnes OA, McKhann GM. Coronary-artery bypass surgery and the brain.
857– 861. N Engl J Med. 2001;344:451– 452.
4. Newman MF, Croughwell ND, Blumenthal JA, Lowry E, White WD, 11. Rumsfeld JS, MaWhinney S, McCarthy M Jr, Shroyer AL, VillaNueva
Spillane W, Davis RD Jr, Glower DD, Smith LR, Mahanna EP, et al.
CB, O’Brien M, Moritz TE, Henderson WG, Grover FL, Sethi GK,
Predictors of cognitive decline after cardiac operation. Ann Thorac Surg.
Hammermeister KE; participants of the Department of Veterans Affairs
1995;59:1326 –1330.
Cooperative Study Group on Processes, Structures, and Outcomes of Care
5. Caplan LR, Hennerici M. Impaired clearance of emboli (washout) is an
important link between hypoperfusion, embolism, and ischemic stroke. in Cardiac Surgery. Health-related quality of life as a predictor of mor-
Arch Neurol. 1998;55:1475–1482. tality following coronary artery bypass graft surgery. JAMA. 1999;281:
6. Moody DM, Bell MA, Challa VR, Johnston WE, Prough DS. Brain 1298 –1303.
microemboli during cardiac surgery or aortography. Ann Neurol. 1990; 12. Cohen RA, Moser DJ, Clark MM, Aloia MS, Cargill BR, Stefanik S,
28:477– 486. Albrecht A, Tilkemeier P, Forman DE. Neurocognitive functioning and
7. Williams-Russo P, Sharrock NE, Mattis S, Szatrowski TP, Charlson ME. improvement in quality of life following participation in cardiac rehabil-
Cognitive effects after epidural vs general anesthesia in older adults: a itation. Am J Cardiol. 1999;83:1374 –1378.
randomized trial. JAMA. 1995;274:44 –50. 13. Lloyd AJ, Boyle J, Bell PRF, Thompson MM. Comparison of cognitive
8. Newman MF, Kirchner JL, Phillips-Bute B, Gaver V, Grocott H, Jones RH, function and quality of life after endovascular or conventional aortic
Mark DB, Reves JG, Blumenthal JA; Neurological Outcome Research aneurysm repair. Br J Surg. 2000;87:443– 447.
Stroke. 2001;32:2874-2881
doi: 10.1161/hs1201.099803
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2001 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/32/12/2874
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.