Outcomes of carotid endarterectomy for
asymptomatic stenosis in Sweden are improving:
Results from a population-based registry
Björn Kragsterman, MD,a Håkan Pärsson, MD, PhD,a Johan Lindbäck, MSc,b
David Bergqvist, MD, PhD,a and Martin Björck, MD, PhD,a on behalf of the Swedish Vascular
Registry (Swedvasc), Uppsala, Sweden
Objective: In large randomized trials, carotid endarterectomy (CEA) for asymptomatic stenosis has shown a net benefit
compared with best medical treatment. To justify an increased number of procedures for this indication, the perioperative
risk of stroke or death must not exceed that of the trials. The aim of this study was to evaluate the outcome in routine
clinical practice in Sweden in a population-based study.
Methods: The Swedish Vascular Registry (Swedvasc) covers all centers performing CEA. Data on all registered CEAs
during 1994 to 2003 were analyzed both for the whole time period and for two 5-year periods to study alterations over
time. Four validation procedures of the registry were performed. Medical records were reviewed for both a random
sample and a target sample (a total of 12% of the CEAs for asymptomatic stenosis). Swedvasc data were cross-matched
with the In-Patient-Registry (used for reimbursement) and the Population-registry (death).
Results: A total of 6182 CEAs were registered, 671 being for asymptomatic stenosis. In the validation process, no missed
registration of major stroke or death was found. Patients with asymptomatic stenosis had, when the whole time-period
was analyzed, a perioperative combined stroke or death rate of 2.1%. Outcome improved over time; the combined stroke
or death rate decreased from 3.3% (11/330) from 1994 to 1998 to 0.9% (3/341) from 1999 to 2003 (P ⴝ .026). During
the second time period, no patient with a perioperative major stroke or death was reported.
Conclusions: This extensively validated national audit of CEA for patients with asymptomatic carotid artery stenosis
showed results well comparable with those of the randomized trials. The results improved over time. ( J Vasc Surg 2006;
44:79-85.)
The efficacy of carotid endarterectomy (CEA) in preventing stroke among patients with severe asymptomatic
stenosis has been demonstrated in large randomized trials.1-3 Compared with those operated on for a symptomatic
carotid artery stenosis,4-7 the perioperative morbidity in
patients with an asymptomatic stenosis has an even more
pronounced influence on the overall benefit of the procedure. On implementing CEA for asymptomatic stenosis,
the results in every day clinical practice must be in agreement with the results from the randomized trials.
Until 2003 in Sweden, about 12% of the CEAs were
performed for asymptomatic stenosis, but after the results
from the Asymptomatic Carotid Surgery Trial (ACST)
were presented,2 the proportion increased to 20% in 2004.8
To justify this increase in CEA for asymptomatic stenosis,
the perioperative results need to be evaluated. The Swedish
Vascular Registry (Swedvasc) covers all centers performing
CEA in the country. The aim of this study was to audit the
perioperative outcome of this procedure for asymptomatic
From the Department of Surgical Sciences, Uppsala University Hospital,a
and Uppsala Clinical Research Center, University Hospital.b
This study was supported by a grant from Gorthon Foundation, Helsingborg, Sweden.
Competition of interest: none.
Reprint requests: Björn Kragsterman, Section of Surgery, Department of
surgical sciences, Uppsala University hospital, SE-751 85 Uppsala, Sweden (e-mail: bjorn.kragsterman@surgsci.uu.se).
0741-5214/$32.00
Copyright © 2006 by The Society for Vascular Surgery.
doi:10.1016/j.jvs.2006.03.003
stenosis on a population basis. The audit also included
validation of the registry.
MATERIAL AND METHODS
CEA. The Swedish Vascular Registry (Swedvasc)
started as a regional registry in 1987 and has had national
coverage since 1994, thus including all centers performing
CEA in the country (population about 9 million). Data
were retrieved on all elective CEAs registered during a
10-year period (1994 to 2003). No combined procedures
(ie, with thoracic surgery) or procedures for restenosis were
included.
In this study, the definition of asymptomatic stenosis
excluded all ipsilateral carotid artery events and nonhemispheric symptoms such as posterior circulation events and
nonspecific symptoms ⱕ6 months.
Comparison was made between the cohorts operated
on for asymptomatic vs symptomatic stenosis regarding
baseline characteristics and perioperative complications (30
days). Data were analyzed during two 5-year periods (1994
to1998 and 1999 to 2003), to study changes over time
regarding treatment and outcome of the patients with
asymptomatic stenosis.
The study received ethics approval from the registry
review board, which according to Swedish law has the
authority concerning registry data.
Swedvasc. Swedvasc contains prospectively registered basic demographic data, risk factors, surgical tech79
JOURNAL OF VASCULAR SURGERY
July 2006
80 Kragsterman et al
nique, and postoperative outcome at 30 days. Data on
the degree of stenosis, status of the contralateral carotid
artery, synchronous procedures, antiplatelet and anticoagulation therapy, or type of anesthesia are not included
in the registry. The register has been described in detail
previously.9
Postoperative neurologic deficits are classified as either
ipsilateral carotid territory events consisting of (1) transitory ischemic attack (TIA), including amaurosis fugax, (2)
transient stroke (deficit ⬎24 hours and ⬍30 days), (3)
permanent stroke (deficit ⬎30 days); or as other perioperative neurologic or cerebrovascular events, including all
perioperative neurologic events not classified as an ipsilateral carotid territory event, for example contralateral carotid territory events or posterior circulation events. Local
nerve injuries, hematomas, and wound infection are registered.
Risk factors in Swedvasc have had the same definitions
throughout the study-period. Cerebrovascular disease is all
present or previous neurologic events (stroke, TIA, or
amaurosis fugax); diabetes, treatment with insulin, oral
medication, or diet; hypertension, medication or a diastolic
blood-pressure ⱖ110; cardiac disease, previous myocardial
infarction, atrial fibrillation, heart failure, angina pectoris,
coronary artery bypass, operation for valvular disease or
signs of ischemia on an electrocardiogram; previous vascular surgery, open or endovascular procedure or amputation
for vascular disease; renal insufficiency, a serum creatinine
⬎150 ␮mol/L; and smoking, regular smoking ⱕ5 years.
Validation. All patients were cross-matched with the
National Population-registry, which was possible by the
unique personal identity code of every Swedish citizen.
This results in 100% accurate data on mortality and date of
death.
From the cohort of patients operated on for asymptomatic stenosis, we collected complete medical records for
both a general and a targeted validation procedure. First, a
random sample of approximately 10% of all procedures was
identified, using the SPSS (SPSS Inc, Chicago, Ill) “random sample of cases” functions. Second, unclear complication registrations were identified in the entire asymptomatic cohort. These included (1) “perioperative thrombosis
or occlusion” and (2) “peripheral embolization” registered
without neurologic complication, as well as those with a
registration of (3) cerebrovascular event and also those (4)
lacking a registration of 30-day complications (missing
field).
A fourth validation procedure was performed to evaluate how many CEAs may not have been registered in
Swedvasc. Permission was granted from the National Board
of Health and Welfare to compare the number of operations in Swedvasc with that of the In-Patient Registry (IPR)
for each hospital for the year 2003. The IPR is used for
reimbursement, and the hospitals have an economic incentive to register their operations in that registry. Unfortunately, we were not granted permission to compare data on
the individual level, as cross-matching data registries using
the unique personal identity code is restricted by Swedish
law. The validation process is also illustrated in the Figure.
In the validation process, unregistered or incorrectly
registered permanent strokes were also classified into minor
or major strokes in an attempt to evaluate the severity of
neurologic deficit in these “missed” registrations. The classification of functional status was retrospective and based
on the data in the medical records. Minor stroke was defined
as minimal focal neurologic deficit persisting ⬎24 hours
but not leading to disability or significant impairment in
activities of daily living. Major stroke was defined as a deficit
lasting ⬎30 days and inducing a change in lifestyle.
Statistics. Analyses were done using SPSS 12.0.1 and
R 2.2 (R-Project for Statistical Computing, Vienna, Austria)10 Continuous variables were summarized by the median (and quartiles), and categoric variables were summarized by percentages. Continuous variables were compared
with the Wilcoxon-Mann-Whitney test and categoric variables with the Pearson ␹2 test. Analyses of risk factors for
the different combined outcomes (listed in Table I) were
performed using multiple logistic regression models. The
selected model was designed based on the number of
adverse events, the completeness of the variables in the
database, and on the clinical relevance. As the numbers of
adverse evens were small, the numbers of variables that may
be included in the model were limited (recommended to be
one variable per ⱖ10 events) Results are presented as odds
ratios (OR) with 95% confidence intervals (CI). A statement of statistically significant implies a P ⬍ .05.
RESULTS
Validation. In the cross-match with the National
Population-registry, all perioperative deaths had been reported to the Swedvasc. Complete medical records were
obtained for 82 (89%) of 92 of the random sample and
target validation, and in reviewing the records we found
two permanent strokes and two transient strokes incorrectly registered, and one transient stroke and one TIA not
registered (Table II) The two incorrectly registered permanent strokes were considered as minor strokes, according to
the severity classification (in Methods).
During 2003, 771 CEAs were registered in Swedvasc
and 782 patients in the IPR. Three hospitals had exactly the
same number of procedures in both registries; 10 had more
in Swedvasc and eight had more in the IPR. In those latter
eight hospitals, 32 surplus operations were registered in the
IPR. A total of 4.1% (32/782) of those in the IPR were
unregistered in Swedvasc. (Cross-match on individual level
was not permitted—see Methods)
The registry was corrected regarding complications
before the analysis within the asymptomatic cohort. In
comparison with the (uncorrected) symptomatic cohort,
both the uncorrected and corrected data were presented.
Thus, we did not identify any unregistered death or major
stroke in the validation process.
CEA. CEA was performed in 22 of 44 vascular units
reporting to the registry, and of these, five did not operate
on patients with asymptomatic stenosis. During this 10-
JOURNAL OF VASCULAR SURGERY
Volume 44, Number 1
Fig. The four different validation procedures of the Swedvasc
registrations for carotid endarterectomy (CEA) of asymptomatic
stenosis.
year period (1994 to 2003), 6182 carotid endarterectomies
on 5821 patients were registered, the numbers with indication asymptomatic stenosis were 671 (10.9%) of 6182
CEAs on 636 (10.9%) of 5821 patients. Bilateral CEAs
were performed on 361 (6%) of 5821 patients, of which
152 (24%) of 636 were among patients with at least one
asymptomatic stenosis. Operations for bilateral asymptomatic lesions were performed in 35 (6%) of 636 patients.
Bilateral operations were staged.
Mean age was 70 years (quartiles 64, 75) for both the
symptomatic and asymptomatic group. There were no gender differences; men comprised 63.8% of patients with
asymptomatic stenosis and 66.6% of those with symptomatic stenosis (P ⫽ .16). The differences in baseline characteristics for the whole time period, with respect to indication, are presented in Table III.
The indications for the symptomatic patients were minor
stroke in 40%, TIA in 37%, amaurosis fugax in 20%, and
nonhemispheric symptoms in 2%. Patch closure was performed in 34% for asymptomatic vs 32% for symptomatic
lesions (P ⫽ .199).
Kragsterman et al 81
The perioperative morbidity and mortality rates are
shown in Table IV, with a combined stroke (permanent and
transient) or death rate of 1.6% for (uncorrected) asymptomatic vs 4.3% for the symptomatic group (P ⫽ .001). Two fatal
strokes (of 3 deaths) occurred among the procedures for
asymptomatic and 17 (of 74 deaths) among those for symptomatic stenosis.
Local nerve injury occurred with similar rate for the two
groups, with 6.7% for asymptomatic vs 7.0% for symptomatic stenosis (P ⫽ .762). Postoperative bleeding or hematomas were registered for 6.4% of asymptomatic vs 6.4% of
symptomatic lesions (P ⫽ .506). The wound infections
rates were 0.6% for the asymptomatic vs 0.9% for the
symptomatic indication (P ⫽ .413).
Among the patients with asymptomatic stenosis, many
were randomized into the ACST. Approximately half of the
532 Swedish patients who were recruited would have been
allocated to immediate CEA (266 ACST/671 Swedvasc ⫽
40%).
CEA for asymptomatic stenosis. Data from the
asymptomatic cohort was divided into two 5-year periods
(1994 to 1998 and 1999 to 2003). During the first period,
330 operations were performed (12.6% of the total number
of CEAs during that time period), compared with 341
procedures (11.7%) during the second. Baseline characteristics were comparable between the two periods, except
that hyperlipidemia (P ⬍ .001) and hypertension (P ⫽
.002) were more frequent in the second period (Table V).
Patch closure was more often used during the second
period (28 % vs 40 %, P ⫽ .001). The frequency of eversion
endarterectomy increased over time (0.3% vs 5.9%, P ⬍ .001)
Perioperative morbidity and mortality rates are shown
in Table I. There were no deaths and only one permanent
(minor) stroke during the second period
Multiple regression analysis. Logistic regression
analysis was performed testing the factors of age, gender,
and patch for different combinations of outcome. When all
stroke, TIA, cardiovascular complications, and death were
used as outcome in the model, the only significant variable
Table I. Perioperative morbidity (neurologic and cardiovascular) and mortality for carotid endarterectomy with
indication asymptomatic stenosis during the two time periods
Complication
1994-1998
n ⫽ 330
1999-2003
n ⫽ 341
P
Permanent stroke
Transient stroke
Stroke (permanent and transient)
TIA
Cerebrovascular event
Death
Cardiovascular complication
Permanent stroke and death
Stroke (permanent and transient), death
Stroke, TIA, death
Stroke, TIA, cerebrovascular event, death
Stroke, TIA, cardiovascular complication, death
1.5% (5)
1.5% (5)
3.0% (10)
0.3% (1)
1.8% (6)
0.9% (3)
4.6% (15)
1.8% (6)
3.3% (11)
3.6% (12)
4.2% (14)
7.6% (25)
0.3% (1)
0.6% (2)
0.9% (3)
0.3% (1)
1.5% (5)
0.0% (0)
3.8% (13)
0.3% (1)
0.9% (3)
1.2% (4)
2.3% (8)
4.7% (16)
.093
.238
.044*
.983
.723
.078
.640
.052
.026*
.036*
.168
.119
TIA, Transient ischemic attack.
*Indicates significance (P ⬍ .05); Pearson ␹2 test.
JOURNAL OF VASCULAR SURGERY
July 2006
82 Kragsterman et al
Table II. Validation of perioperative neurological complication registration after CEA with indication asymptomatic
stenosis
Validation
Random sample
Target sample
Total
N patient-records
obtained/requested
65/71 (93%)
N incorrect
registrations
2
1. Thrombosis or
occlusion; 2/3
1
2. Peripheral embolization; 2/2
3. Cerebrovascular
event; 6/9
0
2
4. Missing 30-day
registration;
7/7
1
82/92 (89%)
5
Classification in
registry
Classification in
validation
1. Minor contralateral finger
dysfunction
(IV-V)
2. Contralateral
facial paresis
(corner of
mouth)
Contralateral arm
dysfunction
(minutes
duration)
NA
Transient stroke
Permanent stroke
Local neurology
Permanent stroke
Not registered
TIA
NA
NA
1. Ipsilateral
(worsening)
arm dysfunction (3 days
duration)
2. Contralateral
hand/arm dysfunction (⬍7
days duration)
Contralateral hand
dysfunction and
aphasia (⬍5
days duration)
Cerebrovascular
event
Transient stroke*
Cerebrovascular
event
Transient stroke
Not registered
Transient stroke
Neurologic deficit
TIA, transient ischemic attack; NA, not applicable.
Results from a random sample as well as target validation of unclear perioperative complication registrations (see Material and Method).
*Ipsilateral stroke but with a contralateral occlusion, most likely a peroperative hypoperfusion injury that should have been registered as a surgical complication
(transient stroke).
Table III. Baseline characteristics for asymptomatic and symptomatic patients undergoing carotid endarterectomy
during 1994 to 2003
Baseline characteristics
Asymptomatic stenosis,
n ⫽ 671 (n)
Symptomatic stenosis,
n ⫽ 5511 (n)
P
Cerebrovascular disease
Diabetes mellitus
Hyperlipidemia
Hypertension
Cardiac disease
Previous vascular surgery
Chronic pulmonary disease
Renal insufficiency
Smoking
62.5% (406)
19.2% (122)
54.0% (218)
65.1% (420)
47.4% (304)
39.5% (254)
8.6% (55)
6.0% (38)
41.4% (251)
92.5% (4926)*
18.9% (942)
49.8% (1399)
60.2% (3081)
44.7% (2282)
16.5% (829)
8.6% (425)
3.7% (179)
42.0% (1956)
⬍.001†
.864
.114
.017†
.184
⬍.001†
.935
.004†
.769
For definitions see Material and Methods.
*Among the patients with symptomatic stenosis registered as indication for the procedure, almost all should have the risk factor cerebrovascular disease, if these
preoperative registrations were complete and correct. The difference of 7.5% represents, almost unexceptionally, missing fields.
†
Indicates significance (P ⬍ .05), Pearson ␹2.
identified was age, where an OR of 1.67 (95% CI, 1.11 to
2.50) was found for each SD of 8.3 years (P ⫽ .01), which
implies that the patient at age 78.3 has a 67% increased risk
compared with the patient at age 70. It can also be described as an OR of 2.85 (95% CI, 1.44 to 5.62) among
those ⬎75 years of age. In the same model, women had an
OR of 1.45 (95% CI, 0.74 to 2.85) and patch OR of 0.77
(95% CI, 0.37 to 1.59).
DISCUSSION
The perioperative stroke or death rate for CEA of
asymptomatic stenosis in this validated population-based
JOURNAL OF VASCULAR SURGERY
Volume 44, Number 1
Kragsterman et al 83
Table IV. Perioperative morbidity (neurological and cardiovascular) and mortality for carotid endarterectomy for the
whole time period (1994 to 2003)
Complication
Symptomatic stenosis
n ⫽ 5511 (not
corrected) (n)
Asymptomatic
stenosis n ⫽ 671
(corrected) (n)
Permanent stroke
Transient stroke
Stroke (permanent and transient)
TIA
Cerebrovascular event
Cardiovascular complication
Death
Permanent stroke and death
Stroke (permanent and transient) and death
Stroke, TIA, death
Stroke, TIA, cerebrovascular event, death
Stroke, TIA, cardiovascular complications, death
2.1% (113)
1.3% (72)
3.4% (185)
1.3% (73)
2.0% (108)
2.5% (135)
1.4% (74)
3.1% (169)
4.3% (239)
5.6% (311)
6.6% (363)
8.3% (460)
0.6% (4)
0.9% (6)
1.5% (10)
0.1% (1)
1.8% (12)
4.2% (28)
0.4% (3)
0.7% (5)
1.6% (11)
1.8% (12)
2.8% (19)
6.3% (42)
P
Asymptomatic
stenosis
n ⫽ 671 (n)
P
.009*
.366
.009*
.008*
.762
.009*
.048*
.001*
.001*
⬍.001*
⬍.001*
.062
0.9% (6)
1.0% (7)
1.9% (13)
0.3% (2)
1.7% (11)
4.2% (28)
0.4% (3)
1.0% (7)
2.1% (14)
2.4% (16)
3.3% (22)
6.9% (46)
.036*
.544
.043*
.020*
.545
.010*
.051
.003*
.005*
⬍.001*
⬍.001*
.182
TIA, Transient ischemic attack.
*Indicates significance (P ⬍ .05), Pearson ␹2 test.
Table V. Baseline characteristics for patients with asymptomatic stenosis, divided in two 5-year periods
Baseline characteristics
1994-1998
n ⫽ 330 (n)
1999-2003
n ⫽ 341 (n)
P
Age, mean (quartiles)
Gender; male
Cerebrovascular disease
Diabetes
Hyperlipidemia
Hypertension
Cardiac disease
Previous vascular surgery
Chronic pulmonary disease
Renal insufficiency
Smoking
70 (64;74)
61.8% (204)
62.2% (201)
16.4% (52)
44.4% (79)
59.2% (189)
49.1% (156)
40.0% (128)
8.0% (13)
4.2% (13)
39.4% (117)
69 (64;75)
65.8% (223)
62.7% (205)
21.9% (70)
61.5% (139)
70.9% (231)
45.8% (148)
39.0% (126)
9.3% (30)
7.9% (25)
43.4% (134)
.715
.286
.903
.079
⬍.001*
.002*
.412
.797
.56
.052
.321
For definitions see Material and Methods.
*Indicates significance (P ⬍ .05), Pearson ␹2 test; Wilcoxon test for age.
study was 2.1% for the whole time period and decreased
over time to 0.9% in the second period (1999 to 2003).
There was no perioperative death or major stroke after CEA
for asymptomatic stenosis during the second time period.
The combined stroke or death rate in this national audit
compared favorably with the results in previous large randomized trials, as in Asymptomatic Carotid Atherosclerosis
Study (ACAS) 2.3% and in ACST 2.8%.1,2 When all neurologic morbidity (irrespective of duration or side) and
mortality was included, the 30-day complication rate was
2.3% in the more recent time period.
The perioperative combined stroke or death rate decreased from 3.3% to 0.9% in the second period (P ⫽ .026),
and this improved outcome over time is in agreement with
other reports. In two recent series, similar results were
obtained for patients undergoing CEA for asymptomatic
stenosis, while analyzing outcome for different time periods.11,12 Bond et al13 recently published a meta-analysis of
CEA, where the pooled estimate of the absolute risk of
stroke or death decreased for the asymptomatic indication
in the more recent articles (⬎1995).
Comparing baseline characteristics between the patients
with asymptomatic and symptomatic stenosis, the high rate of
previous vascular surgery (40%) among those with asymptomatic lesions probably reflects how they were often identified,
namely when investigated or followed-up for other lesions and
procedures (ie, aneurysm, lower extremity ischemia or contralateral symptomatic carotid stenosis). In the asymptomatic
cohort, 24% had bilateral operations, contributing to the fact
that almost 63% of these patients had cerebrovascular disease
registered at baseline as a preoperative risk factor. Others may
have symptomatic contralateral occlusion or previous (⬎6
months ago) cerebrovascular events of various causes. It is
important to bear in mind that these patients represent a
highly selected group with atherosclerotic manifestations and
may therefore bias the results. However, the potentially negative bias introduced was not clearly reflected in the perioperative outcome in this series.
Furthermore, the patients with asymptomatic stenosis
had more registered comorbidities (cerebrovascular discussed
above) compared with patients who had symptomatic stenoses. The incidence of hypertension and renal insufficiency
JOURNAL OF VASCULAR SURGERY
July 2006
84 Kragsterman et al
were significantly higher in this patient group. In a study of
patients with renal insufficiency, Ayerdi et al14 reported increased perioperative mortality after CEA.
Even though most of the baseline characteristics were
similar over time for the patients with asymptomatic stenosis,
the frequency of hypertension and hyperlipidemia increased
during the second period, implying that the improvement in
outcome was not due to a favorable difference in case-mix.
The increased registration of hyperlipidemia may reflect a
more active awareness of this risk factor together with an
increased prescription of statins, but the incomplete registrations, especially in the first period, may introduce a bias. Other
investigators have reported a protective effect of statins in
patients undergoing CEA,15,16 however, and an increased
statin use in the second period could have contributed to the
improved outcome.
A similar, somewhat contradictory, contribution to the
improved outcome may also be explained by hypertension.
A more active treatment of this risk factor may reduce both
the cerebrovascular and cardiovascular risk.
Some concerns have been raised that the advancements in
medical treatment may decrease the stroke risk for patients
with asymptomatic carotid stenosis, resulting in reduced netbenefit of CEA. However, data from this study suggest that
the surgical risk also decreases with contemporary treatment
and that the overall outcome may still be in favor of surgery.
Improvements in surgical technique may also have contributed to improved results. Patch angioplasty is suggested as
being beneficial,17-19 but that was not confirmed in the
present study. A possible explanation may be that although
the frequency significantly increased over time, only 40% had
patch closure in the more recent time period. Another factor
influencing the predictive value may be the small number of
adverse events. Eversion endarterectomy gained popularity
during this series but still only represents 5.9% of all CEAs
during the second period. Furthermore, randomized trials20,21 have not consistently demonstrated benefit with eversion endarterectomy compared with conventional technique.22 The type of anesthesia was not registered in
Swedvasc, so its possible effect on the complication rate could
not be addressed.
Age was the only significant risk factor in the logistical
regression analysis, in agreement with the results from
Goldstein et al23 who reported significantly higher stroke
or death rate for those ⬎75 years. In both ACAS and
ACST, those ⬎75 years had a less favorable outcome,
although not statistically significant.1,2 These trials also
demonstrate less benefit for CEA of asymptomatic stenosis
for women compared with men, partly due to a higher
perioperative risk.1,2 In this series no significant gender
differences were noted in the outcome, which is consistent
with other reports that have specifically analyzed women
operated on for an asymptomatic stenosis.24,25 In another
recent large series of CEAs, Hartun et al26 reported a
higher stroke rate for women in univariate analysis. That
difference was lost with multivariate analysis; however, the
indication for surgery could not be assessed in their study.
There are limitations and strengths in this study. In analyzing risk factors, the small numbers of perioperative events
limit the ability of extensive subgroup or multiple regression
analysis.
A potential weakness with registry studies (open audit)
is self-reported data. Swedvasc is validated in different
aspects annually.8,9,27 CEA has been especially studied on
several occasions with high accuracy in the registry.9,28 In
the validation process of this study, which included cross
matching against two other registries, random sampling,
and targeted validation, we could not identify a single
patient with unreported major stroke or death.
In this study, many of the asymptomatic patients were
included in ACST (40%) and underwent an independent
examination by a neurologist at follow-up (30 days). Patients at an increasing number of Swedish hospitals have a
routine neurologist’s examination at 30 days, which is
recommended by the Steering Committee of Swedvasc.
One of the benefits of an open audit is the possibility to
get updated information on changes in practice and results.
Vascular surgeons in Sweden have generally been restrictive
with CEA for asymptomatic stenosis. Indications for the
procedure have most often been based on the degree of
stenosis, but other factors (eg, extent of contralateral disease) vary between centers and individual surgeons, and
could not be assessed in this study. There has been a stable
level of CEAs for asymptomatic patients (11% to 12%) until
recently, when this indication increased after the results of
ACST were made public (20% and increasing).27
When changes in practice are implemented on the basis
of results from the trials, perioperative complication rates
must be regularly assessed. In doing so, the results from this
population-based study may justify an increase in carotid
endarterectomy for patients with asymptomatic stenosis,
also in this routine clinical setting.
CONCLUSION
The perioperative outcomes in this extensively validated national audit of carotid endarterectomy for asymptomatic stenosis were well comparable with those of the
large randomized trials. Furthermore, there were decreasing complication rates over time.
AUTHOR CONTRIBUTIONS
Conception and design: BK, HP, MB, DB
Analysis and interpretation: BK, HP, MB, DB
Data collection: BK, HP
Writing the article: BK, HP, JL, MB
Critical revision of the article: HP, JL, DB, MB
Final approval of the article: HP, DB, MB
Statistical analysis: JL, BK
Obtained funding: HP
Overall responsibility: DB
REFERENCES
1. Executive committee for Asymptomatic Carotid Atherosclerosis Study.
Endarterectomy for asymptomatic carotid artery stenosis. JAMA 1995;
273:1421-8.
JOURNAL OF VASCULAR SURGERY
Volume 44, Number 1
2. MRC Asymptomatic Carotid Surgery Trial (ACST) Collaborative
Group. Prevention of disabling and fatal stroke by successful carotid
endarterectomy in patients without recent neurological symptoms: randomized controlled trial. Lancet 2004;363:1491-87.
3. Hobson RW, Weiss DG, Fields WS, for the Veteran affairs study group.
Efficacy of carotid endarterectomy for asymptomatic carotid stenosis.
N Engl J Med 1993;328:221-7.
4. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effects of carotid endarterectomy in symptomatic
patients with high-grade stenosis. N Engl J Med 1991;325:445-53.
5. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Benefit of carotid endarterectomy in patients with symptomatic
moderate or severe stenosis. N Engl J Med 1998;339:1415-25.
6. European Carotid Surgery Trialist’s Group. MRC European Carotid
Surgery Trial: interim results for symptomatic patients with severe
(70-99%) or mild (0-29%) carotid stenosis. Lancet 1991;337:1235-43.
7. European Carotid Surgery Trialist’s Group. Randomised trial of endarterectomy for recent symptomatic carotid stenosis: final results of the MRC
European Carotid Surgery Trial (ECST). Lancet 1998;351:1379-87.
8. Swedvasc Yearly Report 2004. Available at: www.swedvasc.se
9. Bergqvist D, Troeng T, Elfström J, Hedberg B, Ljungström K-G,
Norgren L, et al. Auditing surgical outcome. Ten years with the Swedish
Vascular Registry - Swedvasc. Eur J Surg 1998;164:581.
10. R Development Core Team. R: A language and environment for
statistical computing. R Foundation for Statistical Computing, Vienna,
Austria. ISBN 3-900051-07-0;2005. http://www.R-project.org.
11. Kresowik TF, Bratzler DW, Kresowik RA, Hendel ME, Grund SL,
Brown KR, et al. Multistate improvement in process and outcomes of
carotid endarterectomy. J Vasc Surg 2004;39:372-80.
12. Darling RG 3rd, Kreienberg PB, Roddy SP, Paty PSK, Chang BB,
Lloyd WE, et al. Analysis of the effect of asymptomatic carotid atherosclerosis study on outcome and volume of carotid endarterectomy.
Cardiovasc Surg 2000;6:436-40.
13. Bond R, Rerkasem K, Rothwell PM. Systematic review of the risk of
carotid endarterectomy in relation to the clinical indication for and
timing of surgery. Stroke 2003;34:2290-303.
14. Ayerdi J, Sampson LN, Deshmukh N, Farid A, Gupta SK. Carotid
endarterectomy in patients with renal insufficiency: Should selection
criteria be different in patients with renal insufficiency? Vasc Surg
2001;6:429-35.
15. McGirt MJ, Perler BA, Brooke BS, Woodworth GF, Coon A, Jain S, et
al. 3-Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors reduce the risk of perioperative stroke and mortality after carotid endarterectomy. J Vasc Surg 2005;42:829-36.
Kragsterman et al 85
16. Kennedy J, Quan H, Buchan AM, Ghali WA, Feasby TE. Statins are
associated with better outcome after carotid endarterectomy in symptomatic patients. Stroke 2005;36:2072-6.
17. Archie JR Jr. A fifteen-year experience with carotid endarterectomy
after a formal operative protocol requiring highly frequent patch angioplasty. J Vasc Surg 2000;31:724-35.
18. Rockman CB, Halm EA, Wang JJ, Chassin MR, Tuhrim S, Formisano
P, et al. Primary closure of the carotid artery is associated with poorer
outcomes during carotid endarterectomy. J Vasc Surg 2005;42:870-7.
19. Bond R, Rerkasem K, AbuRahma AF, Naylor AR, Rothwell PM. Patch
angioplasty versus primary closure for carotid endarterectomy. Cochrane Database Syst Rev 2004;2:CD000160.
20. Ballotta E, Da Giau G, Saladini M, Abbruzzese E, Renon L, Toniato A.
Carotid endarterectomy with patch closure versus carotid eversion endarterectomy and reimplantation: A prospective randomized study. Surgery
1999;125:271-9.
21. Cao PG, Giordano G, De Rango P, Zannetti S, Chiesa R, Coppi G, et
al. and Collaborators of the EVEREST Study Group. Eversion versus
conventional carotid endarterectomy: late results of a prospective multicenter randomized trial. J Vasc Surg 2000;31:19-30.
22. Cao PG, De Rango P, Zannetti S, Giordano G, Ricci S, Celani MG.
Eversion versus conventional carotid endarterectomy for preventing
stroke. Cochrane Database Syst Rev 2001;1:CD001921.
23. Goldstein LB, Samsa GP, Matchar DB, Oddone EZ. Multicenter
review of preoperative risk factors for endarterectomy for asymptomatic
carotid artery stenosis. Stroke 1998;29:750-3.
24. Rockman CB, Castillo J, Adelman MA, Jacobowitz GR, Gange PJ,
Lamparelo PJ, et al. Carotid endarterectomy in female patients: are the
concerns of the Asymptomatic Carotid Atherosclerosis Study valid? J
Vasc Surg 2001;33:236-41.
25. Lee J-W, Pomposelli F, Park KW. Association of sex with perioperative
mortality and morbidity after carotid endarterectomy for asymptomatic
carotid stenosis. J Cardiothor Vasc Anesth 2003;17:10-6.
26. Harthun NL, Kongable GL, Baglioni AJ, Meakem TD, Kron IL.
Examination of sex as an independent risk factor for adverse events after
carotid endarterectomy. J Vasc Surg 2005;41:223-30.
27. Swedvasc Yearly Report 2005. Available at: www.swedvasc.se.
28. Kragsterman B, Logason K, Ahari A, Troëng T, Pärsson H, Bergqvist
D. Risk factors for complications after carotid endarterectomy - a
population based study. Eur J Vasc Endovasc Surg 2004;28:98-103.
Submitted Jan 12, 2006; accepted Mar 6, 2006.