Carotid Stenting and Endarterectomy
A Clinical and Cost Comparison of Revascularization Strategies
Background and Purpose— Investigational carotid stenting for extracranial carotid stenosis has demonstrated procedural results approaching those reported with endarterectomy, but with limited cost and long-term data. This study compared the in-hospital outcomes and costs of these 2 revascularization strategies at a single institution and the long-term effectiveness of carotid stenting.
Methods— Data for 136 endarterectomies and 136 carotid stent procedures at a tertiary-care community hospital were obtained. The primary clinical outcome measures were in-hospital major ipsilateral stroke and death. In-hospital direct variable costs and length of stay were the primary economic measures. Information on restenosis and late ipsilateral stroke for the stent group was available at 2-year follow-up.
Results— These nonrandomized groups were similar, but the endarterectomy group had more symptomatic patients (42% versus 31%; P=0.0004), and the stent group had more NASCET-excluded patients (68% versus 35%; P<0.0001). In-hospital major ipsilateral stroke and death occurred more frequently in the surgical group, but the difference was not significant (2.9% versus 0%; P=0.1). Minor ipsilateral strokes were similar (2.2% versus 2.9%; P=NS). Cost ($5409 versus $3417; P<0.0001) and length of stay (3.0 versus 1.4 days; P<0.0001) were significantly greater for the surgical group. In the stent group, 6-month angiographic restenosis was 3.1%, and 2-year ipsilateral major stroke rate was 0%.
Conclusions— In-hospital outcomes with carotid stenting were similar to those with endarterectomy but were achieved in patients with significantly more comorbidities. Cost and resource utilization with stenting were substantially less than those with endarterectomy. At 2 years, carotid stenting appeared not only durable but also effective in stroke prevention.
Although first performed in 1953,1 carotid endarterectomy was variably applied in patients with significant extracranial carotid artery stenosis until almost 40 years later,2 when the North American Symptomatic Carotid Endarterectomy Trial (NASCET) confirmed its effectiveness in symptomatic patients.3,4⇓ Subsequently, the Asymptomatic Carotid Atherosclerosis Study (ACAS) demonstrated its utility in the asymptomatic patient.5 These 2 landmark trials firmly established carotid endarterectomy as standard therapy for carotid bifurcation disease. However, NASCET and ACAS surgeons were rigorously screened, the patient cohorts were low risk, and the operations were performed at regional centers of surgical excellence, all of which potentially limit the broad applicability of these study results to clinical practice. Several subsequent surveys of carotid endarterectomy have confirmed that rates of stroke and death are higher than reported in these trials,6 even at participating study institutions.7 Operators and hospitals with low volumes, as well as high-risk patients with significant comorbidities, both of which are commonly encountered in clinical settings, are specifically associated with greater rates of adverse outcomes.8–17⇓⇓⇓⇓⇓⇓⇓⇓⇓
Percutaneous carotid angioplasty was first reported in 1981 for fibromuscular dysplasia18,19⇓ and 2 years later in atherosclerotic disease.20,21⇓ Its use was limited as a result of issues surrounding possible embolic complications, suboptimal angiographic results, acute vessel closure, elastic recoil, and restenosis. In the last decade, however, the introduction of adjunctive stenting has mitigated many of these concerns. Although it is still an emerging technique using evolving equipment, both single-center reports22–26⇓⇓⇓⇓ and worldwide surveys27 of carotid stenting have demonstrated procedural results approaching those of endarterectomy, typically in high-risk patients with significant comorbidities excluded from previous surgical trials.28–33⇓⇓⇓⇓⇓ Complications of anesthesia, infection, hemorrhage, myocardial infarction, and cranial nerve palsy are avoided with stenting, which has also shown the potential to shorten hospital stays and lower costs compared with endarterectomy.34
To determine the current relationship between these 2 therapies, we undertook a comparison of in-hospital outcomes, costs, and resource utilization in nonrandomized, contemporary cohorts undergoing carotid endarterectomy or investigational carotid stenting at our institution. In addition, 2-year outcomes for the carotid stent group were analyzed to assess the durability and efficacy of the initial percutaneous results.
Subjects and Methods
The 136 carotid endarterectomies performed in 1997 at Presbyterian Hospital, a 453-bed tertiary-care community hospital in Albuquerque, NM, serve as the surgical cohort and were not in a study protocol. The calendar year 1997 was chosen because it represented a contemporary clinical experience and was centered in the 30-month (March 1996 to September 1998) economic carotid stent data, thus obviating the need for cost adjustment. These endarterectomy patients were not involved in an investigational study; the surgeons caring for these patients determined treatment, both clinical and operative, on the basis of national and community standards. Patients undergoing hospital-based therapy other than endarterectomy were excluded from the analysis.
An investigational registry was established in March 1996 at the same institution to evaluate the safety, efficacy, and costs of carotid stenting. The first 136 consecutive stent procedures (to equal the number of endarterectomy procedures) were used for comparison. All patients gave informed consent and underwent the procedure according to a protocol approved by the Institutional Review Board, with the exception of 1 patient with multiple medical comorbidities and a 52% stenosis with recurrent transient ischemic events in the same vascular distribution, who had the procedure performed on a compassionate basis. A multidisciplinary team including interventional cardiology (W.A.G., H.J.W.), neurology (D.M.B.), and vascular surgery (G.C.) evaluated and cared for these patients. Patients were eligible for the investigation if they had a symptomatic or asymptomatic extracranial carotid stenosis ≥60%, reflecting both the NASCET3 and ACAS5 eligibility criteria at the time of protocol development. The determination of carotid stenosis severity was made according to the NASCET method.3 Patients were excluded for intracranial tumor, arterial venous malformation or hemorrhage, lesion thrombus, prior disabling stroke, dementia limiting informed consent, total carotid occlusion, or stroke in evolution. Extremely calcified lesions, lack of vascular access, and extremely tortuous carotid artery anatomy were also excluded on the basis of the operator’s assessment that these lesion characteristics increased the risk of stenting. Three patients initially considered carotid stent candidates were referred for carotid endarterectomy on the basis of anatomic or access exclusions.
Carotid Stent Protocol
All patients had baseline blood counts, coagulation profiles, chemistries, urinalysis, and ECG. Evaluation by a neurologist using the National Institutes of Health (NIH) Stroke Scale and the Barthel Index was performed before the procedure (usually on the same day), within 24 hours after the procedure, and 6 months after the procedure. In patients experiencing a procedural neurological event, evaluation was also performed at hospital discharge. Screening for neurological events was performed on the yearly anniversary of the stent procedure. A cranial MRI or CT scan was obtained before the procedure and after stenting only for a procedural event. Carotid Doppler ultrasound, for which we used a laboratory accredited by the Intersocietal Commission for the Accreditation of Vascular Laboratories, was done before the procedure, 2 weeks and 6 months after the procedure, and yearly on the anniversary of the procedure. If Doppler evaluation suggested restenosis, angiography was performed to assess stent patency.
Patients were pretreated with aspirin and either ticlopidine 250 mg PO BID or clopidogrel 75 mg PO every morning for at least 3 days before and 2 to 4 weeks after the procedure. Sedation was avoided to allow for frequent neurological evaluation. All patients had 4-vessel cerebral angiography, most at the same session as the stent procedure. Stenosis severity and reference vessel diameter were determined by quantitative analysis of the angiogram (both computer-aided and caliper measurements). Percutaneous access for the intervention was via the common femoral artery. After heparinization, the lesion was predilated with a low-profile coronary balloon to allow for placement of a self-expanding stent (Wallstent, Schneider USA/Boston Scientific Corporation). Approximately midway through this stent experience, the poststent dilation strategy was altered on the basis of data from other sites, and aggressive dilation after stent deployment was avoided to reduce the potential for emboli. Ipsilateral cerebral angiography followed stent placement to assess for any embolic complications or changes in contrast flow. Femoral access was discontinued within 4 to 6 hours, and the patient was discharged on the same or following day.
Angiographic success was defined as a successful stent deployment and a ≤25% residual stenosis. Primary outcome measures included any major ipsilateral stroke or death during hospitalization. Secondary outcome measures included in-hospital minor ipsilateral stroke for both groups and, for the stent group only, restenosis (defined as ≥50% angiographic stenosis at 6 months) and late ipsilateral stroke.
A major stroke was defined as any new neurological deficit increasing the NIH Stroke Scale score ≥4 and sustained for >7 days. A minor stroke was defined as any new neurological deficit resolving completely within 7 days. In all but 1 patient with a minor stroke, the NIH Stroke Scale score changed ≤3 points, representing a minor neurological deficit.
Carotid Endarterectomy Data Collection
Carotid endarterectomy patient charts were reviewed for medical history, information on preoperative testing (eg, carotid Doppler ultrasonography and/or angiography, cardiac or pulmonary testing), the patient’s admission status (elective, urgent, emergent), NASCET-excluded comorbidities, duration of operating room use, in-hospital morbidity and mortality, blood product administration, and intensive care unit/total length of stay after endarterectomy. Since independent neurological evaluation was not performed routinely on these patients, the reporting surgeon provided the necessary data. No posthospital data are reported on the surgical patients.
Economic Populations and Analysis
The 136 carotid endarterectomies in 1997 and the 99 consecutive carotid stent procedures in the first 30 months of the registry were used in the cost and resource utilization analyses. No charge data were used to determine costs; instead, a system of microcosting was used. The institution’s accounting system (Transition System, Inc) was queried for length of stay and direct variable costs associated with carotid endarterectomy and carotid stent patients from the day of the procedure or operation until discharge; hospital costs were further defined according to departmental cost centers. Indirect costs, representing hospital overhead not related to patient care (eg, parking, security, food service), were not included for either group. No professional fees were included in the cost analysis.
All descriptive data are expressed as either mean±SD or as an absolute count and a percentage of total when appropriate. Multiple procedures in individual patients occurred, making the number of procedures (136) in each cohort greater than the number of patients. In every instance in which a patient underwent multiple procedures, the subsequent procedure was only performed once the induction period for clinical sequelae for the previous procedure had expired, providing grounds for the assumption of statistical independence between successive procedures. Analyses of continuous variables between the clinical carotid stent and carotid endarterectomy cohorts were performed with unpaired, 2-tailed Student’s t test with a sample size of 136 for each cohort. In the case of cost and resource utilization comparisons, the sample size is 99 for the stent group and 136 for the endarterectomy group. For categorical variables, either a χ2 or 2-tailed Fisher’s exact test, when appropriate, was used. The Wilcoxon rank sum test was used to compare median values.
Clinical Characteristics of Endarterectomy and Stent Patients
A total of 136 procedures in each group were analyzed, representing 126 endarterectomy and 122 stent patients as a result of bilateral, but not repeated ipsilateral, treatment. The clinical characteristics of the 2 groups are shown in Table 1. The groups were similar in most respects; however, the surgical group had more symptomatic patients (42% versus 31%; P=0.0004), and the stent group had significantly more patients with NASCET-excluded comorbidities (68% versus 35%; P<0.0001).
Ten surgeons performed endarterectomy at the institution in 1997; volumes and complications by surgeon are displayed in Figure 1. Four surgeons performed <10 endarterectomies that year, but this does not skew group results. Five different operators performed carotid stenting, with the Principal Investigator (W.A.G.) as the primary (103 procedures) or assistant operator in all cases.
Angiographic and Procedural Characteristics for the Stent Group
Detailed angiographic and procedural descriptions for the stent group are displayed in Table 2. Forty-five (33%) of the lesions were considered high risk for endarterectomy because of difficult or inaccessible lesion location, contralateral occlusion, prior endarterectomy with restenosis, prior radiation therapy, or accompanying intracranial stenosis. The administration of atropine during the procedure for bradycardia and/or hypotension from carotid body stimulation during lesion dilation and stenting was common (40% of cases). Less common, but usually more long-lived, was the need for intravenous vasopressors during and after the procedure for persistent hypotension without bradycardia, unresponsive to atropine. No complications resulted from periprocedural hemodynamic instability.
As is occasionally seen during carotid cross-clamping in the electroencephalogram-monitored and anesthetized endarterectomy patient, intraprocedural cerebral ischemia during balloon inflation occurred in 6 patients (4%). It usually manifested as a brief seizure, was predictable by the cerebrovascular anatomy, and in no case resulted in a clinical event. A coughing spasm was provoked by balloon inflation in 9 patients (7%), without identifiable mechanism.
Procedural success was 100%, with all patients receiving a stent and relief of their stenosis. The mean angiographic stenosis before the stent was 75%±11% and after the stent was 4%±6%. Median stenosis was greater for the asymptomatic than for the symptomatic patients (77% versus 74%), although this was not statistically significant. There was no significant change in the mean NIH Stroke Scale (0.80±1.78 versus 0.88±1.96; P=0.34) or Barthel Index (19.57±1.74 versus 19.62±1.79; P=0.82) score after the stent procedure. Distal internal carotid artery flow improved after stenting in 66 (48%) of the cases, as judged by reduced contralateral contribution or more rapid distal vessel filling. Mean external carotid artery stenosis significantly increased from 15% to 32% (P<0.0001), but only in 23 cases (17%) was the external carotid stenosis made significantly worse (>80% stenosis). External artery compromise was asymptomatic in all but 4 patients, who experienced short-lived jaw claudication.
In-Hospital Outcomes for Endarterectomy and Stent Groups
Adverse in-hospital outcomes in hierarchical format (only the most severe complication is reported in each patient) for both groups are listed in Table 3. The number of major complications (death and major ipsilateral stroke) was greater in the surgical group than in the stent group, but the difference was not statistically significant (2.9% versus 0%; P=0.12). While the percentage of minor complications was the same in the 2 groups (6.6%), they differed in type. Total adverse outcomes for the 2 groups were similar (9.6% versus 6.6%; P=0.4).
Nonhierarchical adverse hospital outcomes (multiple complications counted in a single patient) were also assessed. Although not compared statistically, a count of total in-hospital adverse events was greater in the surgical group than in the stent group (17.6% versus 5.9%) and included respiratory failure, pulmonary edema, reoperation, and infection not seen with stenting. When death and stroke are excluded, medical complications occurred more frequently in the surgical group (11.8% versus 2.9%).
Table 4 details each procedural minor stroke in the stent group; there were no neurological events after the procedure.
Costs and Resource Utilization for Endarterectomy and Stent Groups
The direct variable costs and resource utilization for both groups are listed in Table 5. The median lengths of stay for the stent and surgery groups were 1 and 2 days (P<0.0001), respectively, regardless of the level of patient acuity (ie, elective, urgent, or emergent). Patients in the carotid stent group spent significantly less time in the hospital (mean, 1.4 versus 3.0 days; P<0.0001) and incurred significantly less direct costs (mean, $3417 versus $5407; P<0.0001) compared with patients in the endarterectomy group. Procedural costs in both groups were similar, but the overall nonprocedural costs were significantly greater in the surgery group ($2867 versus $555; P<0.0001), reflecting significantly more intensive care nursing, pharmacy, radiology, respiratory therapy, and central supply costs. In addition, both mean intensive care unit length of stay (0.21 versus 0.67 days; P<0.0001) and mean total catheterization laboratory time were significantly less for the stent group than mean operating room time for the endarterectomy group (53 versus 179 minutes; P<0.0001). The mean total direct costs plotted against the mean length of stay for each procedure is displayed in Figure 2.
Posthospital Clinical and Doppler Ultrasound Data for the Stent Group
The postprocedural clinical and ultrasound data for the 136 carotid stent procedures are shown in Table 6. At 30 days there were 2 deaths (1.6%), neither neurological nor related to procedure, and 1 minor ipsilateral stroke due to cerebral hemorrhage 2 days after the procedure, presenting as a headache and seizure with minimal neurological deficit. Only 1 additional ipsilateral stroke (minor) occurred in the 6- to 12-month interval, resulting in a total postprocedural ipsilateral stroke rate of 0.8% at 12 months and 1.3% at 24 months, and there were no major strokes in the stent group at any time during the entire 2-year period of analysis.
Doppler results were divided into quartiles according to systolic velocity; they generally overestimate angiographic stenosis severity in all quartiles when compared with nonstented patients. At 6 months, Doppler velocities suggested significant restenosis (peak systolic velocity ≥2.5 m/s) in 8 treated arteries; subsequent angiography in these 8 demonstrated restenosis (>50% diameter) in only 4 (3.1%), and in those cases peak systolic velocity was ≥3.4 m/s. Three of these patients had a repeated percutaneous intervention, and 1 patient had endarterectomy with stent explantation; there were no clinical events arising from these secondary procedures. Between the 6- and 12-month evaluations, 2 additional stents had progression of neointimal hyperplasia on Doppler ultrasound evaluation. One patient presented with the aforementioned ipsilateral minor stroke and had a successful repeated procedure, and the other patient was asymptomatic. The remaining stents continued patent without progression of neointimal hyperplasia to 24 months.
The Carotid Revascularization Endarterectomy Versus Stent Trial (CREST), an NIH study to evaluate carotid stenting, began in late 2000. However, with an anticipated enrollment of 2500 patients and follow-up ranging to 4 years, it will be several years before randomized data are available to guide therapy selection in extracranial carotid disease. Our study provides insight into the relationship between these 2 treatments not otherwise currently available.
We found no significant difference in clinical outcomes between carotid endarterectomy and carotid stenting but significantly less cost and resource utilization in the stent group. Although not randomized, the 2 groups were well matched except for the greater number of symptomatic patients in the surgical group and of NASCET-excluded patients in the stent group. These differences reflect current referral patterns for investigational stenting, which typically attracts more high-risk patients on an elective basis.22 Hence, the equivalent clinical outcomes with stenting in this study were achieved in a higher-risk cohort with an excess of comorbidities compared with surgical patients.
The endarterectomies in the present study were performed in a less select group of patients than in NASCET and ACAS, and the outcomes are probably more representative of clinical practice.8,16⇓ While the 5.1% in-hospital stroke and death rate for endarterectomy patients in this study is consistent with those from the prior large, neurologically controlled, studies of longer (30-day) outcomes,3,35⇓ it probably does not represent all events. Rothwell et al36 analyzed the published results of almost 16 000 endarterectomies from 51 studies over 15 years and found a mean stroke and death rate of 5.6%, with the highest rates (7.7%) associated with the perioperative audit of surgical results by a neurologist and the lowest rates (2.3%) associated with single surgeon authorship. These findings have been confirmed by other investigators in smaller series.37 Since in-hospital stroke in this study was systematically and independently evaluated by a neurologist only in the stent group, other strokes in the surgical group may not have been identified.
The rate of adverse events for carotid stenting in this single-center experience is similar to prior published data.22–26,33⇓⇓⇓⇓⇓ In a 1999 global survey of 5210 carotid stent procedures from 36 sites, the rate was 1.5% for major stroke, 2.7% for minor stroke, and 0.9% for death.27 Minor strokes also accounted for most of our stenting complications, with 2 of the 4 minor strokes related to limitations of rudimentary equipment inherent in an early experience with a new technique. Specifically, the ability to negotiate tortuosity in access vessels and at lesion sites will improve with the smaller, more flexible, dedicated stent systems now being tested to replace the tracheobronchial device used here. However, in 3 of the 4 events no abnormality was evident on intracranial angiography after the procedure, suggesting that the embolism presumed responsible was of limited size. Cerebral embolic protection devices currently in feasibility trials may further reduce the already low rates of these clinically evident embolic events.
In this study carotid endarterectomy consumed significantly greater institutional resources than carotid stenting, with more than twice the total and intensive care unit lengths of stay, longer operating room times, and greater blood product utilization. One possible contributing factor was the 11.8% medical complication rate in the endarterectomy group. In a review of medical complications in the NASCET study, Paciaroni et al38 found a 30-day medical complication rate of 8.1%, one quarter of which prolonged hospitalization. And while discharge on the first postoperative day after endarterectomy has been shown to be safe and to reduce costs, only 65% to 75% of patients are candidates for such a pathway,39–41⇓⇓ and the remaining patients may significantly lengthen the mean. While some institutions may have shorter lengths of stay for endarterectomy patients, we believe that our lengths of stay are representative since several authors have observed the same or greater lengths of stay after endarterectomy.40–43⇓⇓⇓ It is noteworthy that same-day discharge was feasible and safe in a number of the stent patients and with increasing experience could be applied to the majority of patients, since postprocedural complications in carotid stenting are unusual, and femoral artery closure devices not previously available now allow earlier ambulation.
This is the first study to compare direct hospital costs of these revascularization strategies, with the costs incurred for the surgical group comparable to prior cost analyses of endarterectomy.40,44–47⇓⇓⇓⇓ The greater costs associated with endarterectomy appear to be the result not only of longer hospitalizations but also of increased intensity of services for surgical patients. In an analogous comparison, studies of percutaneous angioplasty versus bypass surgery in patients with multivessel coronary disease have found significantly greater hospital costs associated with the surgical pathway but nearly equivalent costs at 5 years, largely because of the repeated intervention required in the angioplasty arm secondary to a significant rate of coronary restenosis.48 Although long-term costs in this study were not tracked, the low rate of restenosis in the stent arm at 2 years, which is lower than published rates of restenosis after surgery,49 is unlikely to lead to significant late costs in the stent group, thus preserving initial differences. Given the total annual cost for endarterectomy in the United States of approximately $2 billion, the nearly 40% reduction in direct costs with carotid stenting seen here represents significant potential savings. The reduction or elimination of preoperative screening tests could further reduce currently uncounted costs associated with surgery.
The criteria for Doppler/ultrasound evaluation of a stented carotid artery appear to be somewhat different than those for evaluation of a nonstented vessel, with velocities generally greater in the stented artery for the same degree of angiographic stenosis. Although speculative, this likely relates to possible changes in vessel compliance with a stent in place, as well as the longer length of the resistive conduit that promotes continuous velocity acceleration through it with even modest neointimal hyperplasia. Given the generally low embolic potential associated with neointimal hyperplasia, non–hemodynamically significant in-stent restenosis does not appear to increase risk for subsequent stroke and may in fact be advantageous.
In any carotid revascularization strategy, procedural safety is critical, but stroke prevention is the ultimate measure of effective therapy. Stenting appears to be a durable procedure, with angiographic restenosis at 6 months limited to 3.1%. This low rate of in-stent restenosis, confirmed elsewhere,22,23,33⇓⇓ compares favorably with a restenosis rate of 7.6% to 11.4% when systematically studied after endarterectomy in ACAS.49 Also important is the finding that continued in-stent neointimal hyperplasia after 6 months is extremely infrequent.
In the carotid stent group there were no major ipsilateral strokes and only 1 ipsilateral stroke after 30 days in the 2-year follow-up period. On the basis of modeling using NASCET and ACAS medical treatment arms as historical references for the symptomatic and asymptomatic stent patients, respectively, an ipsilateral stroke rate of approximately 11% would have been expected in this population at 2 years. The actual ipsilateral stroke rate for the 77 arteries reaching the 2-year mark in the stent group was 3.9%, most of which were procedural and all of which were minor. These data, along with another large series with long-term follow-up,33 suggest that carotid stenting is effective in stroke prevention.
In conclusion, this single-center, nonrandomized comparison of carotid revascularization strategies demonstrates comparable in-hospital outcomes, but significantly less cost and resource utilization, with carotid stenting compared with endarterectomy and durable 2-year outcomes with stenting. These stent results, in a high-risk cohort, represent an early experience with relatively primitive tools. As dedicated stent equipment emerges, cerebral embolic protection devices are added, and operator technique improves, stenting will likely become even more predictable. We await the results of current randomized trials to further define the relationship between endarterectomy and stenting.
- Received April 5, 2001.
- Revision received October 19, 2001.
- Accepted October 22, 2001.
- ↵Hsia DC, Moscoe LM, Krushat WM. Epidemiology of carotid endarterectomy among Medicare beneficiaries. Stroke. 1998; 29: 346–350.
- ↵Stukenborg AJ. Comparison of carotid endarterectomy outcomes. Arch Neurol. 1997; 54: 426–432.
- ↵Ruby ST, Robinson D, Lynch JT, Mark H. Outcome analysis of carotid endarterectomy in Connecticut: the impact of volume and specialty. Am Vasc Surg. 1996; 10: 22–26.
- ↵Karp HR, Flanders WD, Shipp CC, Taylor B, Martin D. Carotid endarterectomy among Medicare beneficiaries. Stroke. 1998; 29: 46–52.
- ↵Hannan EL, Popp AJ, Tranmer B, Fuestel P, Waldman J, Dhiraj S. Relationship between provider volume and mortality for carotid endarterectomies in New York State. Stroke. 1998; 29: 2292–2297.
- ↵Perler BA, Dardik A, Burleyson GP, Gordon TA, Williams GM. Influence of age and hospital volume of the results of carotid endarterectomy: a statewide analysis of 9918 cases. J Vasc Surg. 1998; 17: 25–31.
- ↵Bockenheimer SA, Mathias K. Percutaneous transluminal angioplasty in arteriosclerotic internal carotid artery stenosis. AJNR Am J Neuroradiol. 1983; 4: 791–792.
- ↵Wiggl U, Gratzl O. Transluminal angioplasty of stenotic carotid arteries: case reports and protocol. AJNR Am J Neuroradiol. 1983; 4: 793–795.
- ↵Yadav JS, Roubin AS, Iyer S, Vitek J, King P, Jordan WD. Elective stenting of the extracranial carotid arteries. Circulation. 1997; 95: 376–381.
- ↵Wholey MH, Wholey M, Mathias K, Roubin GS, Diethrich EB, Henry M, Bailey S, Bergeron P, Dorros G, Eles G, Gaines P, Gomez CR, Gray B, Guimaraens J, Higashida R, Ho DS, Katzen B, Kambara A, Kumar V, Laborde JC, Leon M, Lim M, Londero H, Mesa J, Musacchio A, Myla S, Ramee S, Rodriquez A, Rosenfield K, Sakai N, Shawl F, Sievert H, Teitelbaum G, Theron JG, Vaclav P, Vozzi C, Yadav JS, Yoshimura SI. Global experience in cervical carotid artery stent placement. Catheter Cardiovasc Interv. 2000; 50: 160–167.
- ↵Roubin GS, New G, Iyer SS, Vitek JJ, Al-Mubarak N, Liu MW, Yadav J, Gomez C, Kuntz RE. Immediate and late clinical outcomes of carotid artery stenting in patients with symptomatic and asymptomatic carotid artery stenosis: a 5-year prospective analysis. Circulation. 2001; 103: 532–537.
- ↵Gray WA, Barrett DM, Chandran G, Turner R, White HJ. Carotid stenting and endarterectomy: a cost comparison. J Am Coll Cardiol. 1998; 31 (suppl): 493A.Abstract.
- ↵Rothwell PM, Slattery J, Warlow CP. A systematic review of the risks of stroke and death due to endarterectomy for symptomatic carotid stenosis. Stroke. 1996; 27: 260–265.
- ↵Chaturvedi S, Aggarwal R, Murugappan A. Results of carotid endarterectomy with prospective neurologist follow-up. Neurology. 2000; 55: 769–772.
- ↵Paciaroni M, Eliasziw M, Kappelle LJ, Finan JW, Ferguson GG, Barnett HJM. Medical complications associated with carotid endarterectomy. Stroke. 1999; 30: 1759–1763.
- ↵Hlatky MA, Rogers WJ, Johnstone I, Boothroyd D, Brooks MM, Pitt B, Reeder G, Ryan T, Smith H, Whitlow P, Wiens R, Mark DB, for the Bypass Angioplasty Revascularization Investigation (BARI) Investigators. Medical care costs and quality of life after randomization to coronary bypass surgery. N Engl J Med. 1997; 336: 92–99.
- ↵Moore WS, Kempczinski RF, Nelson JJ, Toole JF. Recurrent carotid stenosis: results of the Asymptomatic Carotid Atherosclerosis Study. Stroke. 1998; 29: 2018–2025.