Chapter  49:  Effectiveness and Cost-Effectiveness of Echocardiography and Carotid Imaging in the Management of Stroke: Evidence Report/Technology Assessment Number 49

Echocardiography

1. Which clinically inapparent abnormalities identified by echocardiography among patients presenting with a new ischemic brain syndrome represent risk factors for recurrent stroke?

2. What is the yield of echocardiography in detecting potential sources of cardioembolism among patients with a new ischemic brain syndrome?

3. What are the operating characteristics (sensitivities, specificities, and likelihood ratios) of transthoracic and transesophageal echocardiography in detecting potential sources of cardioembolic stroke?

4. Whatare the incidence and nature of complications associated with transesophageal echocardiography?

5. Are there clinically identifiable groups of patients with new ischemic brain syndrome who benefit from anticoagulation?

6. Are there echocardiographically identifiable groups of patients with new ischemic brain syndrome who benefit from anticoagulation?

Carotid Imaging

1. What are the operating characteristics of available tests for measuring carotid artery stenosis?

2. What is the incidence of complications associated with cerebral angiography?

3. What is the efficacy of carotid endarterectomy in reducing the rate of recurrent stroke among symptomatic patients with carotid artery stenosis?

4. What is the incidence of complications associated with carotid endarterectomy?

5. Does timing affect the safety of carotid endarterectomy?

Cost Effectiveness

The overarching question for the cost-effectiveness analyses of both echocardiography and carotid imaging in patients with stroke is: what is the cost-effectiveness of routine vs. selective imaging procedures in patients with a new ischemic stroke or transient ischemic attack (TIA)? The following is a list of subquestions:

1. Of routine, selective, and no imaging, what is the most cost-effective strategy to reduce the risk of recurrent stroke associated with modifiable risk factors potentially identifiable by imaging?

2. How do cost-effectiveness estimates change with differences in clinical and demographic factors?

3. How do cost-effectiveness estimates change with differences in treatment effectiveness?

4. How do cost-effectiveness estimates change with differences in other uncertain model parameters?

Echocardiography

The effectiveness of echocardiography as a tool in the evaluation of patients with cerebral ischemia or infarction has not been established directly. Specifically, there have been no clinical trials comparing outcomes among patients managed with and without echocardiography after stroke or TIA. Assessing the usefulness of echocardiography in such patients therefore involves examining the evidence for several assertions that serve as links in a causal pathway between echocardiography and clinical outcomes, particularly recurrent stroke. These assertions are as follows:

• Clinically inapparent abnormalities identified by echocardiography convey increased risk of recurrent stroke.

• The prevalence of these abnormalities is not inconsequential.

• Echocardiography is accurate in diagnosing these abnormalities.

• Adverse events associated with echocardiography are small or infrequent compared to its benefits.

• Efficacious treatments exist that reduce morbidity and mortality associated with potential sources of cardioembolic stroke identified by echocardiography.

• Adverse events associated with these treatments are small or infrequent compared to their benefits.

Key Question 1. Which clinically inapparent abnormalities identified by echocardiography among patients presenting with new ischemic brain syndrome represent risk factors for recurrent stroke?

Approximately 15 percent of strokes are thought to be attributable to cardioembolic sources. Some of the processes that give rise to cardioembolic stroke, most notably atrial fibrillation, are usually clinically apparent at the time a patient presents with a stroke. Other conditions, however, are clinically occult but may be identifiable with the use of imaging procedures such as echocardiography. The usefulness of echocardiography in the management of stroke depends on the ability to identify cardiac lesions and on the presence and modifiability of recurrent stroke risk conveyed by those lesions.

Several different cardiac and aortic abnormalities identifiable by echocardiography have been studied as potential sources of cardioembolic stroke. There is fair overall evidence that left ventricular thrombus (LVT) is associated with an increased risk of systemic embolization, including stroke. Evidence regarding the presence and degree of stroke risk associated with left atrial thrombus (LAT) is insufficient to draw firm conclusions. There is fair evidence that complex aortic atheromas (ulcerated, mobile, or > 4 mm in thickness) represent risk factors for stroke, independent of coexisting carotid artery disease. There is also fair evidence for an association between atrial septal aneurysm (ASA) and stroke, particularly in the presence of coexisting patent foramen ovale (PFO). PFO alone may be an important risk factor for stroke in young patients, but evidence for an association is conflicting. Epidemiological studies of left atrial myxoma and stroke are lacking, but several case series suggest a substantially higher prevalence of stroke in patients with myxoma than in the general population. Evidence for an independent association of left ventricular aneurysm, spontaneous echocardiographic contrast, and valvular strands with stroke is insufficient. Previously documented associations between mitral valve prolapse (MVP) and stroke were likely due to inaccuracy in the determination of MVP with early echocardiographic techniques. Finally, mitral annular calcification appears to be an indicator of atherosclerotic vascular disease rather than an independent cause of stroke.

It must be emphasized that the absence of sufficient evidence regarding an association between a cardiac abnormality and stroke does not necessarily indicate that an association does not exist. When biomedical knowledge and experience suggest a high likelihood that a particular lesion, such as intracardiac thrombus, is an independent risk factor for stroke, that likelihood remains high in the face of inconclusive evidence.

Key Question 2. What is the yield of echocardiography in detecting potential sources of cardioembolism among patients with a new ischemic brain syndrome?

The review of echocardiographic yield focused primarily on intracardiac thrombi, ASA, complex aortic atheroma, and atrial myxoma (lesions for which there was fair to good evidence of an independent association with stroke). Researchers analyzed the prevalence of these lesions as detected by TTE and TEE, in unselected patients, patients with and without cardiac disease, patients without significant carotid artery stenosis, and young patients (under 50) with stroke. The yield of intracardiac thrombi using TTE was highly variable. In one study of good quality from Japan, the prevalence of intracardiac thrombus in consecutive patients with stroke or TIA undergoing TTE was 2.1 percent (95 percent confidence interval [CI], 0.8 to 4.6 percent). In three fair-quality studies, no cases of thrombus were diagnosed in patients without heart disease. Among patients with heart disease, the prevalence of thrombus was highly variable, ranging from 0 to 36 percent. This variability, as well as small sample sizes, made it difficult to derive a reliable estimate of the yield of TTE among patients with a history of cardiac disease or without significant carotid disease. One atrial myxoma was diagnosed among 721 patients in eight studies. In most studies, LAT, ASA, and aortic atheroma were not found using TTE in patients without AF.

In two studies of TTE among young patients (aged 15 to 45) with stroke, the pooled prevalence of intracardiac thrombus in 180 patients was 2.2 percent (similar to that in unselected patients). No left atrial myxomas were detected. The prevalence of any unsuspected thrombus, tumor, valvular vegetation, or cardiomyopathy was 4.4 percent (95 percent CI, 2.1 to 8.2).

The overall yield of intracardiac thrombus using TEE in consecutive stroke patients was 1.7 percent (95 percent CI, 0.5 to 5.3 percent). The prevalence of heart disease was not reported in most studies of TEE, making it difficult to determine the importance of this variable. In four studies of patients without significant carotid disease, the prevalence of intracardiac thrombus on TEE was highly variable, ranging from 1.5 to 18 percent. One myxoma was detected among approximately 1,200 patients examined. The yield of ASA (3.8 to 21.6 percent) and complex aortic atheroma (1.9 to 17.2 percent) using TEE varied widely across studies. One study of patients under 60 with negative TTE reported a prevalence of ASA of 28 percent, with 15 percent having both ASA and PFO. The prevalence of complex aortic atheroma in this study excluding elderly patients was 3.4 percent.

The finding from previous reviews of a higher yield of intracardiac thrombus and other potential sources of stroke using TEE largely reflects the inclusion in those reviews of patients with AF. Findings in the current review suggest that in patients without AF, TEE may be less useful than previously described. TEE may have advantages in patients who have insignificant or no carotid disease or who have a negative TTE. There is little information on the yield of TEE in patients with pre-existing heart disease other than AF.

Key Question 3. What are the operating characteristics (sensitivities, specificities, and likelihood ratios) of transthoracic and transesophageal echocardiography in detecting potential sources of cardioembolic stroke?

Because of the relatively low prevalence of intracardiac thrombus in patients with stroke, it is difficult to assess the accuracy of TTE and TEE in this population. Studies attempting to determine the accuracy of these tests for LAT and LVT, as detected by direct intracardiac inspection, have necessarily examined populations in which the prevalence of thrombus is high. These populations, patients undergoing surgery for severe mitral valve disease or left ventricular aneurysms, are not representative of the general population of patients with stroke, and thrombi occurring in these patients may differ substantially from those likely to affect patients with cardioembolic stroke. It is therefore possible that the reported accuracy estimates in these studies differ from the accuracy of TTE and TEE in patients with stroke.

The average sensitivity and specificity of TEE in detecting LAT in these studies were 93 and 97 percent, respectively. For TTE, sensitivity and specificity averaged 42 and 99 percent. The low sensitivity of TTE was largely due to missed left atrial appendage thrombi. For the diagnosis of LVT, TTE had an average sensitivity of 78 percent and specificity of 87 percent. When results from individual studies were plotted on a summary receiver operating characteristic (SROC) curve, however, it appeared that varying accuracy across studies may have been partly due to differing diagnostic thresholds. Using the SROC curve to estimate the accuracy, the sensitivity and specificity of TTE for diagnosing LVT were 77 and 95 percent, respectively. It should be noted, however, that approximately 15 percent of TTE examinations in studies of LVT were deemed inadequate for interpretation, limiting the diagnostic utility of this test. No studies of the accuracy of TEE in diagnosing LVT were identified.

When the prevalence of intracardiac thrombi in patients with stroke is assumed to be 2 percent or less, as many as or more patients will receive unnecessary treatment due to false positive tests than will receive potentially beneficial treatment for a true positive test, if echocardiographic technology is used to select patients for treatment with anticoagulants. Under current estimates of test accuracy, the prevalence of LAT would have to exceed 15 percent and the prevalence of LVT 37 percent in order to achieve 90 percent predictive value.

Studies examining the accuracy of echocardiography in diagnosing ASA and aortic atheroma are lacking. However, given that the association with stroke has been established for the echocardiographic, rather than anatomic, definitions of these lesions, it may be argued that TEE represents the gold standard for the diagnosis of these lesions as they relate to cardioembolic stroke. Few studies have assessed the accuracy of echocardiography in diagnosing left atrial myxoma. These studies suggest accuracy approaching 100 percent, though one study found disagreement between TTE and TEE in 2 of 11 cases.

Key Question 4. What are the incidence and nature of complications associated with transesophageal echocardiography?

In observational studies of poor and fair quality, the pooled risk of periprocedural death associated with TEE was 0.014 percent. The risk of death in patients specifically undergoing TEE for evaluation of possible cardiac embolus could not be directly calculated. Data were insufficient to determine whether the risk of death was higher in elderly or critically ill patients.

From observational studies of fair quality, the average risk of major (requiring treatment) cardiovascular, pulmonary, and gastrointestinal complications from TEE was 0.7 percent. The rates of major complications in elderly and critically ill patients were 0.4 percent and 0.8 percent, respectively. Neither of these rates was significantly different from the overall rates. No cases of infective endocarditis or systemic infection were found in 775 patients followed after TEE. Approximately 1.9 percent of TEE were unsuccessfully attempted, and an additional 0.9 percent were stopped for complications, most frequently patient intolerance. The rate of minor complications (most commonly patient intolerance) requiring discontinuation of the procedure was not consistently reported, but appears to be about three times the rate of major complications.

Although the estimates of risk came from studies of poor or fair methodological quality (no included study was assessed as having overall good quality), no other data were available to provide more reliable estimates. Data are insufficient to determine whether complication rates are different in patients presenting with particular indications such as cerebral ischemic syndromes.

Key Question 5. What is the efficacy of anticoagulant therapy in reducing the rate of recurrent stroke among patients with potential sources of cardioembolism?

For any given patient with stroke, the potential usefulness of echocardiography in detecting a source of cardioembolism depends on the absence of clinically apparent indications for treatment (i.e., anticoagulation). There is substantial evidence, for instance, that for patients with stroke and AF, anticoagulant drugs confer net benefit, making echocardiographic identification of lesions warranting anticoagulation in these patients superfluous. Whether anticoagulation is beneficial in stroke patients without AF is less clear.

There is fair evidence that unselected patients with stroke do not benefit from anticoagulation as compared to antiplatelet therapy. Evidence from a large, fair-quality international trial suggests that subcutaneous heparin given acutely to patients with stroke is not associated with improved outcomes when compared to aspirin. The two therapies used in combination may confer net benefit, but further study is needed to confirm this finding. A good-quality multicenter trial comparing chronic anticoagulation (target INR 1.4-2.8) and aspirin (325 mg) found no differences in either benefits or harms between the two treatments. Another good-quality trial employing higher degrees of anticoagulation (target INR 3.0-4.5) was stopped early due to increased rates of ICH and death with anticoagulation as compared to aspirin.

No fair- or good-quality studies were found examining the effectiveness of anticoagulation in the prevention of recurrent stroke among patients with stroke and cardiac conditions other than AF. Studies of primary stroke prevention among patients with MI (myocardial infarction) suggest that when compared to aspirin, anticoagulation either alone or in combination with aspirin does not confer net benefit. For patients with dilated cardiomyopathy (DCM), evidence regarding anticoagulation for primary stroke prevention comes from observational studies that provide conflicting results. The only good-quality study found that anticoagulation was more effective than aspirin in the primary prevention of stroke, particularly for patients with moderate and severe cardiomyopathy, after acute MI.

Overall, there was fair evidence that neither acute nor chronic anticoagulation confers net benefit, as compared to aspirin, for unselected patients with stroke. Also, there was insufficient evidence to reach conclusions regarding the effectiveness of anticoagulation for secondary prevention of stroke among patients with stroke and clinically apparent cardiac conditions other than AF. Studies of primary prevention suggest that anticoagulation may be beneficial for patients with DCM but is probably not beneficial in patients with MI; however, results from studies of primary stroke prevention may not be generalizable to patients who have already experienced stroke and are candidates for secondary prevention. Given these findings, it appears that the scope of patients for whom echocardiography may be useful, if it can effectively identify treatable sources of recurrent stroke, includes all stroke patients except those with AF.

Key Question 6. Are there echocardiographically identifiable groups of patients with new ischemic brain syndrome who benefit from anticoagulation?

Studies of the effectiveness of anticoagulation for echocardiographically identifiable lesions were all observational in design. Pooled data from five retrospective cohort studies suggest that warfarin, and possibly surgical PFO closure, may reduce the rate of recurrent stroke or TIA among patients with stroke and PFO. However, these studies were generally of poor quality and did not account for differences in baseline characteristics that may have given rise to differences in outcomes across treatment groups. A small, poor quality cohort study of patients with stroke found to have mobile aortic atheromas revealed a trend toward lower recurrent stroke rates with warfarin as compared to aspirin, but no death. A poor-quality systematic review of primary stroke prevention in patients with intraventricular thrombus after acute MI suggested a net benefit with anticoagulation, but the reviewed studies were observational, and no adjustment for potential confounding was conducted. Moreover, whether or not findings from studies of primary stroke prevention among patients with acute MI can be used to draw conclusions regarding secondary prevention among a general population of patients with stroke is not clear. Researchers found insufficient evidence to draw conclusions about the effectiveness of anticoagulation in reducing morbidity and mortality among stroke patients with echocardiographically identified lesions.

Cost-Effectiveness

Because of the lack of solid evidence for important components of effectiveness, it is difficult to accurately estimate the cost-effectiveness of echocardiography in the management of stroke. Where evidence was lacking or insufficient, informed assumptions were made to enable estimating cost-effectiveness. Assumptions include the following: intracardiac thrombus conveys increased stroke risk for the first year after the initial stroke; thrombus prevalence is 2 percent in unselected patients and 5 percent in patients with heart disease; and anticoagulant drugs reduce the risk of recurrent stroke by one-third. Using those assumptions, one quality-adjusted life year (QALY) can be saved for an approximate incremental cost of $300,000, using TEE only in patients with heart disease. Other strategies were less cost-effective, though TTE in patients with heart disease was the preferred strategy under some plausible assumptions. The cost-effectiveness ratio for either echocardiographic procedure fell below $50,000 per QALY if the assumed relative risk reduction with anticoagulation was increased to 86 percent and the prevalence of thrombus was simultaneously increased to 6 percent. The cost per QALY of all strategies increased as average life expectancy diminished (e.g., with increasing age or comorbidity).

Carotid Imaging

The role of carotid imaging is better established than that of echocardiography in patients with stroke. It is clear that carotid artery stenosis conveys increased risk of stroke and that efficacious treatment exists to reduce that risk. However, the most effective imaging strategy for diagnosing carotid artery stenosis is controversial. The most widely used tests include two noninvasive tests, carotid ultrasound and magnetic resonance angiography, and one invasive test, cerebral angiography. These tests may be used alone or in various combinations. Although the noninvasive tests are not associated with significant complications, their effectiveness in predicting who will benefit from surgical intervention has not been directly established, as it has for angiography. The noninvasive tests therefore carry the potential for false positive and false negative diagnoses and the consequent risk of selecting patients without significant carotid stenosis for ineffective and potentially harmful surgery, or excluding patients with significant stenosis from beneficial treatment. In order to compare the effectiveness of various strategies for carotid imaging, evidence related to the following was examined:

• Operating characteristics (sensitivities, specificities, and likelihood ratios) of available tests for measuring carotid stenosis;

• Harms associated with these tests;

• Efficacy of treatment for varying degrees of carotid stenosis; and

• Harms associated with these treatments.

Key Question 1. What are the operating characteristics of available tests for measuring carotid artery stenosis?

Despite numerous studies of the accuracy of noninvasive carotid imaging, relatively few have been conducted in which all patients undergoing noninvasive tests also undergo diagnostic confirmation with cerebral angiography. The lack of diagnostic verification in these studies creates biased estimates of sensitivity and specificity. Studies can adjust for this bias by angiographically studying a random sample of subjects with negative noninvasive tests. Studies were reviewed of CUS and MRA accuracy that either had no obvious or likely verification bias or that adjusted for this bias.

It is clear from the literature that the accuracy of CUS in diagnosing carotid stenosis varies substantially across centers. It is likely that published reports of the accuracy of CUS from single centers overestimate the accuracy in most settings. This has two important implications. First, it may be inappropriate for individual practitioners or medical centers to assume that the accuracy of CUS in their practices is equivalent to published figures. Second, it is clear that there is potential for CUS to be highly accurate. The sensitivity and specificity of CUS estimated from SROC curves constructed from the results of eight predominantly fair-quality studies were 80 and 91 percent, respectively, for moderate or greater (> 50 percent) stenosis, and 75 and 87 percent for severe (> 70 percent) stenosis. When the largest and only good-quality study was excluded, sensitivity and specificity for severe stenosis rose to 94 and 84 percent. The lower accuracy in the largest study than in other studies may have been due to the use of conventional rather than color-flow duplex imaging, but may also have been due to the representation of multiple centers. Reports from single centers may provide biased estimates of accuracy, as those centers finding low accuracy may choose not to submit their results for publication.

Whether the accuracy of MRA varies by center is not clear. There have not been multicenter studies of MRA. Published data, excluding studies with obvious or likely verification bias, suggest a sensitivity and specificity of 92 and 97 percent for detecting severe stenosis. However, studies of MRA were generally of fair to poor quality. As with CUS, it is possible that centers publishing their accuracy data are not representative of all users of MRA. Until there are more high-quality data on the accuracy of MRA, current estimates of MRA accuracy in measuring carotid stenosis must be interpreted cautiously.

All studies of the accuracy of CUS and MRA used in combination were biased by incomplete verification. In the majority of these studies, sensitivity was 100 percent. However, the studies were generally of poor quality. The specificity of combined CUS and MRA was variable, ranging from 69 to 100 percent. The estimated sensitivity and specificity of combined CUS and MRA for detecting severe stenosis were 95 and 98 percent, respectively. In approximately 18 percent of patients, the results of CUS and MRA in detecting severe stenosis were discordant.

Key Question 2. What is the incidence of complications associated with cerebral angiography?

In prospective studies examining the incidence of stroke and death following cerebral angiography in patients suspected of having cerebrovascular disease and potential candidates for CEA, the overall rate of 0.02 percent for deaths was lower than the 0.08 percent rate previously reported. Only two deaths were found in 10 studies including 3,074 patients.

Significant heterogeneity was found between rates of combined stroke or death from all studies as well as between studies stratified by various methodologic criteria. The rate of combined stroke or death ranged from 0 percent to 4 percent in three studies rated as having good quality, with the study rated as having the highest quality reporting a rate of 1.3 percent (95 percent CI, 0.5 to 2.8 percent).

The risk of complications appears higher in patients with greater degrees of carotid stenosis, who are also those patients most likely to benefit from subsequent CEA.

The magnitude of incremental risk of cerebral angiography (i.e., the risk above the baseline risk of recurrent stroke or death in recently symptomatic patients) cannot be reliably estimated at this time but would be expected to be lower than the rates reported above.

Key Question 3. What is the efficacy of carotid endarterectomy in reducing the rate of recurrent stroke among symptomatic patients with carotid artery stenosis?

In two large, good-quality randomized controlled trials (RCTs), carotid endarterectomy reduced the risk of disabling stroke or death for surgically fit patients with symptomatic ipsilateral stenosis greater than 70 percent as measured by the European Carotid Surgery Trial (ECST) method, and over 50 percent as measured by the North American Symptomatic Carotid Endarterectomy Trial (NASCET) method. In a meta-analysis of these trials, the number needed to treat to prevent one disabling stroke or death over 2 to 6 years was 15 (95 percent CI, 10 to 31) for severe stenosis (70 to 99 percent by NASCET criteria or 80 to 99 percent by ECST criteria) and 21 (95 percent CI, 11 to 125) for moderate stenosis (50 to 69 percent by NASCET or 70 to 79 percent by ECST). No benefit was seen in patients with lesser degrees of carotid stenosis. In the subgroup of patients with severe stenosis, increased degree of stenosis was associated with greater benefit from surgery. The results of the studies are generalizable to surgeons and centers with low perioperative complication rates (30-day stroke or death rate less than 6 percent). The studies did not include angiographic morbidity or mortality in their results.

Although patients over 80 years old, non-whites, and females were underrepresented in these studies, multivariate analysis to determine factors associated with increased benefit was performed on these and other clinical and demographic characteristics in the two trials. In NASCET and ECST, less benefit was seen in females for all degrees of carotid stenosis, and among patients with 50 to 69 percent stenosis, the absolute risk reduction was eight-fold lower in women than in men. The lesser degree of benefit for women may be partially due to a lower baseline recurrent stroke rate compared to men for equivalent degrees of carotid stenosis. Older age was associated with increased benefit in ECST and in the subgroup of patients in NASCET with 70 to 99 percent stenosis.

It must be noted that among patients screened in the NASCET, fewer than one-third were randomized. Approximately one-third did not fulfill baseline criteria, 15 percent were excluded for medical reasons, and another 23 percent were eligible but not randomized. Such exclusions must be considered when trying to generalize data from the endarterectomy trials to individual patients or populations of patients in "real-world" health care settings.

Key Question 4. What is the incidence of complications associated with carotid endarterectomy?

Using data from 12 studies of good quality, the pooled rate of combined perioperative (30-day) stroke or death associated with CEA was 6.8 percent (95 percent CI, 4.6 to 9.5 percent), and from nine studies of good quality, the pooled rate of death alone was 1.6 percent (95 percent confidence interval, 1.0 to 2.5 percent).

In NASCET, the 30-day postrandomization rate of stroke or death ranged from 2.4 percent (for patients with < 70 percent carotid stenosis) to 3.3 percent (70 to 99 percent stenosis) in patients assigned to medical therapy. Therefore, surgery is associated with an additional 35 to 44 perioperative events per 1,000 patients. In NASCET, approximately 60 percent of the strokes that occurred in the perioperative period were nondisabling (Rankin score < 3).

Methodologic characteristics of the studies explained some of the variation in complication rates. Population-based studies, RCTs, studies with independent ascertainment of complications, studies with nonsurgeon authors, and studies published since 1990 were associated with higher combined complication rates. The pooled complication rate in randomized controlled trials was higher than the pooled rate for other studies, suggesting that these trials may have high generalizability despite strict selection criteria. Population-based studies also reported relatively high perioperative complication rates. All of the characteristics associated with higher complication rates appear to occur in studies rated as having higher average methodologic quality.

Key Question 5. Does timing affect the safety of carotid endarterectomy?

The appropriate timing of carotid imaging depends partly on the timing of CEA. CEA is often delayed for several weeks after stroke onset due to concerns about the safety of CEA in the acute period. There is fair evidence that early as compared with delayed CEA is not associated with an increased risk of major complications. Three nonrandomized studies of fair quality suggest that in patients with recent minor or nondisabling stroke, CEA performed earlier than the traditional waiting period of 4 to 6 weeks is not associated with significantly increased adverse events compared to delayed surgery, with a pooled rate of 3.3 percent for early CEA versus 5.3 percent for later CEA. When data from all studies (including seven rated poor quality) are included, the pooled rate of major perioperative complications (stroke or death) is 3.9 percent for early CEA versus 2.7 percent for later CEA. The pooled rate of death alone from all studies was about 1.0 percent in patients undergoing either early or later CEA. There was a nonsignificant trend toward better outcomes for early CEA in studies published since 1990.

There is insufficient evidence to draw conclusions regarding the risk of very early CEA (i.e., less than 1 week after presenting with symptoms). There is also inadequate evidence to draw conclusions for specific subgroups, including patients with specific computed tomography scan findings and greater degrees of carotid stenosis. Patients selected for early CEA in these studies are likely to comprise an overall lower-risk population compared to patients not selected for early CEA, though in higher-quality studies patients undergoing early and later CEA were comparable according to important clinical and demographic criteria.

Cost-Effectiveness

What strategies for using carotid imaging are cost-effective?

The lack of good or consistent evidence regarding the accuracy of noninvasive carotid imaging strategies makes it difficult to accurately determine the most cost-effective strategy for selecting patients with stroke for CEA. Assuming the accuracy of statistics derived from this review, two testing strategies provide the most benefit when compared to no testing: first, administer MRA and refer patients with severe (70-99 percent) stenosis directly to CEA. Second, administer joint CUS and MRA, and when both tests demonstrate moderate to severe (50-99 percent) stenosis, refer patients directly to CEA. When the two tests disagree, request angiographic confirmation. The incremental cost-effectiveness ratios for these two strategies are approximately $250,000 and $700,000 per QALY, respectively.

In sensitivity analyses, the variable with the greatest influence on the results of the carotid imaging model was the prevalence of severe carotid stenosis. At severe stenosis prevalences of 0.15 and below, all testing strategies were dominated by the strategy of no testing or had cost-effectiveness ratios exceeding $250,000 per QALY (0.15 was the prevalence assumed in the base-case analysis). However, as this prevalence increased above 0.15, the cost-effectiveness ratios of two strategies, CUS with angiographic confirmation of severe stenosis (CUS/Angio-70), and MRA with direct CEA referral for severe stenosis (MRA/70 percent), fell precipitously, such that at a prevalence of 0.20, these strategies had cost-effectiveness ratios in the range of $60,000 to $75,000 per QALY. At higher prevalences, these ratios fell further. When compared to the strategy of no testing, CUS/Angio-70 had an incremental cost-effectiveness ratio of less than $50,000 per QALY at a prevalence of 0.25, while the incremental cost-effectiveness of MRA/70 fell below $50,000 per QALY as the prevalence of severe stenosis approached 0.30. These results suggest that carotid imaging may compare unfavorably, in terms of cost-effectiveness, with other commonly endorsed health care interventions, when the prevalence of carotid stenosis is low. Carotid imaging may be most efficient for those with a high pretest probability of severe stenosis, e.g., patients with peripheral vascular disease or audible carotid bruits.

Varying the cost of testing did not substantively affect the results, except in the case where MRA was assumed to cost $2,500 (as opposed to $1249 in the bas-case analysis). In this analysis, the strategy of initial CUS with angiographic confirmation of severe stenosis became undominated, with a cost-effectiveness ratio of $280,000 per QALY. Varying the accuracy of the different testing strategies over wide ranges did not have a substantial overall effect on the results. When the perioperative complication rate was assumed to be zero, noninvasive strategies involving direct referral to CEA of patients with moderate or greater stenosis expectedly became the most cost-effective; without risk of complications, angiographic confirmation to avoid false positives was no longer beneficial, and the marginal benefit of CEA among patients with moderate stenosis was no longer counterbalanced by perioperative risk. Varying the duration of risk reduction associated with CEA between 2 and 10 years also did not substantively affect the cost-effectiveness ratios. Likewise, restricting the cohort to only patients with TIA or minor stroke, which reflects the patient populations in the two large carotid endarterectomy trials, did not have a major impact on cost-effectiveness ratios, though it did produce a different set of undominated strategies.

It is noteworthy that strategies in which patients with moderate (50-69%) stenosis were referred for CEA provided fewer QALYs than strategies in which such patients were treated nonsurgically, despite the fact that the review (and hence the model inputs) reflected an overall benefit from CEA for moderate stenosis. This occurred as a result of the fact that the benefit of CEA over nonsurgical management in patients with moderate stenosis is small, such that when a 3 percent discount rate is applied to account for the fact that health benefits incurred or realized in the future are considered to be of lower value than benefits realized in the present, the future benefits are outweighed by the perioperative complications incurred immediately after surgery. When perioperative complication rates were assumed to be zero, or when the discount rate was removed, strategies involving CEA for patients with moderate stenosis became more cost-effective.

Ultrasonography

The application of ultrasound technology to medical diagnostics has evolved rapidly over the last 4 decades. Early techniques included B-mode ("brightness") ultrasound, which distinguishes structures within the body by translating reflected ultrasound waves of differing intensity into differing levels of brightness, and M-mode ("motion") ultrasound, which allows tracking of object motion (e.g., cardiac valves) across a defined distance. The incorporation of Doppler technology added the ability to characterize the direction, velocity, and turbulence of blood flow. Computer advances now allow translation of ultrasound signals into two-dimensional images, with Doppler signals translated into color patterns that can be superimposed on the image. Three-dimensional ultrasonographic imaging techniques are not yet commonly used in clinical practice.¹⁵

Ultrasound as used in medical diagnostics involves the transmission and reflection of sound waves at frequencies typically ranging from 2 to 10 megahertz (MHz). There is no known risk to human tissues from the ultrasonic waves themselves.¹⁶ Higher ultrasound frequencies are associated with finer resolution. However, at higher frequencies, ultrasound waves are less able to penetrate tissues. Thus, lower frequencies must often be used, particularly for echocardiography in adults, whose skin and soft tissues are typically thicker than those of children. Although this compromises resolution to some degree, frequencies between 2 and 5 MHz provide adequate resolution in most circumstances.

Echocardiography

Echocardiography utilizes ultrasound technology to provide dynamic visual imaging of the cardiac chambers, valves, walls, and septa; the pericardium and pericardial space; and the thoracic aorta. Doppler techniques are particularly useful for assessing valvular function. Echocardiography has become widely used as a diagnostic tool to evaluate patients with cerebral ischemia. Specifically, echocardiography has been advocated as a method of identifying potential sources of stroke, particularly among patients whose ischemic syndrome is unexplained by cerebrovascular disease or whose clinical presentations suggest an embolic source.

Transthoracic and transesophageal echocardiography are the two most commonly used approaches to ultrasonic imaging of the heart. In TTE, the ultrasound transducer is placed on the patient's chest wall. Because the procedure is non-invasive, it is rarely if ever associated with complications. TTE typically allows excellent imaging of the right and left ventricles and the interventricular septum, owing to the proximity of these structures to the anterior chest wall. Images of the atria, interatrial septum, left atrial appendage, and aortic arch are less clear on TTE. In addition, TTE images may be inadequate for interpretation when patients cannot cooperate with the examination or when thick adipose tissue, scar, or air within hyperinflated lungs impedes the transmission of ultrasound waves to the heart.

TEE involves passing a small ultrasound transducer into the esophagus. Because the esophageal wall is thinner than the chest wall, higher frequency transducers can be used in TEE than in TTE, allowing better resolution. Moreover, imaging of posterior structures, including the atria, interatrial septum, left atrial appendage, mitral valve, and aortic arch, is superior with TEE. However, TEE offers less advantage in imaging anterior structures, including the left ventricle. Moreover, because TEE requires intubation of the esophagus and often involves conscious sedation of patients, it is associated with rare but significant complications (reviewed later in this report). TEE probes can be single-plane, biplane, or multiplane. Multiplane probes allow for visualization of multiple cross sections of the heart without probe manipulation but are typically larger than single-plane and biplane probes, making them potentially more difficult to pass into the esophagus.

With both TTE and TEE, shunting of blood from the right to left side of the heart can be visualized with the use of air contrast. Typically, a small amount of saline or gelatin is mixed with air and agitated to create small air bubbles. The agitated solution is then injected intravenously. Because the air bubbles become trapped in the pulmonary capillaries, no bubbles are seen in the left heart unless there is a communication between the right- and left-sided circulation. Contrast is most frequently used to detect an atrial septal defect or patent foramen ovale. Valsalva or other maneuvers to increase intrathoracic pressure may be used to maximize the chance of observing a right-to-left shunt. Complications of contrast echocardiography, including cerebral ischemia from air embolism, have been reported.^{17, 18}

As with nearly all diagnostic uses of ultrasound, the quality of echocardiographic studies depends heavily on the experience and skill of the operator. Most TTEs in the U.S. are performed by non-physician technicians, but physicians are typically responsible for interpretation.¹⁶ Because the quality of echocardiography is so highly dependent on operator and interpreter skill and experience, the American College of Physicians, American Heart Association, and American College of Cardiology commissioned a task force on clinical competence in adult echocardiography, which has issued a set of standard expectations for all echocardiographic practitioners.¹⁶

Carotid Ultrasound

Ultrasound has become the most commonly used modality for imaging the carotid arteries. Doppler technology has been particularly useful in this regard, allowing measurement of the degree of carotid stenosis based on changes in the spectrum and velocity of flow. Many different criteria have been developed to gauge the degree of carotid stenosis. The most frequently used are criteria based on the peak systolic velocity (PSV), end-diastolic velocity (EDV), and the ratio of velocities at the internal carotid artery vs. the common carotid artery (ICA/CCA ratio). Less commonly used is the measurement of "spectral broadening," a measure based on the observation that in stenotic arteries, the distribution, or spectrum, of different flow velocities is greater than in an artery without narrowing. While many have attempted to determine which Doppler criteria are most accurate in diagnosing carotid stenosis,^19-26 others have observed that criteria may need to be lab-specific and updated over time, due to differences in equipment and technique.^27-30 Over the last decade, the combination of color Doppler and B-mode ultrasound, known as color flow duplex, has become the standard method for examining carotid arteries with ultrasound.

The accuracy of ultrasound is highly dependent on skill, experience, and quality assurance. The variation in accuracy across labs can be substantial.³¹ Some centers claiming excellent accuracy have proposed selecting patients for CEA based on ultrasound alone.^{19, 32-36} One limitation of this practice is that ultrasound does not allow for imaging of the intracranial vasculature, but the importance of imaging the intracranial arteries before CEA is questionable.³⁷

Magnetic Resonance Angiography

Magnetic resonance angiography employs magnetic resonance technology to visualize flow within blood vessels. Radiofrequency pulses are applied to "saturate" the tissues in a given region, such that blood flowing into that region represents the only unsaturated substance in the field and can therefore be distinguished from surrounding structures. The most well studied application of this method is known as time-of-flight (TOF) imaging. TOF can be two-dimensional, imaging thin "slices" within the region of interest, or three-dimensional, imaging thick "slabs." These two methods are complementary and often used in conjunction. In most settings, the slices and slabs of 2D and 3D-TOF undergo computer processing to create a two-dimensional, longitudinal image of the blood vessel being studied, through a process known as maximum intensity projection.

When there is a severe stenosis, there may be a disruption in the visualization of flow using TOF MRA. When such "flow voids" occur, they usually indicate arterial occlusion if distal flow is not visualized, or tight stenosis if flow resumes distal to the flow void. Flow voids seen on MRA are typically categorized as severe stenoses but are the source of error in some cases where flow appears disrupted for other reasons, e.g., a tortuous blood vessel. Additionally, flow voids limit the quantification of precise degrees of stenosis, and thus some prognostic information may be lost. Contrast MRA involves the intravenous injection of magnetic contrast material, typically containing gadolinium. Contrast MRA eliminates the problem of flow voids.

Although MRA allows noninvasive imaging of the carotid arteries in a fashion similar to conventional angiography, it has limitations. First, although MRA can be used to image the intracranial arteries, this requires more time and is not routinely done as part of carotid imaging in all centers. Second, MRA is limited in its ability to characterize carotid artery plaques. Third, many patients cannot undergo MRA due to claustrophobia, intraocular metallic objects, cerebral aneurysm clips, or pacemakers. Finally, up to 10 percent of magnetic resonance scans are incomplete or technically inadequate due to motion artifact or signal interference from surgical clips or other metallic objects.³⁸

Conventional Angiography

Conventional cerebral or carotid angiography is considered the "gold standard" for determining carotid artery stenosis. Angiography involves the intra-arterial injection of contrast medium, with X-ray imaging. Angiography can be selective, with only the vessel of interest injected with contrast, or non-selective, with the entire aortic arch injected to provide visualization of the entire cerebral vascular tree. Angiography also allows visualization of carotid plaque ulceration, which may affect prognosis and lower the threshold for surgery.³⁹ The major disadvantage of angiography compared to non-invasive imaging is the potential for complications, most notably stroke and death.

Angiography is two-dimensional and therefore may not be perfectly accurate in measuring the actual degree of stenosis in a particular artery. However, the studies that have determined the association between carotid stenosis and outcomes, including those that have demonstrated the benefits of CEA, have used angiography as the criterion for grading carotid stenosis. Thus, while angiography may not perfectly reflect actual carotid stenosis, it is angiographic measurement of stenosis, not pathological measurement of stenosis, that has been validated as a predictor of outcomes.

Digital subtraction angiography (DSA) employs digital image processing to increase imaging sensitivity to contrast and thereby reduce the amount of contrast needed. Intravenous DSA is not routinely used for carotid imaging because it does not offer improved accuracy when compared to other methods that are less costly and less invasive.^40-42 Intra-arterial DSA is commonly used in conjunction with or as a substitute for conventional film angiography. Although there are disadvantages to DSA that may compromise its accuracy,⁴³ DSA was used to measure carotid stenosis in approximately 30 percent of patients in the European Carotid Surgery Trial (ECST),⁴⁴ which demonstrated a benefit of CEA for patients with severe angiographic stenosis. Intra-arterial DSA has therefore been validated as a tool for selecting patients who will benefit from carotid surgery.

Different methods have been used to measure the severity of carotid stenosis on angiography. The two most commonly used methods are those that were used in the two large, international trials demonstrating a benefit of CEA for patients with carotid stenosis.^{44, 45} The North American Symptomatic Carotid Endarterectomy Trial (NASCET) measured the degree of stenosis by subtracting the diameter of the carotid artery lumen at its narrowest point from the diameter of the internal carotid artery lumen distal to the area of stenosis, and dividing by the latter diameter. The ECST measured stenosis by subtracting the diameter of the carotid artery lumen at its narrowest point from the estimated external diameter of the artery at the point of stenosis and dividing by the latter. A third method uses the diameter of the proximal common carotid (CC) artery as the denominator. Because the diameters of the common carotid artery and of the internal carotid artery at the site of stenosis are typically larger than the diameter of the distal internal carotid artery, the NASCET method tends to estimate a lower degree of stenosis than the ECST and CC methods. Both the NASCET method and the ECST method, however, have been validated as predictors of outcomes with and without surgical intervention. Studies attempting to develop a formula for converting stenosis measured by one method to the other have derived both linear (ECST or CC = 0.6 x NASCET + 40 percent)⁴⁴ and non-linear (ECST = 55.16 + 0.29 x NASCET + 0.002 x NASCET²) equations.⁴⁵

Background

Table 2. Potential sources of cardioembolic stroke

Table 2. Potential sources of cardioembolic stroke

Identified Primarily by Echocardiography	Usually Known or Apparent at Presentation
Intracardiac thrombus	Atrial fibrillation
Left ventricle	Recent myocardial infarction
Left atrium	Dilated cardiomyopathy
Left atrial appendage	Rheumatic heart disease (mitral stenosis)
Left ventricular aneurysm	Prosthetic heart valves
Spontaneous echocardiographic contrast	Infective endocarditis
Atrial septal aneurysm	Non-bacterial thrombotic endocarditis
Patent foramen ovale
Mitral valve prolapse
Mitral annular calcification
Valvular strands
Aortic atheroma
Intracardiac tumors (atrial myxoma)

The presence of some potential sources of cardioembolic stroke is usually known or clinically apparent at the time a patient presents with a cerebral ischemic event. Atrial fibrillation, which accounts for nearly half of strokes classified as cardioembolic,^{47, 48} is diagnosed by electrocardiography, which is routinely performed in patients presenting with stroke. Likewise, the presence of a prosthetic heart valve can usually be established through the history or physical exam. Other potential sources of stroke -- recent myocardial infarction (MI), dilated cardiomyopathy (DCM), infective endocarditis, and rheumatic heart disease -- may be diagnosed echocardiographically but usually first manifest through symptoms other than cerebral ischemia and are therefore typically known at the time a patient presents with stroke or TIA. For instance, in a series of 133 cases of native-valve infective endocarditis, 17 patients (13 percent) presented with cerebral ischemia, but 15 of them also presented with other manifestations indicating the diagnosis of endocarditis.⁴⁹ Among 839 patients with rheumatic mitral valve disease in one study, none developed cerebral ischemia before valvular disease was diagnosed.⁵⁰ Stroke may frequently be the first manifestation of non-bacterial thrombotic endocarditis,⁵¹ but this condition typically occurs in the setting of malignancy, disseminated intravascular coagulation, or other severe acute or chronic illness⁵¹ and is therefore not relevant to the majority of patients presenting with stroke.

Finally, there are some patients for whom echocardiography may identify treatable lesions, but for whom other indications for the same treatment are evident without echocardiography. This is the case for patients with AF, who may have left atrial or ventricular thrombi identifiable by echocardiography, but for whom anticoagulant therapy is indicated regardless of whether or not thrombus is found.⁵² In other words, echocardiography may in some instances identify a source of stroke but not alter therapeutic management. In such cases, echocardiography typically does not add value in terms of management to reduce stroke recurrence, though it may be warranted for other reasons, such as to examine for the presence of structural heart disease. Echocardiography may also be useful in such cases if the decision about whether or not to initiate therapy is difficult; e.g., if anticoagulant therapy is relatively contraindicated in a patient with AF, echocardiography may help to "tip the scale" in making the decision to initiate therapy.

In this section, we address clinically inapparent lesions that are identified primarily by echocardiography and that may represent important sources of cardioembolic stroke. We reviewed the literature with the aim of elucidating the quality and strength of evidence for an independent association linking each lesion to ischemic stroke. We employed a "best evidence" approach, examining the highest quality evidence available for each lesion.⁵³ We generally limited our review to studies of at least fair quality that used designs capable of assessing potential causal associations, i.e., cohort studies (including clinical trials) and case-control studies, and excluded case series and case reports. Because of limitations in the methods for selecting controls in many of the case-control studies, we also reviewed studies that reported the prevalence of lesions among random samples from general populations.

In addition to rating the quality of each study, we assessed the evidence for each lesion for the following features:

• Consistency of association across studies;

• Strength of association;

• Persistence of the size of the association after accounting for potential confounders;

• Incremental association between the severity of the lesion, or characteristics hypothesized to increase stroke risk, and the risk of stroke;

• Appropriate temporal relationship between occurrence of the lesion and stroke.

These factors represent epidemiological criteria that can aid in the judgment of whether or not a statistical association between two variables represents a cause-effect relationship.⁵⁴ Particularly important is the issue of confounding. Because studies of associations between echocardiographic lesions and stroke are necessarily observational rather than experimental, it is essential to consider the possibility that observed associations are explained not by the hypothesized risk factor but rather by confounding factors -- variables that are associated with both risk factor (i.e., the echocardiographic lesion) and outcome (i.e., stroke). We examined not only whether studies accounted for potential confounding factors but also the degree to which the size of an observed association changed with adjustment for confounding factors. Even when an association remains statistically significant after adjustment for potential confounders, a substantial reduction in the size of the association with adjustment suggests that the remaining association may be due to "residual" confounding by unmeasured or inadequately measured variables.

Findings

Summary

Table 3. Evidence for independent association between (more...)

Table 3. Evidence for independent association between clinically inapparent echocardiographic lesions and stroke

Fair to good evidence of association	Insufficient evidence	Fair to good evidence of no association
Atrial septal aneurysm (particularly with coexisting patent foramen ovale)	Left atrial thrombus	Mitral annular calcification
	Left ventricular aneurysm	Mitral valve prolapse
Complex aortic atheroma (> 4mm, mobile, or ulcerated)	Patent foramen ovale
Left atrial myxoma	Spontaneous echocardiographic contrast
Left ventricular thrombus	Valvular strands

It must be emphasized that the absence of sufficient evidence regarding an association between a cardiac abnormality and stroke does not necessarily indicate that an association does not exist. When biomedical knowledge and experience suggest a high likelihood that a particular lesion -- such as LAT -- is an independent risk factor for stroke, that likelihood will remain high in the face of inconclusive evidence. Stated from a Bayesian perspective, if the prior probability that a lesion is associated with stroke is high, and if studies provide insufficient evidence to change that probability in either direction, then the posterior probability of an association remains high. Likewise, if the prior probability of an association is low or moderate, studies providing evidence of an association will increase the posterior probability, but the degree to which that probability rises may not be enough in some cases to firmly establish the presence of a true association. Finally, the presence of an association does not always imply a causal relationship. Establishing causation in epidemiology depends not only on the presence of an association but on other factors, such as the biological plausibility of a cause-effect relationship between the risk factor and disease under study.

Background

The importance of using echocardiography to identify a potential source of cardioembolism rests not only on the existence of an association between a specific lesion and the risk of recurrent stroke but also on the prevalence of that lesion. For instance, some lesions are likely to represent true sources of cardioembolic stroke but may be so uncommon that the number of patients who would need to undergo echocardiography to detect one lesion would be extremely high. It would be difficult to rationalize routine use of echocardiography to detect such lesions. Lesions that are more prevalent, and that therefore may account for a larger burden of illness, may represent an important rationale for the use of echocardiography, provided effective treatments exist to reduce morbidity or mortality associated with those lesions. Because determining actual prevalence would require direct inspection of the heart and aorta through surgical exploration or autopsy, which are not performed on the vast majority of patients presenting with stroke, we report prevalences from echocardiographic series. While such series may under- or overestimate the true prevalence of lesions, they provide important data on the expected yield of echocardiography.

We sought to examine the prevalence of potential sources of cardioembolic stroke in consecutive patients with stroke or TIA and in patient subgroups for whom the selective use of echocardiography has been proposed: patients without significant carotid stenosis, patients with heart disease, and young patients. Although studies generally reported the presence of several lesions representing potential source of cardioembolic stroke, we focus here on the yield of clinically inapparent lesions most likely to be amenable to treatment (LAT, LVT, and intracardiac tumors) and lesions for which we found fair or good evidence of an association with stroke (ASA and complex aortic atheroma).

We examined studies that reported findings from consecutive or random samples of patients with stroke or TIA, or that selectively examined patients based on objective, clearly defined criteria. We excluded studies that examined patients who were referred for echocardiography, unless objective criteria for referral were explicitly stated. The rationale for excluding these studies relates to variability in referral patterns across centers. At some centers, it is routine to refer all patients presenting with stroke for echocardiography, whereas at other centers, only patients for whom the index of suspicion for cardioembolism is high are referred. The implicit criteria that define this index of suspicion, or pre-test probability, vary from clinician to clinician and from center to center. Without knowing these criteria, it is difficult to know to which population the reported prevalence applies.

We sought to discover the yield of echocardiography in patients presenting with stroke or TIA who do not have AF. As discussed earlier, the echocardiographic identification of a potential source of cardioembolic stroke is less relevant in patients with AF than in those without AF, because treatment with anticoagulant therapy is indicated in patients with AF presenting with stroke, regardless of the findings of echocardiography.⁵² We therefore excluded studies that did not allow differentiation of echocardiographic findings between patients with and without AF.

Findings

Summary

In one study of good quality from Japan, the prevalence of intracardiac thrombus in consecutive patients with stroke or TIA undergoing TTE was 2.1 percent (95 percent CI, 0.8 to 4.6 percent). In three fair-quality studies, no cases of thrombus were diagnosed in patients without heart disease. Among patients with heart disease, the prevalence of thrombus was highly variable, ranging from 0 to 36 percent. This variability, as well as small sample sizes, makes it difficult to derive a reliable estimate of the yield of TTE among patients with a history of cardiac disease or without significant carotid disease. One atrial myxoma was diagnosed among 721 patients in eight studies. In most studies, LAT, ASA, and aortic atheroma were not found using TTE in patients without AF.

In two studies of TTE among young patients (aged 15 to 45) with stroke, the pooled prevalence of intracardiac thrombus in 180 patients was 2.2 percent (similar to that in unselected patients). No left atrial myxomas were detected. The prevalence of any unsuspected thrombus, tumor, valvular vegetation, or cardiomyopathy was 4.4 percent (95 percent CI, 2.1 to 8.2).

The overall yield of intracardiac thrombus using TEE in consecutive stroke patients was 1.7 percent (95 percent CI, 0.5 to 5.3 percent). The prevalence of heart disease was not reported in most studies of TEE, making it difficult to determine the importance of this variable. In four studies of patients without significant carotid disease, the prevalence of intracardiac thrombus on TEE was highly variable, ranging from 1.5 to 18 percent. One myxoma was detected in over 1,200 patients undergoing TEE. The yield of ASA (3.8 to 21.6 percent) and complex aortic atheroma (1.9 to 17.2 percent) using TEE also varied widely across studies. One study of patients under 60 with negative TTE reported a prevalence of ASA of 28 percent, with 15 percent having both ASA and PFO. The prevalence of complex aortic atheroma in this study excluding elderly patients was 3.4 percent.

The prevalence of complex aortic atheroma was variable across studies, ranging from 2 to 22 percent. The highest rates were reported by studies in which patients underwent TEE for the expressed purpose of detecting aortic atheroma. In studies of patients undergoing TEE to search for any source of cardioembolic stroke, the prevalence of complex atheromas ranged from 2 to 6 percent, suggesting that when conducted routinely, TEE may not yield the majority of atheromas present.

The finding from previous reviews of a higher yield of intracardiac thrombus and other potential sources of stroke using TEE largely reflects the inclusion in those reviews of patients with AF.^{4, 150} Our findings suggest that in patients without AF, TEE may be less useful than previously described.⁴ TEE may have advantages in patients who are found to have insignificant or no carotid disease or who have a negative TTE. There is little information on the yield of TEE in patients with pre-existing heart disease other than AF.

Background

It is often assumed in clinical practice that echocardiography, particularly TEE, is perfectly accurate in detecting potential cardioembolic sources of stroke. That is, if TEE demonstrates a lesion, it is assumed that the lesion is present, and if TEE does not demonstrate a lesion, it is assumed that the lesion is absent. This is largely due to the lack of a more accurate imaging modality for most intracardiac pathology. Echocardiography has limitations and is almost certainly not perfectly accurate in detecting most cardiac lesions. Studying the accuracy of echocardiography entails comparing it to a more accurate "reference standard" examination. Unfortunately, for the purposes of intracardiac pathology, the reference standard is often direct inspection of the heart through surgical or autopsy examination. Although other cardiac imaging modalities exist, such as contrast ventriculography, computed tomography, magnetic resonance imaging, and nuclear scintigraphy, for most lesions none of these modalities is clearly superior to echocardiography such that it can serve as an appropriate reference standard.

Several studies have compared echocardiography to direct intracardiac inspection. We reviewed these studies for the purpose of estimating the accuracy of TTE and TEE in detecting potential sources of cardioembolic stroke.

Findings

Supplemental Analyses

Summary Receiver Operating Characteristic Curves

Variation in sensitivity and specificity of the same diagnostic test may result from the choice of different diagnostic thresholds. For instance, a group of investigators may be interested in examining varying thresholds for the diagnosis of LAT using TEE. In one analysis, the investigators may want to determine the accuracy of TEE if readings of "definite," "probable," and "possible" thrombus are considered to be positive for LAT; i.e., the investigators may set a low diagnostic threshold in order to assess diagnostic accuracy when the number of undetected thrombi is minimized. Using this low diagnostic threshold, the sensitivity of TEE is likely to be high, while the specificity may be low. In another analysis, the investigators may include only readings of "definite" thrombus as positive. Using this high diagnostic threshold, specificity will likely be high, while sensitivity may be low. If the sensitivity and specificity from these two analyses were averaged, the pooled results may indicate moderate sensitivity and specificity, when in fact the test is capable of both high sensitivity and high specificity, depending on the diagnostic threshold chosen.

Receiver operating characteristic (ROC) curves capture the tradeoff between sensitivity and specificity by plotting a test's true positive rate against its false positive rate at varying diagnostic thresholds. The two-dimensional nature of test accuracy is more appropriately summarized as a ROC curve than as an average of sensitivity and specificity across different thresholds. When a single summary statistic is desired, ROC curves allow determination of sensitivity and specificity at the diagnostic threshold where maximal accuracy is achieved, which reflects a test's diagnostic potential more accurately than average sensitivity and specificity.

Just as a single study may assess a test's accuracy across different diagnostic thresholds, different studies examining the same diagnostic test may also use different thresholds. Thus, when summarizing the results across studies, simple averaging of sensitivity and specificity may not accurately reflect the test's true operating characteristics. To deal with this problem, statistical methods for calculating summary ROC (SROC) curves have been developed (see Appendix F).¹⁷⁰

Figure 1. Summary receiver operating characteristic (more...)

   Figure 1. Summary receiver operating characteristic (ROC) curves for the diagnosis of intracardiac thrombus with echocardiography

1. Summary ROC curves for the diagnosis of left atrial thrombus using TEE. Black circles represent good and fair quality studies. Gray triangle represents poor quality study. Dotted curve is for all studies. Black curve is for good and fair quality studies.
   

2. Summary ROC curves for the diagnosis of left atrial thrombus using TTE. Black circles represent good and fair quality studies. Gray triangles represent poor quality studies. Dotted curve is for all studies. Black curve is for good and fair quality studies.
   

3. Summary ROC curves for the diagnosis of left ventricular thrombus using TTE. Black circles represent good and fair quality studies. Gray triangles represent poor quality studies. Dotted curve is for all studies. Black curve is for good and fair quality studies.

SROC curves for the diagnosis of intracardiac thrombus using echocardiography, constructed from the results of studies tabulated in Evidence Tables 3, 4, and 5, are displayed in Figure 1 (differences between sensitivity and specificity in Evidence Tables 3, 4, and 5 and those plotted on the SROC curves result from adding 0.5 to each cell in the statistical construction of the SROC curves, to avoid empty cells). The curves for TTE (Figure 1 parts B and C) display the typical shape of a ROC curve. The atypical shape of the ROC curve for TEE in detecting LAT reflects the fact that several studies with the highest sensitivity^{152, 153, 157} also had equal or higher specificity than other studies.^154-156 In Figure 1, parts A and B, it is apparent that across most of the studies, accuracy estimates are clustered around similar false positive rates (1-specificity), suggesting that the studies had similar diagnostic thresholds. In Figure 1, part B, the point of maximal accuracy on the ROC curve falls outside the range of data from individual studies. The calculated sensitivity of TTE for diagnosing LAT at the point of maximal accuracy on the ROC curve is 97 percent, while sensitivities in the individual studies range from 0 to 59 percent. In this situation, the average sensitivity and specificity across studies provides a less-biased estimate of test accuracy than sensitivity and specificity at the point of maximal accuracy on the ROC curve. Because the diagnostic threshold appeared similar across studies in the ROC curves for each test in the diagnosis of LAT, average sensitivity and specificity likely reflect the true diagnostic potential in these testing situations. For the diagnosis of LVT, the sensitivity and specificity of TTE at the point of maximal accuracy were 77 and 95 percent, respectively, which provide higher overall accuracy than the pooled average sensitivity and specificity of 79 and 87 percent. As noted above, however, the estimates of sensitivity and specificity in several studies of the accuracy of TTE in diagnosing LVT were derived after exclusion of indeterminate test results. These estimates, and therefore the SROC curve constructed from them, may overstate the accuracy of TTE.

Outcomes of Diagnostic Testing

Figure 2. Estimates of true and false positive test (more...)

   Figure 2. Estimates of true and false positive test results for a hypothetical cohort of 1000 patients with stroke undergoing echocardiography for the detection of intracardiac thrombus: effect of varying prevalence estimates

1. True and false positive tests among a hypothetical cohort of 1000 patients undergoing TEE for detection of left atrial thrombus, according to thrombus prevalence. Black line indicates true positive tests; dotted line indicates false positive tests. Dashed line indicates estimated true prevalence of thrombus in unselected patients.
   

2. True and false positive tests among a hypothetical cohort of 1000 patients undergoing TTE for detection of left atrial thrombus, according to thrombus prevalence. Black line indicates true positive tests; dotted line indicates false positive tests. Dashed line indicates estimated true prevalence of thrombus in unselected patients.
   

3. True and false positive tests among a hypothetical cohort of 1000 patients undergoing TTE for detection of left ventricular thrombus, according to thrombus prevalence. Black line indicates true positive tests; dotted line indicates false positive tests. Dashed line indicates estimated true prevalence of thrombus in unselected patients.

Using estimates of accuracy from our literature review along with varying estimates of prevalence for LAT and LVT, we calculated the number of patients with true and false positive and true and false negative echocardiographic diagnoses of LAT and LVT in a hypothetical cohort of 1,000 patients with stroke undergoing TTE or TEE (Figure 2). For these calculations, we assumed that TEE is equivalent to TTE in diagnosing LVT.

Assuming a sensitivity and specificity of 93 and 98 percent, respectively, for the diagnosis of LAT using TEE, approximately equal numbers would receive a false as compared to a true diagnosis of LAT, assuming a prevalence of 2 percent, which is greater than the estimated prevalence of LAT from our literature review. In order for the positive predictive value of TTE or TEE to reach 90 percent, the prevalence of LAT would have to exceed 15 percent.

For the diagnosis of LVT, assuming a sensitivity and specificity of 77 and 95 percent, respectively (maximal accuracy from SROC curve), false positive tests outnumber true positives unless the prevalence of LVT is greater than 6 percent. To achieve a positive predictive value of 90 percent, the prevalence of LVT would have to exceed 37 percent. It is possible to use very strict criteria for diagnosing LVT with TTE, thereby increasing specificity. In order for true positives to outnumber false positives in the diagnosis of LVT at a prevalence of 2 percent, specificity would need to exceed 98 percent. To reach a positive predictive value of 90 percent, the specificity of TTE would need to be greater than 99.8 percent. To achieve this specificity, sensitivity would likely be sacrificed, increasing the number of false negative tests.

Summary

Because of the relatively low prevalence of intracardiac thrombus in patients with stroke, it is difficult to assess the accuracy of TTE and TEE in this population. Studies attempting to determine the accuracy of these tests for LAT and LVT have necessarily examined populations in which the prevalence of thrombus is high. These populations, patients with severe mitral valve disease or left ventricular aneurysms, are not representative of the general population of patients with stroke, and thrombi occurring in these patients may differ substantially from those likely to affect patients with cardioembolic stroke. It is therefore possible that the reported accuracy estimates in these studies differ from the accuracy of TTE and TEE in patients with stroke.

The average sensitivity and specificity of TEE in detecting LAT in these studies were 93 and 97 percent, respectively. For TTE, sensitivity and specificity averaged 42 and 99 percent. The low sensitivity of TTE was largely due to missed left atrial appendage thrombi. For the diagnosis of LVT, TTE had an average sensitivity of 78 percent and specificity of 87 percent. When results from individual studies were plotted on a SROC curve, however, it appeared that varying accuracy across studies may have been partly due to differing diagnostic thresholds. Using the SROC curve to estimate the accuracy, the sensitivity and specificity of TTE for diagnosing LVT were 77 and 95 percent, respectively. It should be noted, however, that approximately 15 percent of TTE examinations in studies of LVT were deemed inadequate for interpretation, limiting the diagnostic utility of this test. We did not identify any studies of the accuracy of TEE in diagnosing LVT.

When the prevalence of intracardiac thrombi in patients with stroke is assumed to be 2 percent or less, as many as or more patients would receive unnecessary treatment due to false positive tests than would receive potentially beneficial treatment for a true positive test, if echocardiographic technology is used to select patients for treatment with anticoagulants. Under current estimates of test accuracy, the prevalence of LAT would have to exceed 15 percent and the prevalence of LVT 37 percent in order to achieve 90 percent predictive value.

Studies examining the accuracy of echocardiography in diagnosing ASA and aortic atheroma are lacking. However, given that the association with stroke has been established for the echocardiographic, rather than anatomic, definitions of these lesions, it may be argued that TEE represents the gold standard for the diagnosis of these lesions as they relate to cardioembolic stroke.

Few studies have assessed the accuracy of echocardiography in diagnosing left atrial myxoma. These studies suggest accuracy approaching 100 percent, though one study found disagreement between TTE and TEE in 2 of 11 cases.

Findings of Individual Studies

We excluded three studies reporting TEE complications that evaluated only patients following trauma or being evaluated for aortic dissection,¹⁷¹ only patients who underwent cardiac surgery,¹⁷² and only critically ill patients being evaluated for hemodynamic insufficiency, valvular disease, or abscess.¹⁷³

The largest study reported a death rate of 0.01 percent (1/10,218) and major complication rate (pulmonary, cardiac, or bleeding) of 0.2 percent (18/10,218).¹⁷⁶ In the other large series,^{174, 175, 177, 178} the death rates ranged from 0 percent (0/901)¹⁷⁸ to 0.03 percent (1/2,947),¹⁷⁵ and major complication rates ranged from 0.2 percent¹⁷⁴ to 0.9 percent.¹⁷⁵ In the largest series, 1.9 percent (201/10,419) of TEE procedures were unsuccessfully attempted, and 0.9 percent (90/10,218) had to be interrupted for minor or major complications, most (65/90) because of patient intolerance.¹⁷⁶ Similar findings were reported in other large studies.^{175, 178}

In 11 smaller series (sample size range 35 to 566),^{56, 154, 179-187} no deaths were reported, and the rate of major complications ranged from 0 percent^{154, 179-181, 183, 184} to 2.6 percent (2/77).¹⁸² Like the larger studies, the quality of these reports was poor^{154, 179, 180, 182, 184} or fair.^{56, 181, 183, 185-187}

None of the reviewed studies reported complication rates specifically in patients who had experienced recent cerebral ischemia. Three studies^{183, 186, 187} assessed the incidence of complications in elderly patients (defined as either older than 65 years or 70 years), and three assessed the incidence of complications in "critically ill" patients (managed in intensive care unit or emergency room).^{154, 182, 184} In elderly patients, no deaths were reported (n=406), and the rate of major complications ranged from 0 percent (0/35)¹⁸³ to 1.1 percent (1/88).¹⁸⁷ In one study, there appeared to be an association between advanced age and transient hypotension (5 percent vs. 1.4 percent).¹⁸⁶ In critically ill patients, no deaths were reported (n=219), and the rate of major complications ranged from 0 percent^{154, 184} to 2.6 percent (2/77).¹⁸² The rates of complications for other specific populations (including women, non-white, and lower socioeconomic status) and for specific indications were not reported in the reviewed studies.

Transient, asymptomatic bacteremia has been reported to occur in 0 percent (0/101)¹⁸⁸ to 17 percent (4/24)¹⁸⁹ of patients after TEE. In two case reports,^{190, 191} patients with known valvular or congenital heart disease developed infective endocarditis following TEE. However, in eight prospective studies (sample size range 24 to 144, total n=775),^{188, 189, 192-197} no case of endocarditis or systemic infection following TEE was found. There were major methodological flaws in all of these studies with regard to assessment of endocarditis or other systemic infection, including unclear duration and completeness of followup, inadequately defined complications, and poorly described ascertainment techniques. All of these studies were rated as having either poor^{188, 189, 193, 194, 196} or fair^{192, 195, 197} methodological quality.

Pooled Results

We did not find significant heterogeneity between rates of death reported from 15 series of patients (we did not include data from Seward,¹⁷⁴ which overlapped with a later study¹⁷⁷). The pooled periprocedural death rate was 0.014 percent (4/23,123; 95 percent CI, 0.00004 to 0.04 percent). No deaths were reported in the subgroup of studies reporting rates in elderly patients (n=406) and critically ill patients (n=219).

We found significant heterogeneity (chi-square=42.7, df 14, p<.0001) among rates of major periprocedural complications from all 15 series. The between-study variation was reduced by 52 percent (from 0.78 to 0.38) when quality was included as a dependent parameter in the model, and we did not find significant heterogeneity (chi-square=3.6, df 7, p=0.83) between the rates of major complications from the eight fair-quality studies.^{56, 175, 178, 181, 183, 185-187} Pooled findings from these studies demonstrated a major complication rate of 0.7 percent (36/4679, 95 percent CI, 0.3 to 1.0 percent). This rate is significantly higher (chi-square=26.8, df 1, p<0.00005) than the 0.2 percent major complication rate reported in the largest and only multicenter series.¹⁷⁶

We pooled the results of three studies reporting rates of major periprocedural complications in elderly patients^{183, 186, 187} (test for heterogeneity: chi-square=0.15, df 2, p=0.93) and three studies of critically ill patients^{154, 182, 184} (chi-square=1.17, df 2, p=0.22). The rate of major complications in these studies was 0.4 percent (3/406; 95 percent CI, 0.1 to 1.4 percent) in elderly patients and 0.8 percent (2/219; 95 percent CI, 0.1 to 2.6 percent) in critically ill patients. The rates of major complications in elderly patients and critically ill patients were not significantly different than the overall rates (chi-square=0.30, df 1, p=0.58 for elderly versus overall; chi-square=.10, df 1, p=0.75 for critically ill versus overall).

In eight studies^{188, 189, 192-197} assessing rates of transient bacteremia and subsequent infections, no cases of infective endocarditis or systemic infection were found in 775 patients followed after TEE.

Summary

In observational studies of poor and fair quality, the pooled risk of periprocedural death was 0.014 percent. The risk of death in patients specifically undergoing TEE for evaluation of possible cardiac embolus cannot be directly calculated. Data are insufficient to determine whether the risk of death is higher in elderly or critically ill patients.

From observational studies of fair quality, the average risk of major (requiring treatment) cardiovascular, pulmonary, and gastrointestinal complications from TEE was 0.7 percent. The rates of major complications in elderly and critically ill patients were 0.4 percent and 0.8 percent respectively. Neither of these rates was significantly different from the overall rates. No cases of infective endocarditis or systemic infection were found in 775 patients followed after TEE. Approximately 1.9 percent of TEE are unsuccessfully attempted, and an additional 0.9 percent are stopped for complications, most frequently patient intolerance. The rate of minor complications (most commonly patient intolerance) requiring discontinuation of the procedure was not consistently reported, but appears to be about three times the rate of major complications.

Although the estimates of risk come from studies of poor or fair methodological quality (no included study was assessed as having overall good quality), no other data are available to provide more reliable estimates. Data are insufficient to determine whether complication rates are different in patients presenting with particular indications such as cerebral ischemic syndromes.

Background

Because most ischemic strokes are thought to involve thrombosis with or without embolism, it has long been postulated that anticoagulant therapy may confer benefit when applied to all (i.e., unselected) patients with stroke. Numerous studies have been conducted to test this hypothesis, and reviews of these studies have generally found that any benefit of anticoagulation in reducing recurrent ischemic stroke is offset by a higher risk of hemorrhagic complications.^198-203 These findings, however, were derived largely from studies conducted prior to 1975, pre-dating the routine use of computed tomography (CT) to differentiate ischemic stroke from ICH and the use of the international normalized ratio (INR) to monitor anticoagulant effect. Anticoagulation in patients with unsuspected hemorrhagic stroke and excessive anticoagulation may have contributed to high rates of adverse effects in these studies. Moreover, most of these studies compared anticoagulants to placebo, rather than antiplatelet therapy (e.g., aspirin). Because antiplatelet therapy has become standard care for patients with ischemic stroke,^204-206 placebo comparisons are no longer clinically relevant. The Cochrane Collaboration has planned a review of studies of anticoagulants versus antiplatelet therapy in unselected patients with stroke, but this review has not yet been conducted.²⁰⁷ We therefore conducted a systematic review to determine whether unselected patients with stroke benefit from anticoagulant therapy, as compared to antiplatelet therapy, with the view that if anticoagulation were beneficial for unselected patients, echocardiography would be of limited use in guiding therapeutic management.

We also reviewed the literature addressing the efficacy and harms of anticoagulation in patients with stroke and clinically apparent cardiac disease. Certain cardiac conditions, such as acute MI, prosthetic heart valves, rheumatic heart disease and dilated cardiomyopathy, are thought to place patients at high risk of developing cardioembolic stroke. Evidence of therapeutic benefit with anticoagulants in such patients would limit the usefulness of echocardiography in guiding therapeutic decisionmaking.

Findings of Individual Studies

Summary

For any given patient with stroke, the potential usefulness of echocardiography in detecting a source of cardioembolism depends on the absence of clinically apparent indications for treatment (i.e., anticoagulation). There is substantial evidence, for instance, that for patients with stroke and AF, anticoagulant drugs confer net benefit, making echocardiographic identification of lesions warranting anticoagulation in these patients superfluous. Whether anticoagulation is beneficial in stroke patients without AF is less clear.

There is fair evidence that unselected patients with stroke do not benefit from anticoagulation as compared to antiplatelet therapy. Evidence from a large, fair-quality international trial suggests that subcutaneous heparin given acutely to patients with stroke is not associated with improved outcomes when compared to aspirin. The two therapies used in combination may confer net benefit, but further study is needed to confirm this finding. A good-quality multicenter trial comparing chronic anticoagulation (target INR 1.4-2.8) and aspirin (325 mg) found no differences in either benefits or harms between the two treatments. Another good-quality trial employing higher degrees of anticoagulation (target INR 3.0-4.5) was stopped early due to increased rates of ICH and death with anticoagulation as compared to aspirin.

We found no fair- or good-quality studies examining the effectiveness of anticoagulation in the prevention of recurrent stroke among patients with stroke and cardiac conditions other than AF. Studies of primary stroke prevention among patients with MI suggest that when compared to aspirin, anticoagulation either alone or in combination with aspirin does not confer net benefit. For patients with DCM, evidence regarding anticoagulation for primary stroke prevention comes from observational studies that provide conflicting results. The only good-quality study found that anticoagulation was more effective than aspirin in the primary prevention of stroke, particularly for patients with moderate and severe cardiomyopathy, after acute MI.

Overall, we found fair evidence that neither acute nor chronic anticoagulation confers net benefit, as compared to aspirin, for unselected patients with stroke. We found insufficient evidence to reach conclusions regarding the effectiveness of anticoagulation for secondary prevention of stroke among patients with stroke and clinically apparent cardiac conditions other than AF. Studies of primary prevention suggest that anticoagulation may be beneficial for patients with DCM but is probably not beneficial in patients with MI; however, results from studies of primary stroke prevention may not be generalizable to patients who have already experienced stroke and are candidates for secondary prevention. Given these findings, we conclude that the scope of patients for whom echocardiography may be useful, if it can effectively identify treatable sources of recurrent stroke, includes all stroke patients except those with AF.

Findings

We identified no randomized trials examining the efficacy of anticoagulation for secondary prevention among stroke patients with echocardiographic lesions representing potential sources of cardioembolism. We found one systematic review of studies of patients with stroke found to have PFO, comparing outcomes in patients treated with anticoagulants, surgical PFO closure, and antiplatelet therapy.⁸⁸ Another cohort study examined the effectiveness of anticoagulation in patients with systemic emboli and mobile aortic atheromas.²²⁵

The systematic review of secondary stroke prevention among patients with PFO included five retrospective cohort studies and was of fair quality.⁸⁸ The five studies included 378 patients, whose mean age ranged from 38 to 48 and who were treated with warfarin, antiplatelet agents, or surgical PFO closure. None of the individual studies found significant differences across treatment groups. The pooled annual rates of recurrent stroke or TIA over a mean followup period of three years were 1.0 percent with surgical closure, 2.5 percent with warfarin, and 5.1 percent with antiplatelet therapy. Using meta-analytic methods, the reviewers found the difference between warfarin and antiplatelet therapy to be statistically significant (odds ratio 0.37; 95 percent CI, 0.23 to 0.60). Because the number of patients undergoing surgical PFO closure was small, the differences between surgical closure and either warfarin (odds ratio 1.19; 95 percent CI, 0.62 to 2.27) or antiplatelet therapy (odds ratio 0.36, 95 percent CI; 0.04 to 3.09) were not significant. While these results suggest a protective effect with warfarin and possibly with surgical PFO closure, as compared to antiplatelet therapy, it must be noted that the five individual studies were generally of poor quality, treatment assignment in all studies was discretionary, and the effect of potential confounding by differences in patient characteristics across treatment groups was not assessed.

One poor-quality prospective cohort study evaluated the effect of anticoagulation in 31 patients with systemic embolization found to have mobile aortic atheroma on TEE.²²⁵ Twenty-five of these patients (81 percent) presented with stroke. Twenty patients (65 percent) received warfarin (target INR 2.0) at their physicians' discretion, seven (23 percent) received antiplatelet therapy, and four (13 percent) received neither. During a followup period of 10 months, none of the 20 patients receiving warfarin experienced a recurrent stroke, compared to three of the eleven patients (27 percent) not treated with warfarin (p = 0.07); all three patients with recurrent stroke were taking aspirin. However, three patients (15 percent) on warfarin died, one of them from gastrointestinal hemorrhage, compared to one patient (9 percent) not receiving warfarin (not statistically significant). Overall, there were no significant differences in the combined outcome of any vascular event or death. The authors did not report or adjust for baseline differences between those receiving and those not receiving warfarin.

Because there was limited evidence regarding the efficacy of anticoagulation in the secondary prevention of stroke, we evaluated the evidence for primary prevention of stroke in patients with specific echocardiographic lesions. We identified one systematic review of seven studies addressing whether anticoagulation in patients with mural thrombi after anterior MI reduced embolic complications.⁵⁹ Three of these studies demonstrated a statistically significant reduction of embolic events with anticoagulation. In all of these studies, control subjects received placebo rather than antiplatelet therapy. The pooled odds ratio for the effect of anticoagulation on the incidence of embolic events was 0.14 (95 percent CI, 0.04 to 0.52). However, there was significant heterogeneity across the included studies, making pooling of results of questionable validity. None of the studies was a randomized trial.

A cohort study within the SPAF-III trial included 382 patients with non-valvular atrial fibrillation who underwent TEE.⁵⁶ The primary ischemic event rate was 12.9 percent in patients with demonstrated LAT on a combination therapy of low-dose warfarin and aspirin. In patients on high-dose warfarin, the primary ischemic event rate was 17.9 percent. The odds ratio for this comparison was 2.7 (p <0.05). This unexpected finding of a higher stroke rate among patients receiving high-dose anticoagulation may have been attributable to the fact that TEE was conducted approximately 3 weeks after initiation of therapy. Some patients with LAT may therefore have been effectively treated, with resolution of thrombus by the time TEE was performed, leaving only patients with the most severe thrombi or at highest risk of embolism with evident LAT on TEE. One other study assessing embolization in patients with intracardiac thrombus included too few patients with thrombus (n=5) to allow firm conclusions.²²⁶

Summary

Studies of the effectiveness of anticoagulation for echocardiographically identifiable lesions were all observational in design. Pooled data from five retrospective cohort studies suggest that warfarin, and possibly surgical PFO closure, may reduce the rate of recurrent stroke or TIA among patients with stroke and PFO. However, these studies were generally of poor quality and did not account for differences in baseline characteristics that may have given rise to differences in outcomes across treatment groups. A small, poor-quality cohort study of patients with stroke found to have mobile aortic atheromas revealed a trend towards lower recurrent stroke rates with warfarin as compared to aspirin, but no differences in the combined outcome of all vascular events plus death. A poor-quality systematic review of primary stroke prevention in patients with intraventricular thrombus after acute MI suggested a net benefit with anticoagulation, but the reviewed studies were observational, and no adjustment for potential confounding was conducted. Moreover, whether or not findings from studies of primary stroke prevention among patients with acute MI can be used to draw conclusions regarding secondary prevention among a general population of patients with stroke is not clear. Based on our review, we conclude that there is insufficient evidence to draw conclusions about the effectiveness of anticoagulation in reducing morbidity and mortality among stroke patients with echocardiographically identified lesions.

Echocardiography Decision Model

The echocardiography model follows a hypothetical cohort of patients with newly diagnosed ischemic (non-hemorrhagic) stroke over the course of their remaining lifetimes. The base case represents 65-year-old men with newly diagnosed ischemic stroke who have survived acute stroke treatment, and are now at the point where a decision will be made about further diagnostic testing. Patients are assumed not to have a prior indication for anticoagulation therapy, such as atrial fibrillation. They are also assumed not to have contraindications to anticoagulant therapy. The cohort consists of stroke patients with and without a history of cardiac disease, which is defined variably across studies but generally includes history of myocardial infarction, congestive heart failure, valvular disease, and arrhythmia. Patients diagnosed with intracardiac thrombus, either left atrial or left ventricular, receive anticoagulant therapy. Other echocardiographically identifiable lesions, for which indications for treatment are less clear, are not included in the model. The model excludes the clinically unacceptable course of not treating stroke patients with at least standard medical treatment such as antiplatelet therapy. Its time horizon is 30 years, at which time virtually all patients will have died of stroke-related or other causes. Decisions to initiate or change therapy are based on clinical events alone.

Echocardiographic Testing Strategies

Table 4. Echocardiography model testing strategies

Table 4. Echocardiography model testing strategies

Number	Strategy
1	Treat all with standard medical treatment
2	Treat all with anticoagulation plus standard medical treatment
3	All receive TTE
4	All receive TEE
5	All with cardiac history receive TTE -- others receive standard medical treatment (Selective 1)
6	All with cardiac history receive TEE -- others receive standard medical treatment (Selective 2)
7	All receive TTE -- negative TTE prompts TEE (Sequential 1)
8	All with cardiac history receive TTE -- negative TTE prompts TEE (Sequential 2)
9	All with cardiac history receive TTE, negative TTE receives TEE -- all with no history receive TEE (Selective-sequential)

Figure 3. Echocardiography decision tree

   Figure 3. Echocardiography decision tree

Figure 3

Stroke risk for patients with intracardiac thrombus on TTE or TEE is assumed to depend on the thrombus rather than on other patient characteristics such as age or gender, i.e., the stroke risk for all patients with a left atrial thrombus is the same.

Markov Modeling

We constructed a semi-Markov model of stroke to model the time sequence of disease states, survival, and associated costs among the cohort of stroke patients. Semi-Markov models are decision trees that include Markov nodes, branching points in the tree that lead to a Markov process. In a Markov process, both specific health states and the possible transitions between them are defined. The prognosis of the patient (or cohort) in the analysis is described by the health states, the permissible transitions between states, and the rates of transition. Markov models can illustrate the relationship between the risk reduction of an intervention and the cost of a diagnostic or treatment strategy over the appropriate time horizon. Our models use a generalization of Markov processes in which transition rates between states are not fixed, but rather, can change with time.

Table 5. Health state definitions with associated Rankin Disability (more...)

Table 5. Health state definitions with associated Rankin Disability Scale categories

Name	Rankin	Description
TIA*	0-1	No significant disability; able to carry out all usual activities of daily living
Minor	2	Slight disability; unable to carry out some previous activities, but able to look after own affairs without assistance
Moderate	3	Moderate disability, requiring some help but able to walk without assistance
Severe	4-5	Combination of "Moderately severe disability; unable to walk and attend to own bodily needs without assistance" and "Severe disability; bedridden, incontinent, and requiring constant nursing care and attention"

* TIA: transient ischemic attack; category includes completed stroke with full functional recovery.

Figure 4. Markov diagram--Post-treatment health states (more...)

   Figure 4. Markov diagram--Post-treatment health states (echo model)

Figure 4

A transition (event) is the passage from one state to another. For example, a patient that enters the model at age 65 with a TIA, has a severe ischemic stroke at 75 that lowers him to the severe stroke state, and dies at 80 would spend 10 years in the TIA state, then 5 years in the severe stroke state. We assume that the hypothetical patient cohort initially entering a Markov node is distributed among the stroke-related health states (excluding death) according to the prevalence of stroke severity among new stroke patients. Once in a stroke state, a person can remain in the same state for a given annual cycle, experience a recurrent stroke, die from either stroke-related or non-stroke-related causes, or experience short- or long-term treatment complications. A person experiencing a recurrent stroke that is more severe than the most recent stroke transitions into the more severe health state; e.g., a person initially experiencing a minor stroke, who subsequently has a moderate stroke, transitions from the minor stroke health state to the moderate stroke state. When a recurrent stroke is less severe than the most recent stroke, the person accrues a transient (single-cycle) decrement in utility, but thereafter remains in the prior health state; e.g., a person initially experiencing a moderate stroke who subsequently has a recurrent minor stroke will have a single-cycle decrement in health state but will thereafter remain in the moderate stroke state. When a recurrent stroke is of the same severity as the most recent stroke, the person generally accrues a transient (single-cycle) decrement in utility, but thereafter remains in the prior health state. The exception is the case of a person in the severe stroke health state experiencing a recurrent severe stroke. Because of the high mortality associated with this situation, the person is assumed to transition to the death state. Transitions between health states occur at the end of each cycle.

The model simulates the natural history for a large number of hypothetical patients, each of which is followed for relevant health events until death. During this process, health state dependent flows of costs and utilities are recorded. Hence, each patient generates a survival time, a quality-adjusted survival time, and a cost. Although the time course of any one patient occurs randomly, the selection is guided by the parameters of the decision model. If a large enough group of patients is simulated, the overall outcome pattern should reflect well the simulated population's natural history.

Complications of the testing procedures are limited to a small mortality risk from TEE-related esophageal perforation; TTE is assumed to carry no complication risk. We do not assume a short-term loss of quality of life from esophageal perforation. Tests are administered only once. On the basis of a final positive or negative test result, persons are assigned to a specified treatment regimen -- anticoagulation plus standard medical treatment or standard medical treatment alone.

Treatment-related complications are either gastrointestinal bleeding or ICH. All bleeds are assumed to be short-term, i.e., one annual cycle. Any temporary and meaningful loss of quality of life from a bleed within the first 30 days of treatment is assumed to be minor when averaged over the first year. ICHs can be either short- or long-term; long-term ICHs are considered permanent and to be equivalent in severity to a severe ischemic stroke. Short-term complications can occur repeatedly, but in any case patients experiencing anticoagulant-related complications are assumed to immediately and permanently discontinue anticoagulant therapy. The death state is terminal; with enough time all cohort members would die.

The Markov nodes in the echocardiography model are run up to 360 monthly cycles. We use life tables to establish the age- and gender-specific baseline mortality rate for the cohort and adjust these rates to account for the increased risk of cardiac-related deaths among patients with stroke. The echocardiography model includes eight separate Markov nodes that reflect the various prognoses of patients with or without an intracardiac thrombus, on aspirin or anticoagulant therapy, and experiencing or not experiencing a complication of TEE.

Costs

Table 6. Parameter List -- Echocardiography

Table 6. Parameter List -- Echocardiography

	Baseline	Low	High	Source
Clinical/epidemiologic
Prevalence
Health states after stroke/TIA
TIA/full recovery	0.24	0	0.5	Evidence Tables 1, 2; Matchar (2000)
Minor stroke	0.18	0	0.5	Matchar (2000)
Moderate stroke	0.19	0	0.5	Matchar (2000)
Severe stroke	0.39	0	0.5	Matchar (2000)
Cardiac disease	0.36	0.2	0.6	Evidence Tables 1, 2
Intracardiac thrombus
Unselected population	0.02	0.001	0.2	Evidence Tables 1, 2
Patients with cardiac disease	0.05	0.01	0.36	Evidence Tables 1, 2
Patients without cardiac disease	0.007	0.003	0.033	Evidence Tables 1, 2
Test Accuracy
Left Atrial Thrombus
Sensitivity of TEE	0.93	0.80	1.0	Evidence Table 3
Specificity of TEE	0.97	0.90	1.0	Evidence Table 3
Sensitivity of TTE	0.42	0.15	0.75	Evidence Table 4
Specificity of TTE	0.99	0.95	1.0	Evidence Table 4
Left Ventricular Thrombus
Sensitivity of TEE and TTE	0.78	0.60	1.0	Evidence Table 5
Specificity of TTE and TEE	0.87	0.70	1.0	Evidence Table 5
Recurrent Stroke
Patients with untreated intracardiac thrombus, year 1	0.22	0.04	0.60	Comess (1994), SPAF (1998), Stratton (1987), McNamara (1997)
Relative risk (RR) with anticoagulation, year 1	0.67	0.14	1.0	McNamara(1997), Vaitkus(1993)
Patients without intracardiac thrombus, year 1	0.12	0.08	0.15	Petty (1998, issue 50;1), Hankey (1998, Stroke), Sacco (1994), Burn (1994), Chen (1985)
Annual rate for all patients, after year 1	0.03	0.01	0.10	Petty (1998, issue 50;1), Hankey (1998, Stroke), Sacco (1994), Burn (1994), Chen (1985)
Case fatality of recurrent stroke	0.26	0.22	0.30	Tuomilehto (1992), Jorgensen (1997), Ellekjaer (1997)
Complication Rates
Esophageal perforation--TEE	0.0032	0	0.0064	Evidence Table 6
Case fatality of esophageal perforation	0.04	0.03	0.10	Evidence Table 6
Mortality--TEE	0.0001	0	0.0003	Evidence Table 6
ICH incidence--anticoagulation, year 1	0.014	0.001	0.059	Petty (1999), Landefeld (1989)
ICH incidence--antiplatelet therapy, annual	0.0025	0.0002	0.008	Petty (1999), SPIRIT (1997)
Case fatality of ICH	0.50	0.38	0.64	Berwaerts (2000), SPIRIT (1997), Landefeld (1989)
Proportion of ICH with no recovery (severe)	0.09	0.03	0.20	Berwaerts (2000)
GI bleed incidence--anticoagulation, year 1	0.034	0.008	0.068	Petty (1999), Landefeld (1989)
GI bleed incidence--antiplatetelet therapy, annual	0.018	0.009	0.031	Petty (1999), SPIRIT (1997)
Case fatality of GI bleed	0.06	0.008	0.016	Petty (1999), SPIRIT (1997), Landefeld (1989)
Non-stroke Mortality
Baseline	Actuarial data			U.S. Life tables
Annual excess cardiovascular mortality	0.048	0.012	0.12	Dennis (1993), Hankey (2000)
Relative risk of death by health state				Samsa (1999, J Clin Epi)
TIA/full recovery	1.0	1.0	1.0
Minor stroke	1.11	1.0	1.3
Moderate stroke	1.27	1.05	1.4
Severe stroke	2.04	1.35	3.0
Economic
Cost
Diagnosis
TTE	466	233	932	2001 Medicare fee schedule
TEE	564	282	1,128	2001 Medicare fee schedule
Complications
Esophageal perforation	2,198	1,100	4,400	Seto (1997)
Treatment
Annual cost of anticoagulation therapy	900	450	1,800	Matchar (2000)
Annual cost of antiplatelet therapy (aspirin)	60	30	90	Matchar (2000)
Complications
GI Bleed	5,128	2,564	10,256	Matchar (2000)
ICH-acute	17,840	8,920	35,680	McNamara (1997)
Downstream (i.e., cost of stroke)
Acute recurrent stroke	First 30 days
Minor stroke	2,801	1,400	4,200	Samsa (1999, Stroke)
Moderate stroke	5,077	2,538	10,154	Samsa (1999, Stroke)
Severe stroke or severe ICH	14,706	7,353	20,000	Samsa (1999, Stroke)
Chronic stroke (annual)	Month 2-4	Months 5-12	Months 13+
TIA/full recovery	811	364	339	Samsa (1999, Stroke)
Minor stroke	1298	583	543	Samsa (1999, Stroke)
Moderate stroke	2353	1056	984	Samsa (1999, Stroke)
Severe stroke or severe ICH	6816	3059	2850	Samsa (1999, Stroke)
Terminal costs (i.e., life-saving)	10,000	5,000	20,000	Estimate
Utilities
TIA/full recovery	0.9	0.85	1.0	Matchar (2000)
Minor stroke	0.65	0.6	0.7	Matchar (2000)
Moderate stroke	0.5	0.45	0.55	Matchar (2000)
Severe stroke	0.27	0.22	0.32	Matchar (2000)
Death	0	0	0	Convention
ICH-minor	0.65	0.5	0.9	Assumed (utility for minor stroke)
ICH-severe	0.27	0.22	0.32	Assumed (utility for severe stroke)
GI bleed	0.997	0.98	1.0	Matchar (2000)
Discount rate	3%	0	5%	Gold (1996)

Utilities

Both stroke disability and the side effects of treatment can lower quality of life; therefore, we assigned utility values to each health state, ranging from 0.9 for a TIA to 0 for death. In our model, a recurrent stroke is manifested in a transition to a lower health state, and requires a utility loss, which will be to zero if the new stroke results in death. The utility level assigned to a long-term treatment-related ICH is the same as that assigned to a severe stroke, and that of a short-term ICH the same as that assigned to a minor stroke. Our utility values were based on the final, and predominately time-tradeoff (TTO)-based, values reported by the Stroke Patient Outcomes Research Team (PORT) in its discussion of the Stroke Policy Model.²²⁷ We made two modifications, however. First, although the PORT set the utility for TIA equal to that for asymptomatic patients at 1, we chose to acknowledge a small quality of life decrement for TIA patients relative to persons in perfect health, and therefore, conservatively assigned a value of 0.9 to the utility for TIA. Second, the number of health states in our model was based on PORT prevalence data, which distinguished between TIA and minor, moderate, and severe stroke. To map this qualitative scale to the Rankin scale, we averaged data for Rankin levels 4 and 5 to severe stroke. For utility values, we averaged the PORT utilities for Rankin level 4 (.35) and Rankin level 5 (.20) to generate the 0.27 value we used for severe stroke. The PORT team reported that the mean TTO utility for major stroke from its patient survey was 0.30, and that accounting for the significant number of respondents reporting negative utility reduced the mean to 0.23, supporting our use of 0.27 as the utility for patients experiencing severe stroke.

As stated earlier, we assigned a single-cycle decrement in utility to persons experiencing a recurrent stroke of the same or lower severity than their most recent stroke (e.g., persons in the moderate stroke health state who experience a recurrent minor or moderate stroke). We calculated that decrement by multiplying the person's prior health state utility by the difference: one minus the utility associated with the recurrent stroke. For example, a person in the moderate stroke health state (utility 0.5), who experienced a minor stroke (utility 0.65), would be assigned a decrement in utility of (0.5 x (1-0.65)) = 0.175. The rationale for this calculation is that the person having a recurrent stroke is assumed to have a relative decrement in utility similar to that experienced by a healthy person (utility 1.0) having a first stroke.

Discount Rate

We discounted both costs and utilities using a base rate of 3 percent to reflect the common convention that costs and health benefits incurred or realized in the future are assigned lower values than costs and benefits realized in the present.

Major Assumptions and Estimates

Wherever possible, we used estimates in the cost-effectiveness model that were derived from our systematic review. For parameters not identified in our evidence review and for those for which our review revealed insufficient evidence, we either used the best available estimates from the literature or made informed assumptions, guided by our technical expert group, and tested the effects of those assumptions in sensitivity analyses. Importantly, we identified no evidence supporting the use of anticoagulation in the treatment of LAT, and one poor-quality meta-analysis supporting its use to treat LVT. In a previous cost-effectiveness analysis, McNamara et al. addressed this lack of evidence by assuming that the relative risk reduction from anticoagulation among patients with stroke and documented left atrial thrombus was similar to that among patients with atrial fibrillation (33 percent).⁴ They also assumed no treatment benefit for left ventricular thrombus. Given the results of our literature review, we chose to assume equal benefits of anticoagulation for both lesions. We also adopted the 33 percent relative risk reduction in our baseline results to facilitate comparisons between our results and those of McNamara et al. -- which represents the only published cost-effectiveness analysis on this topic that we were able to identify -- and varied this figure widely in sensitivity analyses. Thus, accurately stated, our analysis addresses the question: "If treatment of intracardiac thrombus were effective, what strategies for testing would be cost-effective?"

Influence of Demographics

We did not have sufficient data on either the incidence of recurrent stroke or the effectiveness of treatment within demographic subgroups to comment meaningfully on the specific influence of demographics. As an alternative, we used data from the U.S. Life Tables produced by the National Center for Health Statistics to adjust the all-causes mortality rate in the model, the baseline of which is for 65-year-old white males. We used the Life Tables to adjust baseline mortality according to various combinations of gender, race (specifically, African American), and age (55, 65, 75, and 85 years old).

Figure 5. Effect of Baseline Life Expectancy on Cost-Effectiveness (more...)

   Figure 5. Effect of Baseline Life Expectancy on Cost-Effectiveness of TEE in Patients with a History of Cardiac Disease (Cardiac History/TEE)

Figure 5

Sensitivity Analysis

Table 8. Sensitivity analyses--Echocardiography

Table 8. Sensitivity analyses--Echocardiography

	Strategy	Cost ($000)	Incremental Cost ($000)	QALY	Incremental QALY	Cost-effectiveness ratio ($000)
TEE cost = $282 ($564 at baseline)	History/TEE	138.5	0.2	3.4525	0.0007	165.6
	Aspirin	138.3	--	3.4518	--	--
TEE cost = $1,128	History/TEE	138.8	0.3	3.4525	0.0000	1,091.0
	History/TTE	138.5	0.2	3.4525	0.0007	337.4
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TEE--LAT = .8	History/TEE	138.6	0.1	3.4525	0.0000	386.7
	History/TTE	138.5	0.2	3.4525	0.0007	337.4
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TEE--LAT = 1	History/TEE	138.6	0.3	3.4525	0.0007	272.6
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TEE--LVT = .6	History/TEE	138.6	0.1	3.4525	0.0000	706.4
	History/TTE	138.5	0.2	3.4525	0.0007	337.4
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TEE--LVT = 1	History/TEE	138.6	0.3	3.4525	0.0007	236.2
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TTE--LAT =. 15	History/TEE	138.6	0.3	3.4525	0.0007	294
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TTE--LAT = .75	History/TTE	138.5	0.2	3.4525	0.0007	223.4
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TTE--LVT =.6	History/TEE	138.6	0.3	3.4525	0.0007	294
	Aspirin	138.3	--	3.4518	--	--
Sensitivity of TTE--LVT =1	History/TEE	138.6	0.1	3.4525	--	2,230.2
	History/TTE	138.5	0.2	3.4525	0.0007	251.3
	Aspirin	138.3	--	3.4518	--	--
Specificity of TEE--LAT = .9	History/TTE	138.5	0.2	3.4525	0.0007	337.4
	Aspirin	138.3	--	3.4518	--	--
Specificity of TEE--LAT = 1	History/TEE	138.6	0.3	3.4525	--	253.4
	Aspirin	138.3	--	3.4518	--	--
Specificity of TEE--LVT = .7	History/TEE	138.6	0.3	3.4525	--	253.4
	Aspirin	138.3	--	3.4518	--	--
Specificity of TEE--LVT = 1	History/TEE	138.6	0.3	3.4525	0.0007	172.7
	Aspirin	138.3	--	3.4518	--	--
Specificity of TTE--LAT = .95	History/TEE	138.6	0.3	3.4525	0.0007	294
	Aspirin	138.3	--	3.4518	--	--
Specificity of TTE--LAT = 1	History/TEE	138.6	0.3	3.4525	0.0007	294
	Aspirin	138.3	--	3.4518	--	--
Specificity of TTE--LVT = .7	History/TEE	138.6	0.3	3.4525	0.0007	294
	Aspirin	138.3	--	3.4518	--	--
Specificity of TTE--LVT = 1	History/TTE	138.5	0.2	3.4525	0.0007	170.4
	Aspirin	138.3	--	3.4518	--	--
Relative risk with anticoagulation = .14	History/sequential	138.8	0.2	3.4559	0.0001	1,645.6
(Baseline = .67)	History/TEE	138.6	0.3	3.4558	0.0040	65.0
	Aspirin	138.3	--	3.4518	--	--
Relative risk with anticoagulation = .3	History/TEE	138.6	0.3	3.4548	0.0030	83.4
	Aspirin	138.3	--	3.4518	--	--
Relative risk with anticoagulation = .5	History/TEE	138.6	0.3	3.4536	0.0018	133.4
	Aspirin	138.3	--	3.4518	--	--
TEE complication rate = 0	History/TEE	138.6	0.3	3.4528	0.0010	243.0
	Aspirin	138.3	--	3.4518	--	--
Prevalence of thrombus with no cardiac	All TEE	138.9	0.4	3.4502	0.0004	964.3
history = .033 (.007 at baseline)	History/TEE	138.5	0.2	3.4498	0.0008	294.0
	Aspirin	138.3	--	3.4490	--	--
Prevalence of thrombus with cardiac history = .1	History/TEE	138.6	0.3	3.4512	0.0024	105.5
	Aspirin	138.3	--	3.4488	--	--
Prevalence of thrombus with cardiac history = .2	History/sequential	138.7	0.2	3.4488	0.0004	400.2
	History/TEE	138.5	0.3	3.4484	0.0056	53.3
	Aspirin	138.2	--	3.4428	--	--
Prevalence of thrombus with cardiac history = .3	History/sequential	138.7	0.2	3.4465	0.0008	176.7
	History/TEE	138.5	0.2	3.4457	0.0090	39.2
	Aspirin	138.3	--	3.4367	--	--
ICH incidence with anticoagulation = (.059 year 1)	Aspirin	138.3	--	3.4517	0.0734	12.6
	Anticoagulation	137.4	--	3.3783	--	--
ICH incidence with anticoagulation = (.001 year 1)	Other sequential	139.1	0.4	3.4536	--	40,009
	History/sequential	138.7	0.1	3.4536	0.0002	940.3
	History/TEE	138.6	0.3	3.4534	0.0016	152.5
	Aspirin	138.3	--	3.4518	--	--
Stroke recurrence rate with thrombus--low estimates	Aspirin	138.3	--	3.4518	--	--
Stroke recurrence rate with thrombus--high estimates	History/sequential	139.1	0.6	3.4828	0.0437	1,643,234.0
	History/TEE	138.5	0.2	3.4391	0.0024	95.2
	Aspirin	138.3	--	3.4367	--	--

Figure 6. Cost-effectiveness of TEE as a function of (more...)

   Figure 6. Cost-effectiveness of TEE as a function of 1) intracardiac thrombus prevalence, and 2) the relative risk (RR) of recurrent stroke with anticoagulation compared to aspirin (for patients with thrombus)

Figure 6

We also examined the sensitivity of the results to other parameters such as cost of TTE, chronic cost of stroke, terminal cost, health state utilities, and the rate and cost of complications, but these had little effect on the results.

Background

The accuracy of non-invasive methods for measuring carotid artery stenosis has received a great deal of attention. Because of the small but persistent complication rate associated with carotid angiography, investigators have worked to establish non-invasive techniques and protocols that can accurately predict high-grade carotid stenosis and thereby reduce or eliminate the need for invasive testing in the selection of candidates for carotid endarterectomy. The non-invasive methods that are most commonly used and that consistently have had the highest accuracy are carotid ultrasonography using Doppler or duplex methods and magnetic resonance angiography.²³¹

Findings

We identified one good-quality systematic review of studies assessing the accuracy of non-invasive measurement of carotid artery stenosis.²³¹ From 568 articles retrieved from electronic and manual searches, the authors reviewed 70 studies, published between 1977 and 1993, that allowed quantification of the operating characteristics of B-mode ultrasonography, carotid Doppler, carotid duplex, magnetic resonance angiography, supraorbital Doppler, and oculoplethysmography, as compared to a reference standard of conventional or digital-subtraction carotid angiography. Pooled results demonstrated the superiority of carotid Doppler, carotid duplex, and MRA over the other three non-invasive methods. The sensitivities of the three superior tests ranged from 0.83 to 0.86, and the specificities from 0.89 to 0.94, with narrow 95 percent confidence intervals. Receiver operating characteristic (ROC) curves demonstrated the equivalence of the three methods.

Although the methodological quality of this meta-analysis was high, we had two major concerns about the applicability of its findings. First, technical advances, particularly for the relatively new technology of MRA, may have made the findings of earlier studies obsolete. Second, most of the studies included in the meta-analysis were subject to verification bias. This bias occurs in studies of diagnostic tests when the results of the test under study are used in determining which patients are submitted for testing with the reference standard. An example is a study examining a retrospective sample of all patients who underwent both carotid duplex examination and carotid angiography during a specified time period, in a setting where carotid duplex examination is routinely used to select patients for angiography. Although all patients in the sample underwent both the test under study and the reference standard, most patients with a negative duplex examination have already been excluded because they were not referred for angiography. Excluding negative duplex examinations from the sample artificially reduces the number of both true and false negative tests, thereby inflating sensitivity and reducing specificity. The impact of this bias on estimates of test accuracy can be significant.^232-234 In one study, correction for verification bias diminished sensitivity from 76 to 58 percent and increased specificity from 85 to 92 percent.²³³

We sought to determine whether the results of more recent studies and studies without verification bias differed from the studies included in the meta-analysis by Blakeley et al.²³¹ We therefore retrieved studies of the accuracy of non-invasive measurement of carotid artery stenosis published after 1993 and included only studies in which verification bias did not appear to be present. We also used the exclusion criteria employed by Blakeley et al. in their meta-analysis, which we felt represented a minimum standard for studies of non-invasive carotid testing. Studies were excluded "if 1) results from the test used were not compared with the results of conventional carotid angiography or intra-arterial digital subtraction carotid angiography; 2) the angiographic results were not separated to allow for specific identification of occluded arteries; or 3) the reference standard test results could not be classified into a contingency table according to degree of stenosis."²³¹

Carotid Ultrasound

We identified one good- and four fair-quality studies of carotid Doppler or duplex^{21, 232, 233, 235, 236} and five studies of MRA^{235, 237-240} -- one good, three fair, and one poor -- that met criteria for inclusion. We found that the operating characteristics of these studies were substantially lower than the pooled results in the previous meta-analysis. These studies were reviewed along with the three studies of CUS^241-243 -- one of poor and two of fair quality -- and one poor-quality study of MRA²⁴⁴ included in the meta-analysis by Blakeley et al. that were not affected by verification bias (Evidence Table 8).

Some studies of the accuracy of CUS used several different criteria, typically based on flow velocity, to diagnose and grade carotid stenosis. When multiple criteria were used, we report the criterion that provided the greatest overall accuracy.

In the largest and only good-quality study of CUS, 1,011 patients enrolled in the NASCET were studied with carotid ultrasound and selective carotid angiography.²³² Patients were referred for the study from 50 academic medical centers throughout North America. Patients were eligible for the study if they had at least 30 percent ipsilateral stenosis by angiography. Carotid ultrasound was performed at the time of angiography but was not used for the purpose of selecting patients for entry into the study. The authors reported on several different Doppler criteria used to diagnose stenosis, all of which aimed to define stenosis of 70 percent or greater, and all of which gave similar results. The overall sensitivity and specificity of CUS were 69 and 68 percent, respectively.

At least four factors may explain the lower accuracy in this study as compared to others. First, because patients were selected for study on the basis of angiographic rather than CUS results, there was little potential for verification bias, which may explain the lower sensitivity in this study as compared with others. Second, patients with less than 30 percent stenosis were not examined. This may have diminished the number of true negative studies and thereby reduced specificity. Third, this study was conducted across multiple centers. Favorable estimates of test accuracy in the literature may result from a tendency for only centers demonstrating high accuracy to publish their results. This hypothesis was validated by investigators in the Asymptomatic Carotid Atherosclerosis Study, who found that among 37 participating centers asked to submit data on the accuracy of CUS for quality assurance purposes, the sensitivity of CUS required to achieve a positive predictive value of 90 percent ranged from 4 percent to 97 percent.³¹ Finally, carotid duplex as performed in the NASCET study involved conventional rather than color-flow imaging. It is possible that the newer method of color-flow duplex is more accurate than the older method used in NASCET. One study comparing the two methods found a trend toward greater accuracy with color-flow imaging,²⁴⁵ while another study found the two methods to be equivalent.²⁴⁶ In the two included studies reporting the accuracy of color-flow duplex in detecting 70 percent or greater stenosis, sensitivities were 88 and 93 percent, and specificities were 84 and 93 percent.^{21, 235}

The NASCET was the only study to report the interobserver reliability of CUS. One neurosonographer reviewed a subset of 102 videotaped CUS studies from multiple centers, using different techniques for grading stenosis. Agreement between the study sonographer and the readings reported from the local centers was generally very high (kappa 0.79 to 0.98), particularly for the most accurate method of grading stenosis (IC PSV/CC PSV ratio, kappa 0.92 to 0.97).²³² It should be noted, however, that the reliability of CUS across different operators was not assessed. Differences in skill and technique across operators are likely to be the greatest source of variation in the use of CUS.

Table 9. Pooled accuracy of carotid ultrasound

Table 9. Pooled accuracy of carotid ultrasound

Degree of stenosis	Studies without Verification Bias				Results from Blakeley et al., 1995
	Sensitivity	Specificity	LR+	LR-	Sensitivity	Specificity	LR+	LR-
>50%	76.7% (53.8-93.3)	87.0% (68.2-97.9)	5.90	0.27	91.5% (87.5-94.4)	91.9% (88.0-94.7)	11.3	0.09
> 70%	73.6% (49.5-91.8)	85.1% (65.3-97.2)	4.94	0.31	87.6% (81.1-92.4)	92.7% (88.0-95.9)	12.0	0.13
> 70% (excluding NASCET data)	76.3% (50.8-94.2)	90.6% (84.3-95.5)	8.12	0.26

LR = likelihood ratio, + = positive test, - = negative test. Values in parentheses are 95% confidence intervals. NASCET: North American Symptomatic Carotid Endarterectomy Trial

Table 9 compares the pooled sensitivity and specificity of CUS in our meta-analysis with results from the meta-analysis by Blakeley et al., which included studies affected by verification bias. We defined positive tests in two ways: angiographic stenoses of 50 percent or more and 70 percent or more as defined by the NASCET method. These represent the thresholds for moderate and severe carotid stenosis, conditions for which the benefits of endarterectomy are quantifiable. We combined studies that reported using a 75 percent cutoff for defining a positive stenosis with those that used a 70 percent cutoff. When studies used the ECST or CC rather than NASCET method for grading angiographic stenosis, we converted the degree of stenosis to its NASCET equivalent, according to a published conversion formula (e.g., 70 percent stenosis by ECST or CC method = 50 percent stenosis by NASCET method).⁴⁴

It should be noted that the sensitivities and specificities across the eight included studies of CUS were significantly heterogeneous. Pooled results may therefore be of questionable value. We have included them in the table for the purpose of comparison with the results of Blakeley et al.²³¹ In their meta-analysis, Blakeley et al. also pooled results of studies for which statistical testing indicated significant heterogeneity. We therefore advise caution in interpreting these estimates, which we use only for the sake of comparison.

Pooled estimates of sensitivity and specificity were lower than those reported by Blakeley et al., particularly for the 70 percent stenosis cutoff. When the NASCET study was excluded, accuracy increased. We had expected that the accuracy of CUS in the studies we reviewed might be higher than in those reviewed by Blakeley et al., as a result of technical advances in ultrasound and increasing operator skill as the technology becomes more mainstream over time. The fact that we did not observe higher accuracy is most likely a reflection of our exclusion of studies affected by verification bias. We also expected that excluding studies with verification bias would decrease overall sensitivity but increase specificity. The fact that both sensitivity and specificity in our analysis were lower than those found by Blakeley may indicate that studies with verification bias also suffer from other biases that inflate accuracy.

Magnetic Resonance Angiography

We reviewed six studies of the accuracy of MRA that appeared not to be affected by verification bias (Evidence Table 9).^{235, 237-240, 244} In four of these studies, CUS was explicitly used to select patients for study.^{237-239, 244} While this does not introduce verification bias per se, because the test under study -- MRA -- was not used to determine whether or not a patient underwent the reference standard test, more patients with high-grade stenosis and fewer patients with minimal or no stenosis may have been selected as a result of pre-screening with CUS. This would increase the proportion of true and false positives and decrease the proportion of true and false negatives, inflating sensitivity and decreasing specificity.

Three of the studies excluded patients with indeterminate results,^237-239 and in one study, several included cases were inexplicably missing from the analysis.²³⁹ We contacted the authors of this study but did not receive a clarifying response. Missing or excluded results may have substantively affected estimates of accuracy in two studies.^{238, 239}

Table 10. Pooled accuracy of magnetic resonance angiography (more...)

Table 10. Pooled accuracy of magnetic resonance angiography

Degree of stenosis	Studies without Verification Bias				Results from Blakeley et al., 1995
	Sensitivity	Specificity	LR+	LR-	Sensitivity	Specificity	LR+	LR-
>50%	86.3% (53.9-99.7)	75.7% (40.5-97.3)	3.55	0.18	93.2% (90.0-95.4)	87.3 (78.1-93.7)	5.58	0.08
>50% (excluding studies using only 2D TOF imaging)	89.6% (81.6-95.5)	94.8% (88.4-98.7)	17.2	0.11
> 70%	94.2% (84.2-99.3)	79.3% (52.0-96.5)	4.55	0.07	89.6% (85.4-92.7)	88.9% (80.1-94.7)	8.07	0.13
>70% (excluding studies using only 2D TOF imaging)	91.8% (72.2-99.8)	86.9% (71.7-96.7)	7.01	0.09

LR = likelihood ratio, + = positive test, - = negative test. Values in parentheses are 95% confidence intervals. 2D TOF indicates two-dimensional time-of-flight.

Pooled estimates of accuracy in the six reviewed studies are reported in Table 10, along with estimates from Blakeley et al. In this analysis we combined 50 and 60 percent stenosis cutoffs and included one cutoff of 80 percent in the group reporting accuracy at 70 percent stenosis.

Three studies assessed accuracy at detecting moderate stenosis (50 or 60 percent or more by NASCET criteria).^{237, 239, 244} Two of the studies were of poor quality and used only two-dimensional time-of-flight (2D TOF) MRA. Most studies suggest that 2D TOF is less accurate than the more frequently studied combination of 2D and 3D TOF, or 3D TOF used alone.

Studies assessing MRA accuracy for detecting severe stenosis (70 or 80 percent or more) demonstrated relatively high sensitivity and specificity using 3D TOF imaging, either alone or in combination with 2D TOF.^{235, 238, 240} Specificity in the single study using 2D TOF alone was substantially lower.²⁴⁴ Interobserver reliability of MRA was reported in three studies and was reasonably high (kappa 0.69 to 0.93).^{235, 237, 238}

It should be noted that although we aimed to reduce the influence of verification bias in our review, our results may reflect optimistic estimates of accuracy for MRA. Because MRA is a developing technology, it is likely that investigators reporting on its operating characteristics are those that have the greatest experience and skill in using this technology. Thus, even if unbiased, these estimates of accuracy may not reflect the overall accuracy of MRA in clinical practice.

Supplemental Analyses

Summary Receiver Operating Characteristic Curves

As discussed earlier in Echocardiography Results, variation in sensitivity and specificity of the same diagnostic test may result from the choice of different diagnostic thresholds. This is particularly relevant for CUS. The diagnosis of carotid stenosis using CUS is typically based on the velocity of blood flow at different sites within the carotid artery and during different phases of the cardiac cycle. Many different velocity criteria are used,²³ and it appears that optimal criteria may vary by equipment and clinical site.^27-30 Because variation in the diagnostic accuracy of CUS often reflects differences in diagnostic thresholds, Summary Receiver Operating Characteristic (SROC) curves provide more appropriate summaries of CUS accuracy than do pooled estimates of average sensitivity and specificity.

Figure 7. Summary receiver operating characteristic (more...)

   Figure 7. Summary receiver operating characteristic (ROC) curves for the diagnosis of carotid artery stenosis using carotid ultrasound (CUS), magnetic resonance angiography (MRA), or both

1. Summary ROC curves for the diagnosis of carotid artery stenosis using CUS. Gray triangle represents data from North American Symptomatic Carotid Endarterectomy Trial (NASCET). Black circles represent all other studies. Dotted curve is for all studies including NASCET. Black curve is for studies excluding NASCET.
   

2. Summary ROC curves for the diagnosis of carotid artery stenosis using MRA. Black circles represent good and fair quality studies. Gray triangles represent poor quality studies. Dotted curve is for all studies. Black curve is for fair and good quality studies.
   

3. Summary ROC curves for the diagnosis of carotid artery stenosis using CUS and MRA. Black circles represent good and fair quality studies. Gray triangles represent poor quality studies. Dotted curve is for all studies. Black curve is for fair and good quality studies. (Dotted curve may not be visible due to overlap with black curve.)

Table 11. Aggregate operating characteristics of carotid ultrasound (more...)

Table 11. Aggregate operating characteristics of carotid ultrasound at maximal accuracy and varying diagnostic thresholds

Degree of stenosis	Diagnostic threshold	Sensitivity	Specificity	LR+	LR-
>50%	Maximal accuracy	79.8%	91.2%	9.07	0.22
	Fixed sensitivity	90% 95% 99%	65.6% 29.0% 1.4%	2.62 1.34 1.00	0.15 0.17 0.71
	Fixed specificity	80.9% 74.6% 56.9%	90% 95% 99%	8.09 14.9 56.9	0.21 0.27 0.44
> 70%	Maximal accuracy	75.4%	87.2%	5.89	0.28
	Fixed sensitivity	90% 95% 99%	52.8% 24.2% 2.0%	1.91 1.25 1.01	0.19 0.21 0.50
	Fixed specificity	72.2% 62.4% 38.2%	90% 95% 99%	7.22 12.4 38.2	0.31 0.40 0.62
> 70% (excluding NASCET data)	Maximal accuracy	94.4%	84.0%	5.90	0.07
	Fixed sensitivity	90% 95% 99%	87.0% 83.4% 72.7%	6.92 5.72 3.63	0.11 0.06 0.01
	Fixed specificity	80.4% 36.9% 0.8%	90% 95% 99%	8.04 7.38 0.80	0.22 0.66 1.00

LR = likelihood ratio, + = positive test, - = negative test. NASCET: North American Symptomatic Carotid Endarterectomy Trial.

The SROC curve for CUS, constructed from the results of studies tabulated in Evidence Table 8, is shown in Figure 7, part A. Most of the studies cluster around the ROC curve. The NASCET study,²³² however, is a prominent outlier, mainly because of lower specificity than that observed in other studies with similar sensitivity. This may have resulted from the exclusion in NASCET of patients with carotid stenosis of 30 percent or less. Exclusion of patients with absent or minimal stenosis likely decreased the number of true negative CUS tests and thereby diminished specificity. Table 11 shows the sensitivity and specificity of CUS for detecting 50 percent and 70 percent stenosis, both at the point of maximal accuracy on the SROC curves for those stenosis cutoffs (see Appendix F), and at varying diagnostic thresholds. Maximal accuracy as derived from the SROC curves is generally higher than the pooled averages of sensitivity and specificity (Table 9). The estimates of test accuracy across varying thresholds quantitatively demonstrate the tradeoff between sensitivity and specificity depicted in the ROC curves. To achieve a sensitivity of 95 percent or higher, one must in most cases accept a substantial fall in specificity, and vice versa.

Table 12. Aggregate operating characteristics of magnetic resonance (more...)

Table 12. Aggregate operating characteristics of magnetic resonance angiography at maximal accuracy and varying diagnostic thresholds

Degree of stenosis	Diagnostic threshold	Sensitivity	Specificity	LR+	LR-
> 70%	Maximal accuracy	91.8%	96.5%	26.2	0.08
	Fixed sensitivity	90% 95% 99%	97.8% 90.0% 21.4%	40.9 9.50 1.26	0.10 0.06 0.05
	Fixed specificity	95.0% 93.0% 86.0%	90% 95% 99%	9.50 18.6 86.0	0.06 0.07 0.14

LR = likelihood ratio, + = positive test, - = negative test.

The SROC curve for MRA (Figure 7, part B) shows a relatively wide range of specificity. This may have occurred because of the use of CUS in some studies to select patients for MRA. If only patients with a positive CUS were included, there would likely be fewer true negative MRA studies, which would limit specificity. Nevertheless, the SROC curve was flat over most of the range of specificities, indicating that studies with low specificity did not have a large impact on the shape of the SROC curve. The single outlier study with the lowest sensitivity was a poor-quality study that included only patients with positive CUS and used only 2D-TOF MRA imaging. Table 12 shows the sensitivity and specificity of MRA for detecting 70 percent stenosis, both at the point of maximal accuracy on the SROC curve for this stenosis cutoff (see Appendix F), and at varying diagnostic thresholds. We did not calculate these estimates for 50 percent stenosis because only two studies examined this cutoff, and one of them was of poor quality and used only 2D-TOF imaging.

Table 13. Aggregate operating characteristics of combined carotid (more...)

Table 13. Aggregate operating characteristics of combined carotid ultrasound and magnetic resonance angiography at maximal accuracy and varying diagnostic thresholds

Degree of stenosis	Diagnostic threshold	Sensitivity	Specificity	LR+	LR-
> 70%	Maximal accuracy	95.1%	98.4%	59.4	0.05
	Fixed sensitivity	90% 95% 99%	99.9% 98.5% 26.2%	900 63.3 1.34	0.10 0.05 0.04
	Fixed specificity	97.5% 96.3% 94.3%	90% 95% 99%	9.75 19.2 94.3	0.03 0.04 0.06

LR = likelihood ratio, + = positive test, - = negative test.

The SROC curve for combined CUS and MRA (Figure 7, part C) shows a narrow range of high sensitivity, with a wider range of specificity. As with analysis of MRA alone, the wider array of specificity may have occurred because of the inclusion criteria of a positive CUS in some studies. Similar to the curve for MRA alone, the SROC curve for CUS plus MRA was flat over most of the range of specificities, indicating that studies with low specificity did not have a large impact on the shape of the SROC curve. All of the studies fell on or very near the curve; there were no outliers. Table 13 shows the sensitivity and specificity of combined CUS and MRA for detecting 70 percent stenosis, both at the point of maximal accuracy on the SROC curve for this stenosis cutoff (see Appendix F), and at varying diagnostic thresholds. We did not calculate these estimates for 50 percent stenosis because only three studies examined this cutoff, and two of them were of poor quality.

ROC curves can be used to statistically compare the accuracy of different tests. However, when the tests being compared are applied to different patient populations in different settings, one must be cautious in drawing conclusions from statistical comparisons.¹⁷⁰ In addition, studies of MRA were generally of poorer quality than those of CUS, and studies of combined CUS and MRA were generally of poor quality and hampered by potential biases. We therefore conclude that the evidence regarding the accuracy of the noninvasive carotid imaging strategies discussed is insufficient to allow comparisons across tests.

Outcomes of Diagnostic Testing

Table 14. Outcomes of diagnostic testing for carotid (more...)

Table 14. Outcomes of diagnostic testing for carotid stenosis in 1000 hypothetical patients with stroke: effect of varying prevalence

Description of Testing Strategies:
Description of Testing Strategies:
"CUS or MRA alone" assumes patients undergo CUS or MRA, undergo CEA if test is positive, and are managed non-surgically if test is negative.
"Angiogram if positive" assumes patients undergo CUS or MRA, undergo cerebral angiography if non-invasive test is positive, and are managed non-surgically if non-invasive test is negative; those undergoing angiography are assumed to undergo CEA if angiography is positive.
"CUS+MRA, angiogram if not concordant" assumes patients undergo both CUS and MRA, undergo CEA if both tests are positive, are managed non-surgically if both tests are negative, and undergo cerebral angiography if one test is positive and the other negative; those undergoing angiography are assumed to undergo CEA if angiography is positive.

	Undergo Angiography	Undergo CEA Inappropriately
	Undergo Angiography	Undergo CEA Inappropriately
Prevalence of stenosis > 70% = 10%			Test Characteristic Assumptions   Sensitivity Specificity CUS: 0.944 0.840 MRA: 0.918 0.965 CUS+MRA: 0.951 0.984
CUS alone	0	144
MRA alone	0	32
CUS, angiogram if positive	238	0
MRA, angiogram if positive	123	0
CUS+MRA, angiogram if not concordant	182	12
Prevalence of stenosis > 70% = 30%
CUS alone	0	112
MRA alone	0	25
CUS, angiogram if positive	395	0
MRA, angiogram if positive	300	0
CUS+MRA, angiogram if not concordant	182	9
Prevalence of stenosis > 70% = 50%
CUS alone	0	80
MRA alone	0	18
CUS, angiogram if positive	552	0
MRA, angiogram if positive	477	0
CUS+MRA, angiogram if not concordant	182	7

In order to assess the potential outcomes of different testing strategies, we calculated the number of patients in a hypothetical cohort of 1,000 patients presenting with stroke or TIA who would undergo carotid angiography and CEA under several testing strategies (Table 14). In this analysis we varied the prevalence of carotid stenosis, defined as 70 to 99 percent narrowing by NASCET criteria, from 10 to 50 percent and assumed that patients with 70 to 99 percent stenosis of the ipsilateral carotid artery by angiography would be appropriate for CEA. We assumed that CEA would be inappropriate for patients with less than 70 percent stenosis, although we acknowledge that many patients with "false positive" tests would have moderate degrees of carotid stenosis (50 to 69 percent) and might receive some benefit from surgery.²²⁹

We assumed a sensitivity and specificity of 94 and 84 percent, respectively, for CUS, which represent the maximal accuracy from the SROC curve excluding the NASCET study (Table 11). We used this favorable estimate of accuracy for CUS to allow a more fair comparison with MRA, for which the generally poorer quality of studies and the lack of large, multicenter studies like the NASCET likely provides a higher estimate of accuracy than would be observed in the community. For MRA, the sensitivity and specificity were 92 and 97 percent, as derived from the SROC curve for MRA (Table 12). Finally, for combined CUS and MRA, sensitivity was assumed to be 95 percent and specificity 98 percent (Table 13). The rate of nonconcordant CUS and MRA was assumed to be 18.2 percent, the average rate across the 11 studies of this diagnostic strategy. Although nonconcordance between CUS and MRA may vary with prevalence of stenosis, there was no clear pattern of correlation between prevalence of stenosis and nonconcordance across the studies we reviewed. We therefore assumed the same rate of test nonconcordance across varying degrees of stenosis prevalence.

The analysis demonstrates that at a 10 percent prevalence of stenosis, when CUS is used alone, the number of patients who would undergo CEA for less than 70 percent stenosis would approach 15 percent of the total cohort. If patients with positive CUS proceeded to cerebral angiography, nearly a quarter of patients would undergo angiography, but because angiography is considered the gold standard for selecting patients for surgery, no patients would inappropriately undergo CEA. The risk of harm with CEA is substantially greater than that of angiography, making angiographic confirmation of CUS the safer strategy at low prevalence. As prevalence rises, more patients would undergo angiography, and fewer would undergo CEA inappropriately. The same pattern is observed for MRA, though because of the assumed superior specificity, fewer patients at each level of prevalence would undergo inappropriate CEA.

Assuming a rate of nonconcordance between CUS and MRA of 18 percent, CUS plus MRA is inferior to MRA alone at low prevalence of stenosis. When patients with positive MRA are referred for angiography, approximately 12 percent undergo angiography, and none undergo CEA inappropriately. With CUS and MRA combined, 18 percent (all with nonconcordant tests) undergo angiography, and 1 percent (those for whom both tests are falsely positive) undergo CEA inappropriately. As prevalence rises, combined CUS and MRA becomes a more favorable strategy in comparison to the others.

Summary

Despite numerous studies of the accuracy of noninvasive carotid imaging, relatively few have been conducted in which all patients undergoing noninvasive tests also undergo diagnostic confirmation with cerebral angiography. The lack of diagnostic verification in these studies creates biased estimates of sensitivity and specificity. Studies can adjust for this bias by angiographically studying a random sample of subjects with negative noninvasive tests. We reviewed studies of CUS and MRA accuracy that either had no obvious or likely verification bias or that adjusted for this bias.

It is clear from the literature that the accuracy of CUS in diagnosing carotid stenosis varies substantially across centers. It is likely that published reports of the accuracy of CUS from single centers overestimate the accuracy in most settings. This has two important implications. First, it may be inappropriate for individual practitioners or medical centers to assume that the accuracy of CUS in their practices is equivalent to published figures. Second, it is clear that there is potential for CUS to be highly accurate. The sensitivity and specificity of CUS estimated from SROC curves constructed from the results of eight predominantly fair-quality studies were 80 and 91 percent, respectively, for moderate or greater (> 50 percent) stenosis, and 75 and 87 percent for severe (> 70 percent) stenosis. When the largest and only good-quality study was excluded, sensitivity and specificity for severe stenosis rose to 94 and 84 percent. The lower accuracy in the largest study than in other studies may have been due to the use of conventional rather than color-flow duplex imaging, but may also have been due to the representation of multiple centers. Reports from single centers may provide biased estimates of accuracy, as those centers finding low accuracy may choose not to submit their results for publication.

Whether the accuracy of MRA varies by center is not clear. There have not been multicenter studies of MRA. Published data, excluding studies with obvious or likely verification bias, suggest a sensitivity and specificity of 92 and 97 percent for detecting severe stenosis. However, studies of MRA were generally of fair to poor quality. As with CUS, it is possible that centers publishing their accuracy data are not representative of all users of MRA. Until there are more high-quality data on the accuracy of MRA, current estimates of MRA accuracy in measuring carotid stenosis must be interpreted cautiously.

All studies of the accuracy of CUS and MRA in combination that we identified were biased by incomplete verification. In the majority of these studies, sensitivity was 100 percent. However, these studies were generally of poor quality. The specificity of combined CUS and MRA was variable, ranging from 69 to 100 percent. The estimated sensitivity and specificity of combined CUS and MRA for detecting severe stenosis were 95 and 98 percent, respectively. In approximately 18 percent of patients, the results of CUS and MRA in detecting severe stenosis were discordant.

Findings of Individual Studies

Because the review²⁴⁸ received a poor quality rating, we carried out a systematic review of prospective studies reporting complication rates of cerebral angiography in patients with symptomatic cerebrovascular disease. We limited our search to prospective studies because of the lower complication rates in the retrospective studies reviewed by Hankey.²⁴⁸

We excluded studies on IV-DSA²⁶² or other outdated methods such as direct carotid or brachial puncture.^{263, 264} Two studies that examined the same populations as other included studies^{265, 266} were excluded, as were two studies in which the complication rates for different arteriographic procedures (e.g., cerebral, vertebral, or aortic angiography) were not reported.^{267, 268} In three prospective studies reporting complication rates of cerebral angiography, the rates in symptomatic patients could not be calculated separately from the rates in asymptomatic patients, and these studies were excluded.^269-271

Four studies were rated as having poor quality,^{249, 252, 254, 257} four fair quality,^{253, 255, 260} and three good quality.^{251, 258, 259} The three studies rated as having good quality included a total of 1,219 symptomatic patients who underwent cerebral angiography. The only methodological concerns identified in these studies were unclear independent ascertainment²⁵⁸ and unclear duration of followup.²⁵¹ The study by Hankey²⁵⁹ appeared to be of overall highest quality, and no major methodological limitations were identified. In this study, the rate of cerebrovascular accident or death was 1.3 percent (5/382). No deaths were reported in any of these studies. The rate of stroke or death ranged from 0 percent (0/637)²⁵¹ to 4 percent (8/200)²⁵⁸ in the other studies rated as having good quality.

In studies rated poor, major areas of concern included inadequate characterization of the study population,^{249, 252} unclear selection methods,^{252, 254} unclear techniques to ascertain complication rates,^{252, 254} unclear length of followup,^{252, 254, 257} biased allocation of patients to more experienced angiographer,²⁵⁶ inadequate description of excluded patients,²⁵⁷ exclusion of patients undergoing angiography outside "regular working hours,"²⁵⁷ and lack of statistical analysis of potential confounders or risk factors.^{249, 252, 254, 257} The major methodological areas of concern in studies rated fair quality were inadequate analysis of potential confounders,^{253, 260} unclear ascertainment methods,^{253, 255, 260} and unclear duration of followup.^{253, 260} Rates of stroke or death in these studies ranged from 0 percent^{252, 253} to 5.7 percent (13/230).²⁵⁶ Only two deaths were reported.

Major complications other than stroke and death (e.g., non-fatal cardiovascular and pulmonary events) were infrequently and inconsistently reported. We did not review local complications that typically do not result in long-term sequelae (e.g., hematoma, dysphagia).

Pooled Findings

We did not find significant heterogeneity (chi-square=6.65, df 9, p=0.67) in the rates of death reported from all 10 studies that reported mortality separately. The pooled rate of death was 0.02 percent (2/3074; 95 percent CI, 0 to 0.1 percent).

There was significant heterogeneity between rates of combined stroke or death from all studies (chi-square=24.8, df 8, p=0.002). In order to evaluate potential explanatory factors for the observed heterogeneity, we stratified results by author departmental affiliation (radiology versus non-radiology), study quality (good quality versus fair or poor quality), and year of publication (prior to 1990 versus 1990 or later). Significant heterogeneity remained after stratifying the results by each of these characteristics. Because one study²⁵⁶ reported an unusually high rate of major periprocedural complications (13/230, 5.7 percent), we performed a test of heterogeneity after excluding this study, and still found significant heterogeneity (chi-square=24.8, df 8, p=0.002). Because of these findings, we did not pool the results of the studies for rates of combined stroke or death.

Risk Factors for Complications

There are few data to assess clinical, demographic, or other factors associated with higher complication rates. Three included studies^{248, 251, 258} performed detailed statistical analysis of potential confounders and risk factors. In a univariate analysis,²⁵¹ recent stroke, frequent TIAs (>1/day), chronic renal insufficiency (creatinine greater than 1.2), increasing age, increasing volume of contrast medium, and longer procedure time were associated with higher complication rates. Davies²⁵⁸ found that the rate of complications was not associated with the presenting symptoms (TIA, amaurosis, or cardiovascular accident [CVA]), though Hankey²⁵⁹ reported an increased rate of periangiographic complications associated with CVA as the indication. Both Davies²⁵⁸ and Hankey²⁵⁹ found increasing degree of carotid stenosis associated with increased rates of complications.

Summary

In our systematic review of prospective studies examining the incidence of stroke and death following cerebral angiography in patients suspected of having cerebrovascular disease and potential candidates for CEA, the overall rate of 0.02 percent for deaths was lower than the 0.08 percent rate previously reported.²⁴⁸ Only two deaths were found in 10 studies including 3,074 patients.

We found significant heterogeneity between rates of combined stroke or death from all studies as well as studies stratified by various methodological criteria. The rate of combined stroke or death ranged from 0 percent to 4 percent in three studies rated as having good quality, with the study rated as highest quality²⁵⁹ reporting a rate of 1.3 percent (95 percent CI, 0.5 to 2.8 percent).

The risk of complications appears higher in patients with greater degrees of carotid stenosis, who are also those most likely to benefit from subsequent CEA. The magnitude of incremental risk of cerebral angiography (i.e., the risk above the baseline risk of recurrent stroke or death in recently symptomatic patients) cannot be reliably estimated at this time but would be expected to be lower than the rates reported above.²⁷²

Study Characteristics and Findings, Randomized Controlled Trials

We identified four randomized controlled trials (RCTs) published since 1980 evaluating the efficacy of "best medical treatment plus carotid endarterectomy" versus "best medical treatment" alone for the treatment of symptomatic carotid artery stenosis.^{229, 230, 273-276} In one study, the trial was aborted after only 41 patients had been entered because a high rate of early complications was observed in the surgical arm.²⁷³ Because CEA was routinely performed with a femoro-carotid shunt, a technique now rarely used, this study was excluded from further analysis.

Of the three other RCTs, two were large multicenter trials^{229, 230, 275, 276} and one a small multicenter trial.²⁷⁴ In the VA Cooperative Study (VACSP), 193 recently symptomatic patients with completed stroke or transient ischemic attack and ipsilateral stenosis greater than 50 percent were randomized.²⁷⁴ In NASCET and ECST, a total of 5,950 recently symptomatic patients with non-disabling stroke or transient ischemic attack and some degree of ipsilateral stenosis were entered, with approximately equal numbers of subjects in the two studies.^{230, 275-277} In all three trials, randomization occurred after cerebral angiography; complications due to cerebral angiography were not reported. Participants were randomized to "best medical treatment" alone, generally an antiplatelet agent and risk factor modification, versus "best medical treatment plus carotid endarterectomy," with surgical technique and timing generally left to the discretion of the participating surgeon.

All three trials were of good quality using USPSTF criteria for grading RCTs, with the following considerations: surgeons and patients could not be blinded to treatment, and outcome assessment did not appear to be blinded. In addition, NASCET screened participating centers for minimum volume of procedures (more than 50 CEAs over the preceding 24 months) and low pre-trial morbidity and mortality (less than 6 percent combined perioperative stroke and death), raising concerns about generalizability to other surgical settings. In the trials, the majority of participants were men (around 70 percent) and relatively young (mean age around 63 to 65 years), with low representation of non-whites (less than 10 percent). Although all participants were considered fit for surgery prior to entry, cardiovascular risk factors typically seen in patients with cerebrovascular disease were well represented. Mean followup was nearly 1 year in VACSP, 2 to 5 years in NASCET, and 6.1 years in ECST. Although NASCET and ECST used different methods to calculate the degree of angiographic carotid stenosis, the estimate of stenosis can be converted from one method to the other using validated formulas.^{44, 45}

In each of the three major trials (ECST, NASCET, VACSP), CEA appeared beneficial in patients with at least moderately severe stenosis. In VACSP, patients with >50 percent stenosis experienced a 1-year absolute risk reduction of 11.7 percent (7.7 percent surgical arm vs. 19.4 percent medical arm) for any stroke or crescendo TIA.²⁷⁴ In ECST, there was a 3-year absolute risk reduction of 11.6 percent (14.9 percent surgical arm vs. 26.5 percent medical arm) for the endpoints stroke or death when the stenosis was 80 to 99 percent (equivalent to 60 to 99 percent by NASCET criteria).²³⁰ In NASCET, for patients with 70 to 99 percent stenosis, there was a 2-year absolute risk reduction of 17 percent (9 percent surgical arm vs. 26 percent medical arm) for any ipsilateral stroke; for patients with 50 to 69 percent stenosis, there was a 5-year absolute risk reduction of 6.5 percent (15.7 percent surgical arm vs. 22.2 percent medical arm) for the same outcome.^{275, 277} In both NASCET and ECST, there was either no benefit or a trend toward harm in patients with lesser degrees of stenosis.^{230, 275, 277}

Study Characteristics and Findings of Systematic Review

A recent systematic review²⁷⁸ examined the methodology and results of RCTs of CEA in symptomatic patients. The systematic review identified the same four RCTs described above, and also excluded the study by Shaw. In addition, the systematic review excluded VACSP in its summary measures because the result of "death or disabling CVA" (the major endpoint of the systematic review) could not be calculated from the reported results.²⁷⁴ Because of generally similar results for less serious outcomes (stroke and crescendo TIA), smaller number of patients and events (particularly deaths), and shorter duration of followup compared to NASCET and ECST, including the results of VACSP would likely not alter the findings of the systematic review. We rated this systematic review as being of good quality, finding no substantial concerns regarding methodology. We did not identify any RCT comparing CEA with medical treatment alone published since the systematic review.

Using pooled data from NASCET and ECST, the systematic review found that CEA plus best medical treatment was effective in reducing the risk of disabling stroke or death compared to medical treatment alone, in patients with symptomatic carotid stenosis greater than 50 percent by NASCET criteria (greater than 70 percent by ECST method).²⁷⁸ The degree of benefit increased with greater severity of stenosis. There was no significant heterogeneity between studies.

For patients with severe stenosis (greater than 80 percent by ECST or 70 percent by NASCET method), surgery reduced the relative risk of disabling stroke or death by 48 percent (95 percent CI, 27 to 73 percent). For patients with moderate stenosis (ECST 70 to 79 percent or NASCET 50 to 69 percent), surgery reduced the relative risk by 27 percent (95 percent CI, 15 to 44 percent). The number needed to treat to prevent one disabling stroke or death over 2 to 6 years was 15 (95 percent CI, 10 to 31) for severe stenosis and 21 (95 percent CI, 11 to 125) for moderate stenosis. For patients with lesser degrees of carotid stenosis (less than 70 percent by ECST or 50 percent by NASCET method), surgery increased the risk of disabling stroke or death by 20 percent (95 percent CI, 0 to 44 percent), with a number needed to harm of 45 (95 percent CI, 22 to infinity).

In both ECST and NASCET, multivariate analysis was performed to determine clinical and demographic factors associated with increased benefit.^{229, 230, 275} In ECST, increased age was associated with improved outcomes by a complex function; women had significantly less benefit from surgery than men.²³⁰ In NASCET, for subjects with 70 to 99 percent stenosis, advanced age (over 70 years old) and male gender were associated with increased benefit. For patients in NASCET with 50 to 69 percent stenosis, age was not associated with increased benefit. In this subgroup, male gender was associated with increased benefit, with the number needed to treat to prevent one disabling stroke or death being 16 for men compared to 125 for women. No significant benefit was seen in the subgroup of women in NASCET with 50 to 69 percent stenosis; endarterectomy reduced the baseline risk of stroke from 15 percent to 14 percent, compared with a reduction of 25 percent to 17 percent in men.

In the subgroup of patients with 70 to 99 percent (i.e., severe) stenosis, the degree of benefit was related to the severity of stenosis. In NASCET, for stenosis 70 to 79, 80 to 89, and 90 to 99 percent, respective absolute risk reductions were 12, 18, and 27 percent, with corresponding numbers needed to treat 8, 5, and 4.²⁷⁵

Although both ECST and NASCET evaluated the relationship between race and outcomes by multivariate analysis and found no significant association, non-whites were underrepresented in these studies.^{230, 275-277}

It must be noted that among patients screened in the NASCET, fewer than one third were randomized.²⁷⁹ Approximately one third did not fulfill baseline criteria, 15 percent were excluded for medical reasons, and another 23 percent were eligible but not randomized. Such exclusions must be considered when trying to generalize data from the endarterectomy trials to individual patients or populations of patients in "real-world" health care settings.

Summary

In two large, good-quality RCTs, carotid endarterectomy reduced the risk of disabling stroke or death for surgically fit patients with symptomatic ipsilateral stenosis greater than 70 percent as measured by ECST, and over 50 percent as measured by NASCET. In a meta-analysis of these trials, the number needed to treat to prevent one disabling stroke or death over 2 to 6 years was 15 (95 percent CI, 10 to 31) for severe stenosis (70 to 99 percent by NASCET criteria or 80 to 99 percent by ECST criteria) and 21 (95 percent CI, 11 to 125) for moderate stenosis (50 to 69 percent NASCET or 70 to 79 percent ECST). No benefit was seen in patients with lesser degrees of carotid stenosis. In the subgroup of patients with severe stenosis, increased degree of stenosis was associated with greater benefit from surgery. The results of the studies are generalizable to surgeons and centers with low perioperative complication rates (30-day stroke or death rate less than 6 percent). The studies did not include angiographic morbidity or mortality in their results.

Although patients over 80 years old, non-whites, and females were underrepresented in these studies, multivariate analysis to determine factors associated with increased benefit was performed on these and other clinical and demographic characteristics in the two RCTs. In NASCET and ECST, less benefit was seen in females for all degrees of carotid stenosis, and in women with 50 to 69 percent stenosis, the absolute risk reduction was eight-fold lower in women than in men. The lesser degree of benefit for women may be partially due to a lower baseline recurrent stroke rate compared to men for equivalent degrees of carotid stenosis.²²⁹ Older age was associated with increased benefit in ECST and in the subgroup of patients in NASCET with 70 to 99 percent stenosis.^{230, 275, 276}

Study Characteristics and Findings, Systematic Review

We identified one systematic review²⁸⁰ that examined the perioperative (30-day) complication rate of CEA in symptomatic patients in studies published since 1980. This systemic review found a total of 51 studies in which the complication rate in symptomatic patients could be calculated separately from the rate in asymptomatic patients.

The systematic review did not rate the quality of the included papers, and included all studies irrespective of methodology. No other major methodological areas of concern were identified using USPSTF criteria for systematic reviews, and the study was given an overall fair quality rating. The summary outcomes were perioperative (30-day) deaths, and perioperative combined deaths or strokes. The study evaluated the effect of the following methodological features on reported complication rates: the use of prospective or retrospective data, author departmental affiliations, independent or non-independent assessment of outcomes, and year of publication.²⁸⁰ Author departmental affiliation and independence of ascertainment were assessed together in four categories (see below).

The systematic review found significant heterogeneity in the reported rates of perioperative complications (chi-square=203, df=49, P<.001). The overall mortality due to CEA for symptomatic stenosis was 1.6 percent (95 percent CI, 1.3 to 1.9, n=17,105); the overall risk of stroke or death was 5.64 percent (95 percent CI, 4.4 to 6.9, n=15,956). Single surgeon author/non-independent assessment studies were associated with the lowest rate of complications (2.3 percent stroke and/or death), followed by multiple surgeon authors/non-independent assessment (5.5 percent), non-surgeon authors/non-independent assessment (6.4 percent), and any author/independent assessment studies (7.7 percent).²⁸⁰ Although there was a trend toward higher complication rates in prospective studies, this was not significant. Including studies published prior to 1980 (from a prior study by the same author), no difference in complication rates according to year of publication was found (including studies published before 1980). If the analysis was limited to studies published since 1980, however, there appeared to be a trend toward higher reported complication rates in more recent studies. The difference in complication rates according to authorship/independent assessment of outcomes was significant when examined in multivariate analysis looking at other study characteristics (retrospective or prospective and year of publication) and was thought to explain much of the heterogeneity in reported rates.²⁸⁰

Findings of Individual Studies

In order to assess the influence of methodological quality and other study characteristics not examined in the prior systematic review,²⁸⁰ we performed a literature search and selective review of studies examining the rate of perioperative complications in patients with symptomatic carotid artery stenosis. We limited our review to studies in which the complication rate for symptomatic patients with carotid stenosis could be calculated separately from the rate in asymptomatic patients, because of evidence that symptomatic patients are at higher risk for perioperative complications.²⁸¹ We identified studies published since 1980 that met our inclusion criteria.

We included all prospective studies in our review. We limited inclusion of retrospective studies to several defined subgroups, because on our initial review of abstracts we found large numbers of retrospective studies that generally appeared to be of poorer quality than the prospective studies. We included population-based retrospective studies because these appeared more likely to have independent assessment, multidisciplinary or non-surgeon authorship, and larger sample sizes than non-population-based retrospective studies. We defined "population-based" as studies that attempted to capture all or a clearly defined subset (e.g., Medicare-insured or a random sample) of patients from a specified geographic area. We also included retrospective single surgeon author studies in order to explore possible reasons for the very low complication rates found in the previous systematic review.²⁸⁰ Because we also reviewed all studies on timing of CEA (see next section), we included these studies, assuming that they represented a non-biased sample of retrospective studies of varying methodological quality that were not population-based or single surgeon author studies. Finally, we included studies classified as prospective by Rothwell²⁸⁰ but re-classified as retrospective after detailed review of study methods. We believed that these studies would provide another non-biased sample of retrospective studies that were not population-based or single surgeon author studies.

We reviewed all 19 studies classified as prospective in the previous systematic review,²⁸⁰ substituting final results from ECST²³⁰ for the earlier interim results²⁷⁶ and Magee²⁸² for Earnshaw²⁸³ (a less detailed report on the same population). In eight of these studies^284-291 we found no clear evidence that the data had been collected prospectively, and we re-classified these studies as retrospective. Two of these "re-classified" studies^{284, 289} were found to be single surgeon author studies. One prospective study was population-based.²⁹² We identified four prospective studies published since 1996^{277, 293-295} and four studies published prior to 1996 not included in the original systematic review,^296-299 resulting in a total of 18 included prospective, non-population-based studies and six "re-classified" retrospective, non-single surgeon author studies.

We identified a total of six single surgeon author studies. Five studies^{284, 300-303} were identified as such in the previous systematic review,²⁸⁰ and one was a study previously classified as having multiple surgeon authors but found on review to have a single surgeon author.²⁸⁹ We did not identify any prospective single surgeon author studies published since 1990. All single surgeon author studies were retrospective.

We identified a total of eight population-based studies. We reviewed the two population-based studies^{292, 304} previously identified by Rothwell, but used data from Brott³⁰⁵ instead of Kempczinski³⁰⁴ because of more detailed reporting of complications from the same population. We identified four additional population-based retrospective case series published since 1996^306-309 and two population-based studies published prior to 1996 not included in the systematic review.^{310, 311} One of the population-based studies²⁹² was prospective.

We included nine retrospective studies^312-320 that examined the association between timing of CEA and perioperative complication rates. We excluded one by Gasecki,³²¹ a post-hoc analysis of NASCET, because its results were reported in other included studies.^{275, 277}

In the studies that we reviewed, we did not distinguish between those routinely using a shunt during surgery, those that used or did not use particular patching procedures, or studies that used different anesthetic techniques. Systematic reviews on these subjects^322-325 have shown inadequate evidence to suggest differences in clinical outcomes between these techniques. We reviewed recent trials on primary closure versus patching³²⁶ and the practice of eversion endarterectomy^{327, 328} and did not find sufficient evidence to suggest that the conclusions of the systematic reviews would be different with their inclusion.

Twelve studies were rated as good quality. Of these, five studies (reporting results from three trials) were RCTs,^{229, 230, 274, 275, 299} six were population-based studies,^{305-307, 309-311} and one was a non-population-based observational study.³²⁹ NASCET and ECST, the two largest RCTs, were found to have no major methodological areas of concern,^275-277 adequately meeting all quality ratings criteria for studies reporting complication rates (see Methods). In these studies, the perioperative complication rates ranged from 6.7 percent (73/1,087)²²⁹ to 8.0 percent (19/327)²⁷⁵ for stroke or death, and 0.6 percent (2/327)²⁷⁵ to 1.3 percent (22/1,787)²⁷⁶ for death alone. In the largest of the good-quality population-based studies, which also had no significant methodological areas of concern, the rate of CVA or death was 6.4 percent (60/943), and death alone 1.7 percent (16/943).³⁰⁷ The single non-population-based study rated as good quality did not adequately define complications and had inadequate statistical analysis of potential confounders, but otherwise appeared to adequately meet all quality ratings criteria.³²⁹ This study reported a 14.5 percent (16/110) rate of CVA or death, and a 3.6 percent (4/110) rate of death alone.

Other important perioperative complications (non-fatal cardiovascular or pulmonary complications) were not routinely assessed or reported, although a rate of 0.9 percent (3/327) was reported in NASCET.²⁷⁵

We rated 21 studies as poor quality^{284, 286-291, 295, 298, 300-303, 312-314, 316, 318-320, 330} and 14 as fair quality.^{282, 285, 292-294, 296, 297, 308, 315, 317, 331-334} All six single surgeon author studies were classified as poor quality.^{284, 289, 300-303} In the studies rated as poor or fair quality, there was substantial variation in reported complication rates. Excluding studies with smaller (n<100) sample sizes, the rates of death ranged from 0.4 percent (1/274)³¹² to 3.5 percent (15/427),³³¹ and the rates of combined stroke or death ranged from 1.1 percent (3/274)³¹² to 10.3 percent.²⁸²

Pooled Findings

We did not pool rates of perioperative outcomes from all studies that we reviewed, because we found significant heterogeneity for the rates of the perioperative outcomes of death alone (chi-square=115, df 64, p<0.0001) in 65 studies and combined stroke or death (chi-square=338, df 68, p<0.00005) in 69 studies (outcomes available from 47 studies that we abstracted and 23 studies previously reviewed by Rothwell²⁸⁰). In the studies we reviewed, we found much less heterogeneity between the results of studies rated good quality for both the outcomes of death alone (12 studies, chi-square=21.5, df 11, p=0.03) and combined stroke or death (nine studies, chi-square=13.8, df 8, P=0.09). Pooled event rates from these studies were 1.6 percent (95 percent CI, 1.0 to 2.5 percent) for death alone and 6.8 percent (95 percent CI, 4.6 to 9.5 percent) for combined stroke or death.

Risk Factors for Perioperative Complications

A systematic review of 14 potential risk factors for perioperative stroke or death from CEA combined data from 36 studies to calculate odds ratios using stepwise logistic regression analysis.³³⁵ The systematic review did not rate the quality of included studies, and included rates for both symptomatic and asymptomatic patients. The review found that the odds of stroke and death were decreased in patients with ocular ischemia alone compared to those with cerebral TIA or stroke (OR 0.49). The odds were increased in women (OR 1.44); subjects aged over 75 years (OR 1.36); and patients with systolic blood pressure over 180 (OR 1.82), peripheral vascular disease (OR 2.19), occlusion of the contralateral internal carotid artery (OR 1.91), stenosis of the ipsilateral internal carotid siphon, and stenosis of the ipsilateral external carotid artery (OR 1.61). Operative risk was not significantly related to presentation with cerebral TIA vs. stroke, diabetes, angina, recent myocardial infarction, current cigarette smoking, or plaque surface irregularity at angiography.

Detailed multivariate analyses of potential risk factors for perioperative complications of CEA were performed in ECST,²⁷⁶ and in a recent good-quality population-based study of Medicare patients.³⁰⁶ The latter found that surgery at a higher-volume hospital was associated with a 71 percent reduction in risk of stroke or death at 30 days after CEA. Other factors associated with increased risk of perioperative complications in multivariate analysis were TIA as indication (OR 2.9), history of angina (OR 2.4), and the presence of renal insufficiency (OR 3.3). Age and gender were not independent predictors of perioperative risk. In ECST, female gender (HR 2.39) and age (HR 0.959/years at randomization) were associated with increased risk 0 to 5 days after surgery.

Supplemental Analyses

We performed several analyses to assess the relative influence of study characteristics on reported complication rates. Because the pooled absolute rate of death did not vary much when studies were stratified according to various study characteristics (range 1.1 to 2.1 percent), we report the stratified results only for the combined endpoint of stroke or death, which showed more substantial variation (range 2.3 to 9.0 percent).

In order to determine whether the conclusions of the previous systematic review²⁸⁰ remain robust after the inclusion of additional studies and re-classified data from studies previously analyzed, we stratified the studies that we abstracted by year of publication, retrospective or prospective collection of data, and by author departmental affiliation. For these analyses, we included the results of studies that we had reviewed (n=46) as well as those studies reported in the previous meta-analysis (n=23).²⁸⁰ In contrast to the findings of Rothwell,²⁸⁰ there was a significantly lower pooled rate of combined stroke and death in retrospective studies (4.5 percent; 95 percent CI, 3.8 to 5.3 percent) compared to prospective studies (7.0 percent; 95 percent CI, 5.3 to 9.0 percent). Like Rothwell, we found a non-significant trend toward a higher pooled rate of complications in more recent (published 1990 or after) studies (5.5 percent; 95 percent CI, 4.2 to 6.9 percent) compared to earlier (1980 to 1989) studies (4.8 percent; 95 percent CI, 3.9 to 5.8 percent). We also found that complication rates were significantly different according to author departmental affiliation: 2.3 percent (95 percent CI, 1.4 to 3.4 percent) for single surgeon author studies, 4.5 percent (95 percent CI, 3.8 to 5.4 percent) for multiple surgeon author studies, and 6.6 percent (95 percent CI, 5.4 to 7.8) for studies with at least one non-surgeon author. Finally, like the previous meta-analysis,²⁸⁰ we found a higher rate of complications in studies with independent ascertainment (7.6 percent; 95 percent CI, 5.9 to 9.6 percent) compared to those with non-independent ascertainment (4.3 percent; 95 percent CI, 3.6 to 5.0 percent).

Table 15. Perioperative (30-day) complication rates (more...)

Table 15. Perioperative (30-day) complication rates of carotid endarterectomy, stratified by study characteristics

	Number of studies	Stroke or death % (95% CI)
Prospective studies	18	4.5% (4.2-4.9%)
Retrospective studies	52	4.9% (4.6-5.3%)
RCTs	8	7.0% (6.2-7.9%)
Not RCTs	62	4.4% (4.2-4.7%)
Population-based	8	6.2% (5.5-6.9%)
Not population-based	54	4.4% (4.2-4.7%)
Single surgeon author	6	2.3% (1.7-3.1%)
Surgeon only author	35	4.1% (3.8-4.4%)
Non-surgeon author	29	6.3% (5.7-6.7%)
Ascertainment independent	20	6.7% (6.1-7.2%)
Ascertainment not independent or unclear      (all studies)	50	4.0% (3.7-4.2%)
Published 1990 or earlier	44	4.1% (3.8-4.4%)
Published 1991 or later	26	6.0% (5.5-6.4%)
Poor quality	21	2.9% (2.6-3.2%)
Fair quality	14	5.1% (4.6-5.7%)
Good quality	12	6.7% (6.2-7.4%)
Selection biased or unclear	9	3.2% (2.8-3.6%)
Selection not biased	38	5.2% (4.9-5.6%)
Population inadequately described	12	3.4% (2.7-4.3%)
Population adequately described	35	4.6% (4.4-4.9%)
Followup incomplete or unclear	0	No data
Followup complete	47	4.5% (4.3-4.8%)
Complications not defined	30	3.5% (3.3-3.8%)
Complications defined	17	6.7% (6.2-7.3%)
Techniques not described	32	3.4% (3.1-3.8%)
Techniques described	15	6.1% (5.7-6.6%)
Ascertainment not independent	27	3.4% (3.1-3.7%)
Ascertainment independent      (abstracted studies only)	20	6.6% (6.1-7.2%)
Statistical analysis inadequate	35	3.9% (3.6-4.2%)
Statistical analysis adequate	12	6.0% (5.5-6.6%)
Duration of followup unclear or inadequate	20	3.4% (3.1-3.8%)
Duration of followup adequate	27	5.5% (5.1-5.9%)

RCTs had a perioperative complication rate of 9.0 percent (95 percent CI, 5.6 to 13.8 percent), and non-RCTs 4.4 percent (95 percent CI, 3.9 to 6.0 percent). Higher complication rates in the RCTs may have been due to better methodological techniques for identifying complications. Population-based studies had a rate of 6.4 percent (95 percent CI, 4.0 to 9.5 percent), and non-population-based studies 5.2 percent (95 percent CI, 4.0 to 6.5 percent).

The association between higher rates of perioperative complications for RCTs, population-based studies, non-surgeon author studies, and prospective studies may be suggestive of an association between assessed study quality and reported complication rates, as these studies all have in common higher average quality.

Validation of Quality Ratings Criteria

To validate the quality ratings criteria that we developed to assess studies reporting complication rates and to further explore the clustering of higher complication rates that we observed in our initial supplemental analyses, we performed analyses on each of the eight quality ratings criteria, as well as on the overall quality rating (poor, fair, or good) derived from these criteria.

Summary

Using data from 12 studies of good quality, the pooled rate of combined perioperative (30-day) stroke or death was 6.8 percent (95 percent CI, 4.6 to 9.5 percent), and from nine studies of good quality, the pooled rate of death alone was 1.6 percent (95 percent CI, 1.0 to 2.5 percent).

In the NASCET trial, the 30-day post-randomization rate of death or stroke ranged from 2.4 percent (for patients with <70 percent carotid stenosis) to 3.3 percent (70 to 99 percent stenosis) in patients assigned to medical therapy.^{275, 277} Therefore, surgery is associated with an additional 35 to 44 perioperative events per 1,000 patients. In NASCET, approximately 60 percent of the strokes that occurred in the perioperative period were non-disabling (Rankin score <3).

There was significant variation in complication rates across studies, which was partly explained by methodological characteristics of the studies. Population-based studies, randomized controlled trials, studies with independent ascertainment of complications, studies with non-surgeon authors, and studies published since 1990 were associated with higher combined complication rates. The pooled complication rate in RCTs was higher than the pooled rate for non-RCTs, suggesting that RCTs may have high generalizability despite strict selection criteria. Population-based studies also reported relatively high perioperative complication rates. All of the characteristics associated with higher complication rates appear to occur in studies rated as having higher average methodological quality.

Background

The timing of carotid imaging in patients with symptomatic cerebrovascular disease has important implications for cost and effectiveness in everyday practice. Published guidelines recommend prompt evaluation of TIAs with carotid imaging (within 1 week), and in some clinically higher-risk patients, hospitalization to facilitate workup.³³⁶ Advocates of early testing argue that if the study is not done during the hospitalization or otherwise expedited, there is a risk that it may never be done because of poor communication between inpatient and outpatient providers, lack of available facilities in some settings, or because the patient fails to seek followup. Keeping patients in the hospital for diagnostic testing, admitting patients with symptoms for testing, and maintaining readily accessible and reliable outpatient carotid imaging services is costly, however, and may not improve rates of diagnostic testing. We did not identify any studies reporting the rates of carotid imaging in patients with recently symptomatic cerebrovascular disease who did not receive "expedited" carotid imaging.

A second argument for early diagnostic testing is that early CEA might be beneficial. The safety of early CEA, however, remains controversial. Early experience with CEA suggested a high rate of complications (in particular ICH) when performed early for symptomatic carotid stenosis, compared with delayed or elective surgery.³³⁷ A recommended waiting period of 4 to 6 weeks evolved from these early observations. More recently, however, some surgeons have questioned the necessity of waiting for 4 to 6 weeks because of clinical observations and published series of early CEA without excess morbidity or mortality. Proponents of early CEA have argued that the time immediately following symptoms may be a high-risk period for recurrent cerebrovascular events, and delays in performing CEA could attenuate the beneficial effects of the procedure. In NASCET, for example, 2.4 percent of medically treated patients with less than 70 percent stenosis and 3.3 percent with 70 to 99 percent stenosis had recurrent stroke or death within 30 days of entering the trial.^{229, 275} Uncertainty also remains regarding the existence and magnitude of excess risk of early CEA in patients with different categories of ischemic brain injury (TIA, CVA with normal or abnormal CT scan). If early CEA is not associated with excess morbidity or mortality, it may be important to obtain diagnostic imaging (i.e., CUS, MRA, or angiography) early after presentation with cerebrovascular symptoms.

Findings

The definition of "early" CEA varied between studies, ranging from less than 2 weeks to less than 6 weeks after presenting with symptoms. In three studies,^{312, 318, 319} patients undergoing early CEA had the procedure within 2 to 3 weeks after becoming clinically stable. No paper specifically addressed the issue of increased risk of performing CEA in a more urgent manner, i.e., within several days to a week of presenting with symptoms. The degree of risk of very early CEA could have a greater impact on the need to perform urgent diagnostic testing of the carotid arteries. Although Paty³¹⁶ reported the rate of complications of CEA performed within one week of symptoms, only 31 procedures were performed in this subgroup.

We excluded one study that reported early CEA in a series of three subjects.³³⁹ Another paper³²⁰ was an update of a previously published report,³⁴⁰ and we used data from the later study. Of the 10 included studies, one was a post-hoc subgroup analysis of patients randomized to CEA (NASCET),³²¹ and the remainder were retrospective reports of patients undergoing early CEA at one or two centers.^{312-314, 318-321} We did not identify any randomized trial specifically designed to address safety of early CEA. Ad-hoc cohorts (i.e., no true inception cohort) of patients undergoing early CEA versus later CEA could be analyzed for all studies except one³¹³ in which there were no data regarding patients undergoing later CEA.

Of the three studies rated as having overall fair quality, one,³²¹ a post-hoc analysis from the NASCET, was rated as having the highest methodological quality. This study included prospective entry of subjects, independent ascertainment of events, and comparable clinical and demographic characteristics between early and later CEA groups. Even in this study, however, methods of selection for early surgery were unclear, and no statistical analysis of potential confounders or risk factors was performed. In this study, the rate of perioperative stroke was 4.8 percent (2/42) for early CEA vs. 5.2 percent (3/58) for later CEA (difference not significant); no deaths were reported in either group.³²¹

In the other included studies, there was no clear pattern of increased complications in either the early or later CEA groups. Some studies^{314, 316, 320} reported an increased rate of complications in the early CEA group; others^{312, 315, 317} reported an increased rate in the later CEA group. In general, the numbers of events and differences in rates between groups was too small to be statistically significant, except for one study³¹⁴ that found a statistically significant difference between a rate of stroke or death of 18.5 percent (5/27) for early CEA vs. 0 percent (0/22) for later CEA. In the one study reporting complication rates in patients undergoing very early CEA (<1 week after the initial cerebral ischemic event), the rate of perioperative stroke was 3.2 percent (1/31); perioperative deaths were not reported according to timing of surgery.³¹⁶

Although the studies in general included stable patients without nondisabling stroke, interpretation of the findings is complicated by differences in the specific inclusion criteria in each study. For example, one study³²¹ included patients with 70 to 99 percent stenosis with or without abnormal CT scans; another³¹² studied only those with normal CT scans. Other studies did not adequately define patient characteristics, making it difficult to generalize their results, as the risk of perioperative complications may vary according to the presenting cerebrovascular symptoms. In those studies^{312, 318, 321} reporting outcomes according to findings on CT scan, the number of events was too low and the difference in complication rates too small to detect significant differences in perioperative complications for early CEA according to CT scan findings. Results were also not reported for important clinical subgroups, including those with greater degrees of symptomatic carotid stenosis.

Pooled Findings

The overall (later and early CEA combined) complication rates (death 1.1 percent [16/1451], stroke or death 3.1 percent [56/1815]) from these studies are comparable to the CEA complication rates reported in other recent studies of comparable quality which did not specifically address the issue of timing.

Supplemental Analyses

In studies with non-surgeon authors, early CEA was not associated with a higher pooled risk of adverse events (early CEA 4.2 percent [95 percent CI, 0.5 to 14.3 percent]; later CEA 5.2 percent [95 percent CI, 2.3 to 9.8 percent]). In studies with only surgeon authors, in contrast, the pooled risk of early CEA (4.0 percent; 95 percent CI, 1.2 to 8.9 percent) was higher than the pooled risk of later CEA (1.9 percent; 95 percent CI, 0.8 to 3.3 percent), though this result was not significant. Publication bias could explain this trend if surgeon authors were more likely to publish reports showing poorer outcomes of early CEA.

In studies published prior to 1990, early CEA was associated with higher pooled complication rates (5.8 percent; 95 percent CI, 1.3 to 14.8 percent) than later CEA (2.4 percent; 95 percent CI, 0.9 to 4.6 percent). In contrast, in studies published in 1990 or later, the pooled rate of complications was lower for early CEA (2.7 percent; 95 percent CI, 0.5 to 7.3 percent) than for later CEA (4.0 percent; 95 percent CI, 1.3 to 7.4 percent). Neither of these differences was statistically significant. The pooled rate from pre-1990 studies is not significantly different from the pooled rate of post-1990 studies (chi-square=1.31, df 1, p=0.25), though the results suggest a trend toward improved selection methods for early CEA or better surgical techniques over time.

Summary

There is fair evidence that early CEA is not associated with an increased risk of major complications. Three non-randomized studies of fair quality suggest that in patients with recent minor or non-disabling stroke, CEA performed earlier than the traditional waiting period of 4 to 6 weeks is not associated with significantly increased adverse events compared to delayed surgery, with a pooled rate of 3.3 percent for early CEA versus 5.3 percent for later CEA. When data from all studies (including seven rated poor quality) are included, the rate of pooled complications is 3.9 percent for early CEA versus 2.7 percent for later CEA. The pooled rate of death alone from all studies was about 1.0 percent in patients undergoing either early or later CEA. There was a non-significant trend towards better outcomes for early CEA in studies published since 1990.

There is insufficient evidence to draw conclusions regarding the risk of very early CEA (i.e., less than 1 week after presenting with symptoms). There is also inadequate evidence to draw conclusions for specific subgroups, including patients with specific CT scan findings and greater degrees of carotid stenosis. Patients selected for early CEA in these studies are likely to comprise an overall lower-risk population compared to patients not selected for early CEA, though in higher-quality studies patients undergoing early and later CEA were comparable according to important clinical and demographic criteria.

Carotid Imaging Decision Model

The carotid imaging decision model follows a hypothetical cohort of patients with newly diagnosed, anterior circulation (carotid territory), ischemic (non-hemorrhagic) stroke over the course of their remaining lifetimes. The base case represents 65-year-old men with newly diagnosed ischemic stroke who have survived acute stroke treatment and are now at the point where a decision will be made about further diagnostic testing. They are assumed to have neither a prior indication for nor contraindication to CEA. The cohort is comprised of patients with 0 to 49 percent stenosis, 50 to 69 percent stenosis, 70 to 99 percent stenosis, or 100 percent occlusion. The model excludes the clinically unacceptable course of not providing at least standard medical treatment to all stroke patients. Its time horizon is 30 years, at which time virtually all patients will have died of stroke-related or other causes. Decisions to initiate or change therapy are based on clinical events alone.

Carotid Testing Strategies

Table 16. Carotid imaging model testing strategies

Table 16. Carotid imaging model testing strategies

Number	Strategy
1	Test no one -- standard medical treatment
2	All receive cerebral angiography -- treat positives (50%-99% stenosis)
3	All receive cerebral angiography -- treat positives (70%-99% stenosis)
4	All receive CUS -- treat positives (50%-99% stenosis)
5	All receive CUS -- treat positives (70%-99% stenosis)
6	All receive MRA -- treat positives (50%-99% stenosis)
7	All receive MRA -- treat positives (70%-99% stenosis)
8	All receive CUS -- angiography for positives (50%-99% stenosis)
9	All receive CUS -- angiography for positives (70%-99% stenosis)
10	All receive MRA -- angiography for positives (50%-99% stenosis)
11	All receive MRA -- angiography for positives (70%-99% stenosis)
12	All receive CUS/MRA -- treat positives (50%-99% stenosis); angiogram for disparate results
13	All receive CUS/MRA -- treat positives (70%-99% stenosis); angiogram for disparate results

Figure 8. Carotid Imaging decision tree

   Figure 8. Carotid Imaging decision tree

Numbers in parentheses refer to Table 16. *SMT -- standard medical treatment CEA -- carotid endarterectomy

The various testing strategies are listed in Table 16. The objective of testing is to accurately identify those stroke patients who are most likely to benefit from CEA. The baseline strategy is to provide standard medical treatment, including antiplatelet therapy (aspirin), to everyone (strategy 1). Alternate strategies include testing everyone using cerebral angiography (strategies 2 & 3), CUS (strategies 4 & 5), or MRA (strategies 6 & 7), with positives receiving a CEA. Strategies 8 to 11 are sequential and involve testing everyone with either CUS or MRA, then following up with cerebral angiography when the noninvasive test is positive. Finally, strategies 12 to 13 involve performing CUS and MRA jointly on everyone, reserving cerebral angiography for cases where CUS and MRA results differ. For each primary testing strategy, we examine the effect of including as "test positive" (and therefore referring for CEA) those with 70 to 99 percent stenosis (severe), and those with 50 to 99 percent stenosis, which includes moderate stenosis as well (Figure 8).

Markov Modeling

Figure 9. Markov diagram--Post-treatment health states (carotid (more...)

   Figure 9. Markov diagram--Post-treatment health states (carotid model)

The Markov nodes in the carotid imaging decision model are similar to those used in the echocardiography decision model and include five stroke-related health states: TIA, minor ischemic stroke, moderate ischemic stroke, severe ischemic stroke, and death (Figure 9). Transitions between states are as described for the echocardiography model. CUS and MRA are assumed not to be associated with significant complications. Cerebral angiography has a complication risk ranging from transient neurologic events to permanent neurologic deficits and death. Persons undergoing CEA assume a risk of periprocedural MI, stroke, or death. Persons who experience a MI undergo a permanent decrement in utility due to the MI.

The Markov nodes in the carotid imaging model are run up to 360 monthly cycles. As with the echocardiography model, we use life tables to establish the age- and gender-specific baseline mortality rate for the cohort and adjust these rates to account for the increased risk of cardiac-related deaths among patients with stroke. These rates are assumed to be highest for those with severe carotid stenosis or occlusion, intermediate for those with moderate stenosis, and lowest for those with mild or no stenosis. The carotid imaging model includes nine separate Markov nodes that reflect the various prognoses of person with four different degrees of stenosis (mild, moderate, severe, occluded), receiving standard medical treatment alone or along with CEA, and those experiencing complications of angiography, who are assumed not to undergo CEA.

Costs

Table 17. Parameter List -- Carotid Imaging

Table 17. Parameter List -- Carotid Imaging

	Baseline	Low	High	Source
Clinical/epidemiologic
Prevalence
Health states after stroke/TIA
TIA/full recovery	0.24	0	0.5	Evidence Tables 3.1, 3.2; Matchar(2000)
Minor stroke	0.18	0	0.5	Matchar(2000)
Moderate stroke	0.19	0	0.5	Matchar(2000)
Severe stroke	0.39	0	0.5	Matchar(2000)
Stenosis among symptomatic patients
0%-49%	0.55	0.2	0.6	Inference Lindgren(1994, issue 25;12),
50%-69%	0.2	0.1	0.4	Bogousslavsky(1988), Hankey (1990, BMJ), Alexandrova(1996) Lindgren(1994, issue 25;12),
70%-99%	0.15	0.05	0.5	Bogousslavsky(1988), Hankey (1990, BMJ), Alexandrova(1996) Lindgren(1994, issue 25;12),
Occlusion	0.1	0.02	0.2	Bogousslavsky(1988), Hankey (1990, BMJ), Alexandrova(1996)
Annual progression from 50-69% to 70-99%	0.043	0.02	0.07	Matchar(2000)
Test Accuracy
CUS
Sensitivity, 50%-99% stenosis	0.8	0.5	0.95	Evidence Table 4.1
Specificity, 50%-99% stenosis	0.91	0.7	1	Evidence Table 4.1
Sensitivity, 70%-99% stenosis	0.75	0.5	0.9	Evidence Table 4.1
Specificity, 70%-99% stenosis	0.87	0.65	1	Evidence Table 4.1
MRA
Sensitivity, 50%-99% stenosis	0.91	0.5	1	Evidence Table 4.2
Specificity, 50%-99% stenosis	0.74	0.4	1	Evidence Table 4.2
Sensitivity, 70%-99% stenosis	0.92	0.75	1	Evidence Table 4.2
Specificity, 70%-99% stenosis	0.97	0.5	1	Evidence Table 4.2
CUS + MRA
Sensitivity, 50%-99% stenosis	0.96	0.75	1	Evidence Table 4.3
Specificity, 50%-99% stenosis	0.95	0.75	1	Evidence Table 4.3
Proportion of discordant tests, 50%-99% stenosis	0.2	0.03	0.45	Evidence Table 4.3
Sensitivity, 70%-99% stenosis	0.95	0.75	1	Evidence Table 4.3
Specificity, 70%-99% stenosis	0.98	0.75	1	Evidence Table 4.3
Proportion of discordant tests, 50%-99% stenosis	0.18	0.1	0.3	Evidence Table 4.3
Recurrent Stroke
0%-49% stenosis
Standard medical treatment
Nondisabling (minor) stroke
30 days	0.0062	0.0012	0.0155	Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.027	0.0054	0.0676	Barnett(1998), ECST(1998)
Disabling stroke
30 days	0.0088	0.0018	0.0195	Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.014	0.0028	0.035	Barnett(1998), ECST(1998)
Fatal stroke
30 days	0.0012	0.0002	0.0029	Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.0074	0.0015	0.0185	Barnett(1998), ECST(1998)
CEA (false positive)
Nondisabling (minor) stroke
30 days	0.0351	0.004	0.0405	Evidence Table 14, Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.027	0.0054	0.0676	Barnett(1998), ECST(1998)
Disabling stroke
30 days	0.0169	0.002	0.0195	Evidence Table 14, Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.014	0.0028	0.035	Barnett(1998), ECST(1998)
Perioperative or stroke-related death
30 days	0.016	0.004	0.04	Evidence Table 14
30 days to 4 years (annual rate)	0.0074	0.0015	0.0185	Barnett(1998), ECST(1998)
50%-69% stenosis
Standard medical treatment
Nondisabling (minor) stroke
30 days	0.0072	0.0014	0.0179	Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.0389	0.0078	0.0972	Barnett(1998), ECST(1998)
Disabling stroke
30 days	0.0124	0.002	0.0198	Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.0191	0.0038	0.0478	Barnett(1998), ECST(1998)
Fatal stroke
30 days	0.0014	0.0003	0.0036	Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.0101	0.002	0.0253	Barnett(1998), ECST(1998)
CEA
Nondisabling (minor) stroke
30 days	0.0348	0.004	0.0402	Evidence Table 14, Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.029	0.0058	0.0725	Barnett(1998), ECST(1998)
Disabling stroke
30 days	0.0172	0.002	0.0198	Evidence Table 14, Barnett(1998), ECST(1998)
30 days to 4 years (annual rate)	0.0077	0.0015	0.0192	Barnett(1998), ECST(1998)
Perioperative or stroke-related death
30 days	0.016	0.004	0.04	Evidence Table 14
30 days to 4 years (annual rate)	0.0053	0.0011	0.0133	Barnett(1998), ECST(1998)
70%-99% stenosis
Standard medical treatment
Nondisabling (minor) stroke
30 days	0.0144	0.0029	0.0361	NASCET(1991), ECST(1998)
30 days to 4 years (annual rate)	0.0572	0.0114	0.143	NASCET(1991), ECST(1998)
Disabling stroke
30 days	0.0036	0.0007	0.009	NASCET(1991), ECST(1998)
30 days to 4 years (annual rate)	0.0462	0.0092	0.1156	NASCET(1991), ECST(1998)
Fatal stroke
30 days	0.0018	0.0004	0.0045	NASCET(1991), ECST(1998)
30 days to 4 years (annual rate)	0.0128	0.0026	0.0319	NASCET(1991), ECST(1998)
CEA
Nondisabling (minor) stroke
30 days	0.0292	0.0034	0.0337	Evidence Table 14, NASCET(1991), ECST(1998)
30 days to 4 years (annual rate)	0.0208	0.0042	0.052	NASCET(1991), ECST(1998)
Disabling stroke
30 days	0.0228	0.0026	0.0263	Evidence Table 14, NASCET(1991), ECST(1998)
30 days to 4 years (annual rate)	0.0096	0.0019	0.0239	NASCET(1991), ECST(1998)
Perioperative or stroke-related death
30 days	0.016	0.004	0.04	Evidence Table 14
30 days to 4 years (annual rate)	0.0056	0.0011	0.0141	NASCET(1991), ECST(1998)
Occlusion
Standard medical treatment
Nondisabling (minor) stroke
30 days	0	0	0.0361	Faught (1993)
30 days to 4 years (annual rate)	0.0096	0.0019	0.0241	Faught (1993), Matchar (2000)
Disabling stroke
30 days	0	0	0.009	Faught (1993)
30 days to 4 years (annual rate)	0.0311	0.0062	0.0777	Faught (1993), Matchar (2000)
Fatal stroke
30 days	0	0	0.0045	Faught (1993)
30 days to 4 years (annual rate)	0.0143	0.0029	0.0358	Faught (1993), Tuomilehto(1992), Jorgensen(1997), Ellekjaer(1997)
CEA (false positive)
Nondisabling (minor) stroke
30 days	0.0292	0.0034	0.0337	Evidence Table 14, NASCET(1991), ECST(1998)
30 days to 4 years (annual rate)	0.0096	0.0019	0.0241	Faught (1993), Matchar (2000)
Disabling stroke
30 days	0.0228	0.0026	0.0263	Evidence Table 14, NASCET(1991), ECST(1998)
30 days to 4 years (annual rate)	0.0311	0.0062	0.0777	Faught (1993), Matchar (2000)
Perioperative or stroke-related death
30 days	0.016	0.004	0.04	Evidence Table 14
30 days to 4 years (annual rate)	0.0143	0.0029	0.0358	Faught (1993), Tuomilehto(1992), Jorgensen(1997), Ellekjaer(1997)
Annual recurrent stroke after 4 years, all
Nondisabling (minor) stroke	0.0053	0.0011	0.0131	Petty(1998, issue 50;1), Hankey(1998, Stroke), Sacco(1994), Burn(1994), Chen(1985), Matchar(2000) Petty(1998, issue 50;1), Hankey(1998, Stroke),
Disabling stroke	0.0169	0.0034	0.0424	Sacco(1994), Burn(1994), Chen(1985), Matchar(2000) Petty(1998, issue 50;1), Hankey(1998, Stroke),
Fatal stroke	0.0078	0.0016	0.0195	Sacco(1994), Burn(1994), Chen(1985), Tuomilehto(1992), Jorgensen(1997), Ellekjaer(1997)
Proportion of disabling strokes that are moderate	0.33	0.20	0.70	Matchar (2000)
Proportion of disabling strokes that are severe	0.67	0.30	0.80	Matchar (2000)
Rate of perioperative (30-day) stroke or death	0.068	0.01	0.10	Evidence Table 14
Rate of perioperative (30-day) death	0.016	0.004	0.04	Evidence Table 14
Years of risk reduction with CEA,50-99% stenosis	4	2	10	Barnett (1998), ECST, 1998
Complication rates
Stroke after cerebral angiography	0.0128	0.005	0.027	Evidence Table 13
Death after cerebral angiography	0.0002	0	0.001	Evidence Table 13
MI after CEA	0.009	0.001	0.02	NASCET (1991)
Non-stroke Mortality
Baseline	Actuarial data			U.S. Life tables
Annual excess cardiovascular mortality				Norris (1991), Dennis (1993), Hankey (2000)
0-49% stenosis	0.026	0.012	0.12
50-69% stenosis	0.048	0.012	0.12
70-100% stenosis	0.095	0.012	0.12
Relative risk of death by health state				Samsa (1999, J Clin Epi)
TIA/full recovery	1.0	1.0	1.0
Minor stroke	1.11	1.0	1.3
Moderate stroke	1.27	1.05	1.4
Severe stroke	2.04	1.35	3.0
Economic
Cost ($)
Diagnosis
Carotid ultrasound (duplex)	225	112	450	2001 Medicare fee schedule
MRA	1,249	625	2,500	2001 Medicare fee schedule
Cerebral angiography	3,238	1,619	6,476	Matchar(2000)
Treatment
CEA	19,312	10,000	25,000	Matchar(2000)
Downstream (i.e., cost of stroke)
Acute recurrent stroke
Minor stroke	2,801	1,400	4,200	Samsa (1999, Stroke)
Moderate stroke	5,077	2,538	10,154	Samsa (1999, Stroke)
Severe stroke	14,706	7,353	20,000	Samsa (1999, Stroke)
Chronic stroke (annual)	Month 2-4	Months 5-12	Months 13+
TIA/full recovery	811	364	339	Samsa (1999, Stroke)
Minor stroke	1298	583	543	Samsa (1999, Stroke)
Moderate stroke	2353	1056	984	Samsa (1999, Stroke)
Severe stroke	6816	3059	2850	Samsa (1999, Stroke)
Terminal costs (i.e., life-saving)	20,000	5,000	30,000	Estimate
Utilities
MI	0.73	0.62	0.83	Matchar (2000)
TIA/full recovery	0.9	0.85	1	Matchar (2000)
Minor stroke	0.65	0.5	0.7	Matchar (2000)
Moderate stroke	0.5	0.3	0.5	Matchar (2000)
Severe stroke	0.27	0.1	0.3	Matchar (2000)
Death	0	0	0	Convention
Discount rate	3%	0	5%	Gold(1996)

As in the echocardiography model, we take the perspective of direct medical costs related to stroke evaluation and management, and exclude indirect costs related to lost work productivity. Cost estimates were derived from best estimates from the literature or from Medicare fee schedules (Table 17). All financial outcomes are adjusted to 2000 dollars using the medical component of the Consumer Price Index. For the purposes of calculation, the model assumes that events occur in the middle of each year.

Major Assumptions and Estimates

Table 18. Baseline carotid imaging results

Table 18. Baseline carotid imaging results

Strategy	Cost ($000)	Incremental Cost ($000)	QALY	Incremental QALY	Cost-effectiveness ratio ($000)
Angio 50%	153.5	0.1	3.7114	-0.0005	Dominated
MRA 50%	153.4	1.1	3.7119	-0.0086	Dominated
MRA/Angio 50%	152.3	0.1	3.7205	-0.0037	Dominated
Joint/Angio 50%	152.2	2.2	3.7242	0.0039	577.0
CUS/Angio 50%	150.0	0.0	3.7203	0.0199	522.8
Angio 70%	150.0	0.2	3.7004	-0.0194	Dominated
CUS 50%	149.8	1.3	3.7198	0.0045	930.4
Joint/Angio 70%	148.5	0.6	3.7153	-0.0021	Dominated
MRA/Angio 70%	147.9	0.3	3.7174	-0.0001	Dominated
MRA 70%	147.6	-0.1	3.7175	0.0088	211.1
CUS 70%	147.7	1.1	3.7087	-0.0037	Dominated
CUS/Angio 70%	146.6	3.2	3.7124	0.0114	278.9
No testing	143.4	--	3.7010	--	--
After all dominated alternatives are removed:
Joint/Angio 50%	152.0	4.0	3.7242	0.0067	692.8
MRA 70%	148.0	5.0	3.7175	0.0165	257.9
No testing	143.0	--	3.7010	--	--

Wherever possible, we used estimates in the cost-effectiveness model that were derived from our systematic review. For parameters not identified in our evidence review and for those for which our review revealed insufficient evidence, we either used the best available estimates from the literature or made informed assumptions, guided by our technical expert group, and tested the effects of those assumptions in sensitivity analyses. Major assumptions in our base-case analysis are listed in Table 17. Other assumptions included: the prognosis of a patient with occlusion is the same as that of a patient with severe (70 to 99 percent) stenosis receiving standard medical treatment; the duration during which the rate of recurrent stroke among patients with moderate or severe stenosis is lower for those receiving as compared to those not receiving CEA is four years, and after four years, the rate of recurrent stroke is similar between the two groups, such that their stroke-free survival curves run parallel; the severity distribution of non-fatal strokes is the same as that for initial strokes, i.e., 24 percent minor, 25 percent moderate, and 51 percent severe; the annual excess cardiovascular mortality among patients with stroke varies by degree of stenosis but does not vary with stroke severity; and persons with more severe strokes, and therefore worse health states, have a higher rate of death not attributable to cardiac causes or recurrent stroke.

Results

Table 19. Sensitivity analyses--Carotid imaging

Table 19. Sensitivity analyses--Carotid imaging

	Strategy	Cost ($000)	Incremental Cost ($000)	QALY	Incremental QALY	Cost-effectiveness ratio ($000)
Cost of CUS = $112 ($225 at baseline)	Joint/Angio 50% MRA 70% Aspirin	152.2 147.7 143.4	4.5 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	686.5 257.8 --
Cost of CUS = $450	Joint/Angio 50% MRA 70% Aspirin	152.5 147.7 143.4	4.8 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	737.9 257.8 --
Cost of MRA = $625 ($1,249 at baseline)	Joint/Angio 50% MRA 70% Aspirin	151.7 147.0 143.4	4.7 3.6 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 220.0 --
Cost of MRA = $2,500	Joint/Angio 50% CUS/Angio 50% CUS 50% CUS/Angio 70% Aspirin	153.5 150.0 149.8 146.5 143.4	3.5 0.2 3.3 3.1 --	3.7242 3.7200 3.7200 3.7125 3.7008	0.0042 0.0000 0.0075 0.0117 --	901.4 536.1 436.4 279.0 --
Sensitivity of CUS--50% = .5	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Sensitivity of CUS--50% = .95	CUS 50% MRA 70% Aspirin	150.7 147.7 143.4	3.0 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	451.0 257.8 --
Sensitivity of MRA--50% = .5	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Sensitivity of MRA--50% = .1	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Joint sensitivity of CUS/MRA--50% = .75	MRA/Angio 50% CUS/Angio 50% MRA 70% Aspirin	152.3 150.0 147.7 143.4	2.3 2.3 4.3 --	3.7200 3.7200 3.7175 3.7008	0.0000 0.0025 0.0167 --	10,940.9 883.0 257.8 --
Joint sensitivity of CUS/MRA--50% = 1	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7250 3.7175 3.7008	0.0075 0.0167 --	639.1 257.8 --
Specificity of CUS--50% = .7	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Specificity of CUS--50% = 1	Joint/Angio 50% Aspirin	148.8 143.4	5.4 --	3.7258 3.7008	0.0250 --	217.2 --
Specificity of MRA--50% = .4	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Specificity of MRA--50% = 1	MRA 50% Aspirin	150.5 143.4	7.1 --	3.7292 3.7008	0.0284 --	252.7 --
Joint specificity of CUS/MRA--50% = .75	MRA/Angio 50% CUS/Angio 50% MRA 70% Aspirin	152.3 150.0 147.7 143.4	2.3 2.3 4.3 --	3.7200 3.7200 3.7175 3.7008	0.0000 0.0025 0.0167 --	10,940.9 883.0 257.8 --
Joint specificity of CUS/MRA--50% = 1	Joint/Angio 50% MRA 70% Aspirin	151.8 147.7 143.4	4.1 4.3 --	3.7267 3.7175 3.7008	0.0092 0.0167 --	453.2 257.8 --
Sensitivity of CUS--70% = .5	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Sensitivity of CUS--70% = .9	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Sensitivity of MRA--70% = .75	Joint/Angio 50% CUS/Angio 50% CUS 50% MRA 70% CUS 70% Aspirin	152.3 150.0 149.8 147.2 146.6 143.4	2.3 0.2 2.6 0.6 3.2 --	3.7242 3.7200 3.7200 3.7142 3.7125 3.7008	0.0042 0.0000 0.0058 0.0017 0.0117 --	579.4 536.1 465.8 343.7 279.0 --
Sensitivity of MRA--70% = 1	Joint/Angio 50% MRA 70% Aspirin	152.3 147.9 143.4	4.4 4.5 --	3.7242 3.7192 3.7008	0.0050 0.0184 --	881.3 247.6 --
Joint sensitivity of CUS/MRA--70% = .75	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Joint sensitivity of CUS/MRA--70% = 1	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Specificity of CUS--70% = .65	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Specificity of CUS--70% = 1	Joint/Angio 50% CUS 70% Aspirin	152.3 145.7 143.4	6.6 2.3 --	3.7242 3.7158 3.7008	0.0084 0.0150 --	793.6 156.1 --
Specificity of MRA--70% = .5	Joint/Angio 50% CUS/Angio 50% CUS 50% CUS/Angio 70% Aspirin	152.3 150.0 149.8 146.6 143.4	2.3 0.2 3.2 3.2 --	3.7242 3.7200 3.7200 3.7125 3.7008	0.0042 0.0000 0.0075 0.0117 --	579.3 536.1 436.4 279.0 --
Specificity of MRA--70% = 1	Joint/Angio 50% MRA 70% Aspirin	152.3 147.2 143.4	5.1 3.8 --	3.7242 3.7192 3.7008	0.0050 0.0184 --	1,029.7 209.7 --
Joint specificity of CUS/MRA--70% = .75	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Joint specificity of CUS/MRA--70% = 1	Joint/Angio 50% MRA 70% Aspirin	152.3 147.7 143.4	4.6 4.3 --	3.7242 3.7175 3.7008	0.0067 0.0167 --	703.7 257.8 --
Prevalence of initial stroke severity TIA--50%; minor stroke--50%	CUS/Angio 70% CUS 50% MRA 70% Aspirin	84.0 81.0 79.0 75.0	3.0 2 4.0 --	5.6115 5.607 5.6023 5.5772	0.0045 0.0047 0.0251 --	535.1 402.2 150.4 --
Prevalence of stenosis 65%--0-49% (55% at baseline) 20%--50%-69%   5%--70%-99% (15% at baseline) 10%--Occlusion	Aspirin CUS/Angio 70%	148.0 147.0	1.0 --	3.8558 3.7408	0.1150 --	2.0 --
20%--0-49%20%--50%-69%50%--70%-99%10%--Occlusion	MRA 70% CUS/Angio 70% Aspirin	142.0 141.0 129.0	1.0 12.0 --	3.5275 3.5217 3.1600	0.0058 0.3617 --	196.3 34.7 --
Zero perioperative (30-day) stroke and death rate	MRA 50% Joint/Angio 50% CUS 50% Aspirin	154.0 153.0 150.0 143.0	1.0 3.0 7.0 --	3.7656 3.7626 3.7561 3.7010	0.0030 0.0065 0.0551 --	445.3 387.6 123.0 --
Years of risk reduction after CEA = 2 (4 at baseline)	Joint/Angio 50% MRA/Angio 70% MRA 70% Aspirin	152.0 148.0 148.0 143.0	4.0 0.0 5.0 --	3.7174 3.7156 3.7154 3.7008	0.0018 0.0002 0.0146 --	2,457.3 1,823.3 294.6 --
Years of risk reduction after CEA = 10	Joint/Angio 50% CUS 50% MRA 70% Aspirin	152.0 150.0 148.0 143.0	2.0 2.0 5.0 --	3.7362 3.7298 3.7208 3.7010	0.0064 0.0090 0.0198 --	390.5 238.0 216.5 --

Influence of Demographics

Figure 10. Effect of Baseline Life Expectancy on Cost-Effectiveness (more...)

   Figure 10. Effect of Baseline Life Expectancy on Cost-Effectiveness of MRA with Direct Referral of Patients with Severe (70-99 percent) Stenosis for CEA (MRA-70 percent)

As with the echocardiography model, we used data from the U.S. Life Tables produced by the National Center for Health Statistics to adjust the baseline all-causes mortality rate in the model, which is for 65-year-old white males. We used the Life Tables to adjust baseline mortality according to various combinations of gender, race (specifically, African American), and age (55, 65, 75, and 85 years old). Figure 10 illustrates the effect of increasing baseline mortality (measured by declining average life expectancy) on the cost per QALY for the strategy of testing with MRA and directly referring to CEA those patients with severe stenosis. Readers can apply this information to the specific circumstances of their patients.

Sensitivity Analysis

We performed sensitivity analyses on the model parameters listed in Table 17. Table 19 presents selected results. Varying the cost of testing did not substantively affect the results, except in the case where MRA was assumed to cost $2,500. In this analysis, the strategy of initial CUS with angiographic confirmation of severe stenosis became undominated, with a cost-effectiveness ratio of $280,000 per QALY. Varying the accuracy of the different testing strategies over wide ranges did not have a substantial overall effect on the results. When the perioperative complication rate was assumed to be zero, noninvasive strategies involving direct referral to CEA of patients with moderate or greater stenosis expectedly became the most cost-effective; without risk of complications, angiographic confirmation to avoid false positives was no longer beneficial, and the marginal benefit of CEA among patients with moderate stenosis was no longer counterbalanced by perioperative risk. Varying the duration of risk reduction associated with CEA between 2 and 10 years also did not substantively affect the cost-effectiveness ratios. Likewise, restricting the cohort to only patients with TIA or minor stroke, which reflects the patient populations in the two large carotid endarterectomy trials, did not have a major impact on cost-effectiveness ratios, though it did produce a different set of undominated strategies.

Figure 11. Cost effectiveness of 1) CUS with Angiographic Confirmation (more...)

   Figure 11. Cost effectiveness of 1) CUS with Angiographic Confirmation for Severe (70-99 percent) Stenosis (CUS/Angio-70 percent), and 2) MRA with Direct Referral to CEA for Severe Stenosis (MRA-70 percent), as a Function of the Prevalence of Severe Stenosis

The variable with the greatest influence on the results of the carotid imaging model was the prevalence of severe carotid stenosis. At severe stenosis prevalences of 0.15 and below, all testing strategies were dominated by the strategy of no testing or had cost-effectiveness ratios exceeding $250,000 per QALY. However, as this prevalence increased above 0.15, the cost-effectiveness ratios of two strategies -- CUS with angiographic confirmation of severe stenosis (CUS/Angio-70), and MRA with direct CEA referral for severe stenosis (MRA-70) -- fell precipitously, such that at a prevalence of 0.20, these strategies had cost-effectiveness ratios in the range of $60,000 to $75,000 per QALY (Figure 11). At higher prevalences, these ratios fell further. When compared to the strategy of no testing, CUS/Angio-70 had an incremental cost-effectiveness ratio of less than $50,000 per QALY at a prevalence of 0.25, while the incremental cost-effectiveness of MRA-70 fell below $50,000 per QALY as the prevalence of severe stenosis approached 0.30. These results suggest that carotid imaging may compare unfavorably, in terms of cost-effectiveness, with other commonly endorsed health care interventions, when the prevalence of carotid stenosis is low. It may be most efficient, therefore, to refer for carotid imaging those with a high pretest probability of severe stenosis, e.g. patients with peripheral vascular disease or audible carotid bruits.

Research Team

• Richard T. Meenan, PhD, MPH, Principal Investigator

• InvestigatorCenter for Health ResearchKaiser Permanente NorthwestPortland, Oregon

• Mark Helfand, MD, MPH

• Director, Evidence-based Practice CenterAssociate Professor of Internal Medicine and Medical Informatics & Outcomes ResearchOregon Health & Science UniversityPortland, Oregon

• Mark C. Hornbrook, PhD

• Associate Director for Health Services ResearchCenter for Health ResearchKaiser Permanente NorthwestPortland, Oregon

• Somnath Saha, MD, MPH

• Assistant Professor of MedicineOregon Health & Science UniversityPortland, Oregon

• Roger Chou, MD

• Assistant Professor of MedicineOregon Health & Science UniversityPortland, Oregon

• Kari Swarztrauber, MD, MPH

• Assistant Professor of NeurologyOregon Health & Science UniversityPortland, Oregon

• Kathryn Pyle Krages, AMLS, MA

• Administrator, OHSU Evidence-based Practice CenterDivision of Medical Informatics & Outcomes ResearchOregon Health & Science UniversityPortland, Oregon

• Benjamin K.S. Chan, MS

• Research AssociateDivision of Medical Informatics & Outcomes ResearchOregon Health & Science UniversityPortland, Oregon

• Maureen O'Keeffe-Rosetti, MS

• Research AssociateCenter for Health ResearchKaiser Permanente NorthwestPortland, Oregon

• Marian McDonagh, PharmD

• Clinical Research PharmacistCenter for Health ResearchKaiser Permanente NorthwestPortland, Oregon

• Daphne Plaut, MLS

• Research Center LibrarianCenter for Health ResearchKaiser Permanente NorthwestPortland, OR

• Martha Swain

• Senior Technical Writer-EditorCenter for Health ResearchKaiser Permanente NorthwestPortland, Oregon

• Susan Wingenfeld

• Administrative AssistantDivision of Medical Informatics & Outcomes ResearchOregon Health & Science UniversityPortland, Oregon

• Carole Reule

• SecretaryCenter for Health ResearchKaiser Permanente NorthwestPortland, Oregon

Technical Expert Advisory Group

• American Academy of Neurology Representative:

• Philip B. Gorelick, MD, MPH, FACPProfessor of Neurological SciencesRush Medical CollegeChicago, Illinois

• Neurologist/Stroke Center Director:

• Bruce M. Coull, MDProfessor of NeurologyUniversity of Arizona College of MedicineTucson, Arizona

• Cardiologist:

• George D. Giraud, MD, PhDPortland VA Medical CenterPortland, Oregon

• Vascular Surgeon:

• James Edwards, MDPortland VA Medical CenterPortland, Oregon

• HMO Perspective:

• Bruce W. Goldberg, MDMedical Director, CareOregonDepartment of Family MedicineOregon Health & Science UniversityPortland, Oregon

• Patient/Community Perspective:

• Alberta M. SimmonsPortland, Oregon

Peer Reviewers

• Robert J. Adams, MD

• Presidential Distinguished Chair and Regents' Professor of NeurologyMedical College of GeorgiaAugusta, Georgia

• William Armstrong, MD, FACC

• Representing the American College of CardiologyProfessor of Internal MedicineDirector, Echocardiology LaboratoryUniversity of Michigan HospitalAnn Arbor, Michigan

• Kenneth Brummel-Smith, MD

• Representing the American Geriatric SocietyProvidence Center on AgingPortland, Oregon

• Chris Gibbons, MD, MPH

• Senior Policy FellowOffice of the AdministratorCenters for Medicare and Medicaid ServicesBaltimore, Maryland

• Larry B. Goldstein, MD

• Duke Center for Cerebrovascular DiseaseDuke University Medical CenterDurham, North Carolina

• Phillip B. Gorelick, MD, MPH, FACP

• Representing the American Academy of NeurologyProfessor of Neurologic SciencesRush Medical CollegeChicago, Illinois

• Edward Grant, MD

• Representing the American College of RadiologyProfessor of Radiological SciencesChief of UltrasoundUniversity of California, Los Angeles School of MedicineLos Angeles, California

• Robert G. Hart, MD

• Professor of Medicine and NeurologyUniversity of Texas Health Science CenterSan Antonio, Texas

• Arthur J. Labovitz, MD

• Division of CardiologySt. Louis University School of MedicineSt. Louis, Missouri

• David B. Matchar, MD

• Director, Duke Evidence-based Practice CenterDirector, Duke Center for Clinical Health Policy ResearchDuke University Medical CenterDurham, North Carolina

• Gregory P. Samsa, PhD

• Associate Director, Duke Center for Clinical Health Policy ResearchDuke University Medical CenterDurham, North Carolina

• William B. Schroedter, BS, RVT, FSVT

• Representing the Society of Vascular TechnologyTechnical DirectorQuality Vascular Imaging Inc.Venice, Florida

Key to Symbols and Abbreviations

exp = exploded subject heading or MeSH term

.fs = floating subheading

.gw = group name

.kw = keyword

.me = subject heading or MeSH term

.mp = word in title, abstract, registry number word, or subject heading

.pt = publication type

.ti = title word

/ (after a word) = subject heading or MeSH term

$ (as a suffix) = truncation

* (as a suffix) = truncation

* (as a prefix) = major focus subject heading or MeSH term

Note: When possible, MEDLINE citations were excluded when searching HealthSTAR and the Cochrane Controlled Trials Register.

Carotid Endarterectomy Complications

Carotid Endarterectomy

Carotid Endarterectomy

Carotid Endarterectomy Economics

Carotid Endarterectomy Economics

Cerebral Angiography Complications