Article, see p 946

Consider the case of C.S., a 69-year-old man who has a moderately severe 50% to 69% left internal carotid stenosis, which has been followed for years by his physician with serial duplex scans at accredited Vascular Laboratory A. When that physician retired, C.S. transitioned his care to another physician, who ordered his annual carotid duplex scan to be performed at accredited Vascular Laboratory B. C.S. was alarmed when the report released to his electronic medical record stated that he now had a >70% left internal carotid stenosis. The peak systolic velocity in the left internal carotid artery is 210 cm/s on both examinations.

Carotid duplex ultrasound was developed in the 1970s at the University of Washington by a team led by Dr D. Eugene Strandness Jr,¹ and although catheter-based arteriography remains the gold standard imaging modality for carotid artery disease, major societal guidelines now recommend carotid duplex ultrasound as the initial diagnostic imaging modality to evaluate the severity of carotid stenosis.^2,3 Furthermore, if carotid duplex ultrasound is unequivocal in the identification of 50% to 99% stenosis in symptomatic patients and 70% to 99% stenosis in asymptomatic patients, carotid duplex findings are sufficient to make decisions regarding further management, including intervention.² Because carotid duplex ultrasound is widely available and noninvasive, it is the most common imaging examination performed worldwide to diagnose carotid disease.⁴ Consequently, the criteria used for classifying stenosis severity by carotid duplex ultrasound must be accurate and reproducible.

Ultrasound accrediting organizations such as the Intersocietal Accreditation Commission (IAC) and the American College of Radiology attempt to standardize the performance of vascular ultrasound examinations, but they do not require the use of a single set of carotid duplex criteria for the classification of carotid stenosis. Instead, they stipulate that each accredited laboratory have a set of interpretation criteria that are used by all members of the technical and medical staff and are either derived from the literature or developed and validated internally.^5,6 Therefore, it is not surprising that among accredited vascular laboratories, there is wide variability in carotid interpretation criteria.⁷ Even with the publication of proposed standardized velocity criteria for carotid duplex ultrasound by a multispecialty panel in 2003,⁴ only 24% of IAC-accredited vascular laboratories were using these standardized criteria in 2011, and there were 17 sets of diagnostic criteria in use among 117 laboratories, with the remaining facilities using locally developed (6 laboratories) or unreferenced or hybrid criteria (29 laboratories).⁷ Whereas there is clearly wide variability in the diagnostic criteria used in accredited facilities for carotid duplex ultrasound, the clinical implications of this variability are unclear.

The current study by Columbo et al⁸ not only describes the variation in velocity thresholds used for determining carotid stenosis severity in accredited laboratories but also estimates the potential clinical impact of this variability. Using data from a random sample of 338 vascular testing centers accredited by the IAC, diagnostic velocity thresholds were applied to peak systolic velocity values obtained from 2 populations: a population-based cohort (4791 patients ≥65 years of age) representative of patients who are typically tested because of risk factors for carotid disease (the Cardiovascular Health Study⁹) and a population-based cohort (28 483 patients) who underwent surgery for asymptomatic carotid stenosis (the Vascular Quality Initiative registry [www.vqi.org]). Diagnostic thresholds for ≥50% stenosis were applied to the Cardiovascular Health Study group, given that this is the stenosis threshold at which patients would usually begin long-term surveillance; and diagnostic thresholds for ≥70% stenosis were applied to the Vascular Quality Initiative registry group, given that this is the stenosis threshold at which carotid revascularization may be considered in the asymptomatic patient.

The results of this investigation on the diagnostic velocity thresholds for carotid duplex ultrasound are disturbing. The authors found that the 2003 Society of Radiologists in Ultrasound Consensus Conference⁴ criteria were being applied by 46% of facilities, an increase from the 24% reported in 2011.⁷ However, among the 338 IAC-accredited vascular laboratories, 60 discrete stenosis category/peak systolic velocity pairs were in use.⁸ When applied to the Cardiovascular Health Study group, the 5th percentile peak systolic velocity threshold for ≥50% stenosis (125 cm/s) would assign a diagnosis of “moderate carotid stenosis” twice as often as the 95th percentile peak systolic velocity threshold (150 cm/s). In the Vascular Quality Initiative registry group, 9.8% of patients who underwent a carotid endarterectomy for asymptomatic stenosis had a peak systolic velocity falling between the 5th and 95th percentile peak systolic velocity thresholds for ≥70% stenosis (230 cm/s and 275 cm/s, respectively). This implies that 1 in 10 patients may not have been considered for carotid revascularization had they received their carotid duplex ultrasound in an accredited vascular laboratory with different interpretation criteria. These findings are similar to those of a study published in 2014 in which diagnostic velocity thresholds from 10 institutions in New England were applied to a series of 15 534 carotid duplex scans performed at the University of Massachusetts.¹⁰ The differences in diagnostic criteria used to interpret these scans resulted in significant variation in the classification of carotid artery stenosis and, by implication, significant variation in the potential number of subsequent carotid interventions.

A number of factors contribute to variability in the results of carotid duplex ultrasound examinations, including scanning protocol, sonographer skill, and instrumentation used. Accreditation seeks to minimize this variability by requiring participating vascular laboratories to comply with detailed standards for ultrasound imaging, qualifications of sonographer and physician personnel, selection and maintenance of equipment, quality assurance, and reporting.^5,6 Assuming that the accredited laboratories included in this study were in good standing and adhered to the standards outlined by the IAC, the findings of this study are even more impressive, as the only major factor left to explain the variability in carotid stenosis classification is the interpretation criteria or diagnostic velocity thresholds used. With 60 distinct stenosis category/peak systolic velocity pairs in use in the 388 accredited vascular laboratories, it is not surprising that our patient, C.S., had a high probability of receiving disparate reports on the severity of his carotid stenosis, even when the peak systolic velocity remained the same on serial tests. A study published in 2014 found that of 7327 outpatient facilities billing Medicare for cerebrovascular testing in a 5% random Outpatient Limited Data Set for the United States in 2011, only 22% were IAC-accredited, with significant variation in rates of accreditation by region.¹¹ Data are not available for the prevalence of accreditation by all accrediting bodies, but it should be noted that accreditation is not a requirement for reimbursement of carotid duplex ultrasound examinations by the Centers for Medicare and Medicaid Services.¹²

In the decades since it was introduced in the 1970s, duplex scanning has become an integral part of the management of patients with carotid disease, from initial screening to follow-up after intervention. Although the instrumentation has improved significantly and scanning protocols have been refined, considerable variability remains in the velocity thresholds used to classify the severity of internal carotid artery stenosis. Standardization of the criteria used for interpreting carotid duplex ultrasound examinations would avoid the distressing predicament of our patient, C.S., and reduce the variability identified in the results of examinations performed in different vascular laboratories. However, it is important to point out that standardization of diagnostic velocity thresholds would not improve the overall accuracy of carotid duplex scanning. Although there is a definite correlation between the velocities obtained by duplex ultrasound and percent stenosis on arteriography, this relationship is highly variable.¹³ Therefore, it is unlikely that further refinements in the diagnostic velocity thresholds will lead to improved accuracy compared to arteriography. The principal rationale for standardization of velocity thresholds is to achieve consistency in the interpretation and reporting of carotid duplex scans.

The study by Columbo et al⁸ illustrates how the wide variability in carotid velocity thresholds can lead to differences in clinical care and contributes to the growing body of work supporting standardization. An initiative is currently being undertaken by IAC Vascular Testing, which is designed to validate a specific set of carotid velocity criteria that can be recommended for use by all IAC-accredited facilities.^14,15 The endorsement of IAC Vascular Testing and its sponsoring organizations would be a major step toward standardized velocity criteria and improved consistency for carotid duplex ultrasound. This would avoid confusion and improve vascular laboratory practice for sonographers, interpreting physicians, referring providers, and our patients.

Dr Kim is a member of the Board of Directors of the Intersocietal Accreditation Commission: Vascular Testing. Dr Zierler is a former member of the Board of Directors and past President of the Intersocietal Commission for the Accreditation of Vascular Laboratories.

The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.

https://www.ahajournals.org/journal/circ

文章，请参见第946页

以CS的情况为例，这名69岁的男子患有中度严重的50％至69％的左颈内动脉狭窄，此后多年，他的医生在经认可的血管实验室A中进行了连续双相扫描。退休后，CS将他的护理转到了另一位医生那里，后者命令在经认可的血管实验室进行他的年度颈动脉双侧扫描。当向他的电子病历发布的报告称他现在有70％以上的颈内动脉狭窄时，BCS感到震惊。在两次检查中，左颈内动脉的最高收缩速度均为210 cm / s。

颈动脉双工超声是由D. Eugene Strandness Jr，¹博士领导的团队于1970年代在华盛顿大学开发的，尽管基于导管的动脉造影术仍然是颈动脉疾病的金标准成像方式，但主要的社会指导原则现在推荐颈动脉双工超声作为诊断颈动脉狭窄严重程度的初步诊断影像学手段。^2,3此外，如果在有症状的患者中50％至99％的狭窄和在无症状的患者中70％至99％的狭窄的识别中，颈动脉双工超声检查是明确的，则颈动脉双工检查结果足以就进一步治疗（包括干预）做出决策。²因为颈动脉双工超声是广泛可用的并且是非侵入性的，所以它是全世界诊断颈动脉疾病最常见的影像学检查。⁴因此，用于通过颈动脉双工超声对狭窄严重程度进行分类的标准必须准确且可重复。

诸如社会间认可委员会（IAC）和美国放射学院的超声认可组织试图对血管超声检查的性能进行标准化，但它们并不需要使用一套单独的颈动脉双工标准来对颈动脉狭窄进行分类。取而代之的是，他们规定，每个获得认可的实验室都有一套解释标准，技术人员和医务人员的所有成员都可以使用这些解释标准，这些解释标准要么来自文献，要么在内部进行开发和验证。^5,6因此，在经过认可的血管实验室中，颈动脉解释标准存在很大差异也就不足为奇了。⁷即使在2003年由多专业小组公布了提议的颈动脉双工超声标准化速度标准，2011年，只有⁴个经过IAC认证的血管实验室使用了这些标准化标准，在117个诊断标准中有17套正在使用实验室，其余设施使用本地开发的实验室（6个实验室）或未引用或混合的标准（29个实验室）。⁷尽管在认可的颈动脉双工超声检查设施中使用的诊断标准存在明显的差异，但这种差异的临床意义尚不清楚。

Columbo等[ ^8]的当前研究不仅描述了在认可实验室中用于确定颈动脉狭窄严重程度的速度阈值的变化，而且还估计了这种变化的潜在临床影响。使用IAC认可的338个血管检测中心的随机样本数据，将诊断速度阈值应用于从两个人群中获得的峰值收缩速度值：基于人群的队列（4791例年龄≥65岁的患者）代表了以下人群：通常由于颈动脉疾病的危险因素而进行测试（心血管健康研究⁹）和基于人群的队列研究（28 483例患者），他们接受了无症状颈动脉狭窄手术（“血管质量计划”注册表[www.vqi.org]）。≥50％狭窄的诊断阈值适用于心血管健康研究组，因为这是通常开始长期监测的狭窄阈值；血管质量计划注册小组采用了≥70％狭窄的诊断阈值，因为这是无症状患者可以考虑进行颈动脉血运重建的狭窄阈值。

这项关于颈动脉双工超声诊断速度阈值的研究结果令人不安。作者发现，在46％的设施中采用了2003年放射医师学会在超声波共识会议上提出的^4个标准，较2011^年报告的24％有所增加^。7但是，在338个获得IAC认可的血管实验室中，有60个离散性狭窄类别使用/峰值收缩速度对。⁸当应用于“心血管健康研究”组时，≥50％狭窄（125 cm / s）的第5个百分位收缩峰值速度阈值将被诊断为“中度颈动脉狭窄”的频率是第95个百分位收缩率峰值速度阈值（150）的两倍厘米/秒）。在“血管质量计划”注册小组中，接受无症状性狭窄的颈动脉内膜切除术的患者中，有9.8％的收缩期峰值速度在狭窄度≥70％的收缩期峰值速度阈值的第5个百分点至第95个百分点之间（230 cm / s和275 cm / s s）。这意味着，如果每10名患者中有1名患者在具有不同解释标准的认可血管实验室中接受了颈动脉双工超声检查，则可能没有考虑过进行颈动脉血运重建。¹⁰用于解释这些扫描的诊断标准的差异导致颈动脉狭窄的分类发生显着变化，并暗示随后进行的颈动脉干预的潜在数量也发生显着变化。

许多因素导致颈动脉双工超声检查结果的差异，包括扫描协议，超声检查者的技能和所使用的仪器。认证试图通过要求参与的血管实验室遵守超声成像的详细标准，超声医师和医师人员的资格，设备的选择和维护，质量保证以及报告来最大程度地减少这种差异。^5,6假设本研究中包括的认可实验室信誉良好并遵守IAC概述的标准，则本研究的结果将更加令人印象深刻，因为解释颈动脉狭窄分类变异性的唯一主要因素是标准或使用的诊断速度阈值。在388个获得认可的血管实验室中使用了60种不同的狭窄类别/收缩期峰值速度对，即使患者的收缩期峰值达到峰值，我们的患者CS也很可能收到有关颈动脉狭窄严重程度的不同报告，这并不奇怪在串行测试中，速度保持不变。¹¹尚无所有认可机构普遍认可的数据，但应注意的是，医疗保险和医疗补助服务中心并不一定要报销颈动脉双工超声检查。¹²

自从1970年代引入以来的几十年中，从最初的筛查到干预后的随访，双面扫描已成为治疗颈动脉疾病患者不可或缺的一部分。尽管仪器已得到显着改善，并且扫描方案也得到了改进，但是用于对颈内动脉狭窄程度进行分类的速度阈值仍存在较大差异。用于解释颈动脉双工超声检查的标准的标准化将避免我们的患者CS的困扰，并减少在不同血管实验室进行的检查结果中确定的变异性。但是，必须指出的是，诊断速度阈值的标准化不会提高颈动脉双工扫描的整体准确性。¹³因此，与动脉造影术相比，诊断速度阈值的进一步完善不太可能导致准确性提高。速度阈值标准化的基本原理是要在颈动脉双工扫描的解释和报告中实现一致性。

Columbo等人[ ^8]的研究表明，颈动脉速度阈值的巨大差异如何导致临床护理方面的差异，并有助于支持标准化的工作量不断增长。IAC血管测试公司目前正在采取一项举措，旨在验证一组特定的颈动脉速度标准，这些标准可建议所有IAC认可的设施使用。^14,15对IAC血管测试及其赞助组织的认可将是朝着标准化速度标准和改善颈动脉双工超声的一致性迈出的重要一步。这将避免混淆，并改善超声检查医师，口译医生，转诊医生和我们患者的血管实验室操作。

金博士是社会认证委员会：血管检测委员会的成员。Zierler博士是血管实验室认证社会间委员会的前董事会成员和前任主席。

本文表达的观点不一定是编辑者或美国心脏协会的观点。

https://www.ahajournals.org/journal/circ