Non-invasive imaging to detect vertebral artery stenosis

  • Protocol
  • Diagnostic

Authors


Abstract

This is the protocol for a review and there is no abstract. The objectives are as follows:

To estimate and compare the diagnostic accuracy of DUS, TCD/TCCD, MRA and CTA for detecting relevant vertebral artery stenosis in symptomatic patients with (posterior circulation) ischaemic stroke or TIA as well as in asymptomatic patients. A patient is considered asymptomatic in this review when he or she did not have any clinical signs or symptoms related to stroke or TIA of the posterior circulation in the last six to 12 months. This includes patients that were in fact symptomatic of anterior circulation pathology and were investigated by a combination of the aforementioned image modalities (see Participants below).

As the accuracy of imaging may depend on the location of the stenosis in the vertebral artery we are planning to analyse the data according to the location of the stenosis. The vertebral artery can be divided into four anatomical parts: V1 to V3 form the extracranial vertebral artery and V4 forms the intracranial vertebral artery (Khan 2007, Figure 2).

Background

Target condition being diagnosed

About one third of all ischaemic strokes occurring in the posterior circulation of the brain (the territory supplied by the vertebral arteries) can be attributed to thromboembolism from atherosclerotic stenosis in one of the vertebral arteries. In people with posterior circulation events, the presence of a stenosis ≥ 50% in the vertebrobasilar system is associated with a higher risk of early recurrent stroke (Gulli 2009; Marquardt 2009). Although the stenosis is measured on a continuous scale, the threshold value of 50% was chosen. Accurate grading of vertebral artery stenosis is essential because of the new endovascular treatment options that are emerging.

Studies of sufficient quality measuring the accuracy of diagnosing and grading vertebral artery stenosis with noninvasive imaging techniques against the gold standard of intra-arterial digital subtraction angiography (DSA) are scarce. In a systematic review from 2007, computed tomographic angiography (CTA) and contrast enhanced magnetic resonance angiography (MRA) were most sensitive in diagnosing vertebral artery stenosis compared with duplex ultrasound (DUS) (Khan 2007). The sensitivity and specificity of CTA were 100% (95% confidence interval (CI) 15.8 to 100) and 95.2% (95% CI 83.8 to 99.4), respectively. The pooled sensitivity and specificity of contrast enhanced magnetic resonance imaging (MRI) were 93.9% (95% CI 79.8 to 99.3) and 94.8% (95% CI 91.1 to 97.3). In Khan 2007 the number of high quality studies was limited. Since 2007, several new studies of imaging of vertebral artery stenosis have been performed.

Visualisation of the vertebral artery, and especially the vertebral artery origin, is complicated by the deep, tortuous course, breathing movements and pulsations from nearby vessels. Current treatment options for vertebral artery stenosis are medical treatment, surgical revascularisation and percutaneous transluminal angioplasty with or without stenting. A Cochrane review of percutaneous transluminal angioplasty and stenting for vertebral artery stenosis concluded that the role of the different procedures in preventing stroke from vertebral artery stenosis has not been fully established (Coward 2005). This diagnostic test accuracy review is relevant, since little is known about the accuracy of the non-invasive imaging modalities for imaging of the vertebral arteries.

The main objective of this review will be to meta-analyse the diagnostic accuracy of non-invasive imaging compared with the reference standard, and to investigate potential sources of variation. A clinical correlation is not needed for this analysis, since stenosis in a vertebral artery of a patient who has had symptoms originating from pathology from the posterior circulation has the same characteristics as a stenosis in a vertebral artery of a patient who did not have symptoms.

Index test(s)

  • DUS is used for assessment of the extracranial vertebral artery and can be performed with or without colour coding. Peak systolic velocity (PSV) is the most commonly used parameter for the assessment of vertebral artery stenosis, although the exact relation between PSV and the grade of stenosis has not been fully elucidated. DUS is quick and inexpensive, but very operator-dependent. Therefore, intensive training and quality control of the technicians and doctors performing the test are important factors to consider when interpreting test results.

  • Transcranial Doppler (TCD) and transcranial colour-coded duplex sonography (TCCD) are ultrasound neuroimaging modalities that can provide useful diagnostic information on the intracranial vertebral artery with a high degree of safety. Similar to DUS this technique is non-invasive and the diagnostic value is known to be operator-dependent. In addition, the use of TCD and TCCD is limited to the evaluation of the intracranial segments of the vertebral artery. If a severe extracranial stenosis is present, the pattern of intracranial vertebral flow may confirm this stenosis. However, the operator would not be able to localise the extracranial stenosis on TCD or TCCD alone.

  • CTA allows visualisation of the total vertebral artery. CTA is relatively cheap and widely available, but is contraindicated in patients with renal failure. In this review we consider imaging techniques in which contrast fluid is administered to the patient through an intravenous catheter as 'minimally invasive'.

  • MRA allows visualisation of the total vertebral artery similar to CTA and can be performed with or without contrast enhancement. MRI scans often prove difficult for patients suffering from claustrophobia. If a patient is unable to lie still the imaging will most likely suffer from movement artefacts that will compromise the quality and reliability of the test. MRA is contraindicated in patients with pacemakers and metallic implants. Furthermore, it is expensive.

Clinical pathway

In patients with vertebrobasilar ischaemia, intra-arterial DSA is the gold standard for diagnosing vertebral artery stenosis. The current diagnostic pathway preceding DSA varies, depending on the preferences of the local neurologist. Either DUS, TCD, (contrast enhanced) MRA, CTA or a combination of these imaging modalities is used. DSA has a small but non-negligible risk of morbidity and mortality (Hankey 1990). Consequently, if DSA is replaced by a noninvasive or minimally invasive imaging modality, the potential overall benefit of a possible treatment would increase. In the best diagnostic pathway DSA would be replaced by a single non-invasive imaging modality or the combination of one or more imaging modalities. Alternatively, one or more of the non-invasive or minimally invasive imaging modalities might be suitable for screening for vertebral artery stenosis. Another advantage of replacing DSA by (a combination of) index tests is that DSA is mostly unavailable in non-tertiary centres, whereas DUS, CT and MRI are widely available in most hospitals.

Currently, there is no standard treatment for patients with a symptomatic vertebral artery stenosis. However, there are different treatment options as reported above. If the result of the index test is false negative, the patient will not be eligible for surgical or interventional treatment. This in turn could lead to a new transient ischaemic attack (TIA) or stroke. If the result of the index test is false positive, surgical or endovascular treatment might be considered. This is expensive and invasive. However, missing a potentially treatable stenosis (i.e. false negative result) would be worse.

Rationale

Intra-arterial DSA is the gold standard test to estimate stenosis in the vertebral artery. However, in line with diagnosing carotid artery stenosis, the question is whether it remains ethical to use DSA for diagnostic purposes only, due to the small but non-negligible risk of morbidity and even mortality (Hankey 1990). Therefore, non-invasive and minimally invasive tests such as DUS, TCD/TCCD, CTA, or MRA are increasingly used in the diagnosis of vertebral artery stenosis. However, unlike in the diagnosis of carotid artery stenosis (Debrey 2008; Koelemay 2004; Nederkoorn 2003; Wardlaw 2006) the pooled diagnostic accuracy is unknown. Anatomical differences between the carotid and vertebral artery (e.g. smaller size and deeper course of the vertebral artery, greater variation between the two sides of the neck) prevent extrapolation of results from diagnosing carotid artery stenosis to vertebral artery stenosis. In 2007 a systematic review was published on the same subject; since then several new studies have been published and treatment of vertebral artery stenosis has gained ground (Khan 2007).

Objectives

To estimate and compare the diagnostic accuracy of DUS, TCD/TCCD, MRA and CTA for detecting relevant vertebral artery stenosis in symptomatic patients with (posterior circulation) ischaemic stroke or TIA as well as in asymptomatic patients. A patient is considered asymptomatic in this review when he or she did not have any clinical signs or symptoms related to stroke or TIA of the posterior circulation in the last six to 12 months. This includes patients that were in fact symptomatic of anterior circulation pathology and were investigated by a combination of the aforementioned image modalities (see Participants below).

Secondary objectives

As the accuracy of imaging may depend on the location of the stenosis in the vertebral artery we are planning to analyse the data according to the location of the stenosis. The vertebral artery can be divided into four anatomical parts: V1 to V3 form the extracranial vertebral artery and V4 forms the intracranial vertebral artery (Khan 2007, Figure 2).

Methods

Criteria for considering studies for this review

Types of studies

We will include diagnostic studies on humans that meet the following criteria:

  • DUS, TCD/TCCD, CTA or MRA performed to estimate the severity of vertebral artery stenosis;

  • DSA used as the reference standard;

  • absolute numbers of true positives, false negatives, true negatives, and false positives are available or derivable from the presented data for at least one cut-off criterion for the degree of stenosis based on DSA.

We will include studies that evaluate a single test and studies that evaluate a combination of tests (i.e. a diagnostic strategy). For instance, we will include studies that evaluate a combination of tests as a single modality, such as DUS plus MRA. Another example of studies that evaluate a combination of tests are those which evaluate a sequence of multiple tests, such as DUS followed by CTA if DUS was inconclusive. We will include comparative studies as well, in which two or more index tests are done in the same artery versus the reference test.

We will include studies that determine the diagnostic accuracy from case series of consecutively enrolled patients, or using a case-control design. We will exclude case reports and studies in which less than five patients were included. We plan to include studies published in any language.

Participants

Studies and case reports including adult patients presenting with a clinical diagnosis of TIA or stroke who were evaluated for stenosis in the posterior circulation are eligible for this systematic review. A patient will be considered 'symptomatic' when a posterior circulation stroke or TIA occurred in the six to 12 months preceding the imaging of a vertebral artery stenosis. In clinical practice a posterior circulation stroke or TIA is diagnosed by the combination of clinical findings, imaging and follow-up. The presence of a stroke or TIA should be assessed by a neurologist. We will also include patients presenting with an anterior circulation stroke or TIA in which evaluation of the posterior circulation was done. Although these patients have had symptoms clinically located as originating from the anterior circulation, we consider these patients as being asymptomatic since we are defining 'symptomatic' on the clinical details of posterior stroke signs. Since a clinical correlation is not needed for the analysis, we are also interested in asymptomatic patients. A patient is considered asymptomatic when he or she did not suffer from disease originating from the vertebrobasilar circulation.

Index tests

  • DUS, with or without colour coding.

  • TCD and TCCD.

  • CTA.

  • MRA, with or without contrast enhancement.

For the DUS, we will use systolic velocity (PSV) or the peak velocity ratio (PVR, calculated as intrastenotic PSV divided by proximally recorded PSV). We will make use of the measurement the author of the original article used, both for ≥ 50% stenosis and ≥ 70% stenosis. In the final analysis we will pool together PSV-data and PVR-data.

For the TCD/TCCD we will use the mean flow velocities (MFV).

For CTA and MRA we will use a NASCET-like method: % stenosis = [1 - minimum residual lumen/normal distal cervical vertebral artery diameter] × 100 (Rothwell 2003).

Target conditions

Presence of relevant vertebral artery stenosis defined as ≥ 50% (50% to 99%) or ≥ 70% (70% to 99%).

Reference standards

The reference standard (and gold standard) for diagnosing vertebral artery stenosis will be intra-arterial DSA. Different methods can be applied to measure the degree of stenosis. The most applied method is the NASCET-like method using per cent diameter ratios (see Index tests for the formula).

Search methods for identification of studies

Electronic searches

To find the studies, we will search the following electronic databases:

  • Cochrane Central Register of Controlled Trials (CENTRAL);

  • Cochrane Register of Diagnostic Test Accuracy Studies (CRDTAS);

  • MEDLINE (Ovid);

  • EMBASE;

  • MEDION;

  • BIOSIS;

  • Health Technology Assessment Database (HTA);

  • Database of Abstracts of Reviews of Effects (DARE).

We have used both relevant free text and subject headings (MeSH terms) to develop the MEDLINE search strategy. We will adapt the MEDLINE strategy to search other relevant electronic databases. Details of the MEDLINE search strategy are presented in Appendix 1.

We will import all citations identified by the search strategy into Reference Manager.

Searching other resources

We will check the reference lists of eligible original and review publications on this subject. We will use previous relevant published literature reviews as a source of citations. We will conduct citation searches for included studies via the 'Related Articles' feature in PubMed, the Conference Proceedings Citations Index and the ISI citations index. We will consult experts on the subject to find additional (unpublished) studies. We will also handsearch relevant neurological and radiological journals from the date the first included study was published (Appendix 2).

Data collection and analysis

Selection of studies

One review author (AGK) will assess the titles and abstracts of the records retrieved from the searches and exclude irrelevant reports. We will attempt to identify reports that update previously published reports which use the same study population at different recruitment points. We will use the most complete data set that supersedes previous publications to avoid double counting participants or studies. Two authors will independently apply the selection criteria to the selected full-text reports. We will resolve any disagreements by discussion. We will list relevant excluded studies with the primary reason for exclusion. We will not be blinded to study authors, institution, and study results during the selection process.

Data extraction and management

Two review authors (AGK, AC) will extract data on study characteristics and measures of accuracy using a standardised data extraction form. We will record the following data:

  • journal name, year of publication, study design (cross-sectional survey, prospective study, retrospective study, case report);

  • setting, number and characteristics of participants, (mean age and range, sex, previous history of stroke), inclusion of consecutive patients, diagnosis of posterior circulation stroke or TIA in the last six or 12 months, duration of symptoms to test;

  • study level characteristics: index test parameters: method of measuring stenosis, cut-off values for the index tests (PSV value for threshold of ≥ 50% or ≥ 70% stenosis may vary when analysing DUS), location of the vertebral artery stenosis (V1 to V4); the index test(s) performed, imaging technique (DUS with or without colour, MRA with or without contrast enhancement), participants (symptomatic patients versus asymptomatic patients), scanning machine used (Tesla number for MRA; non-helical versus helical versus helical multi-slice scanner for CTA), manufacturer of medical imaging equipment, time interval from index test and reference standard, method of intra-arterial DSA (selective, aortic arch), method of defining degree of stenosis in reference test (NASCET (Rothwell 2003) or other calculation of lumen diameter); information related to the clinicians who read and interpret imaging results (background speciality, level of expertise, whether the imaging was read whilst blinded to clinical information and the reference standard);

  • for each index test we will put the absolute number of true positive cases, false positive cases, false negative cases and true negative cases into a 2x2 table.

Where investigators published several reports based on data from a single study population, we will select the updated or most complete report. If data are only available for a subset of patients, we will include this subset. When absolute numbers for true positives, false negatives, true negatives and false positives are not provided by the study author and cannot be derived from the available data, we will request the missing date from the study author. If insufficient data are provided we will exclude the study.

The unit of analysis will be each individual vertebral artery. We chose to use this unit of analysis because the primary objective of this study is to determine the diagnostic test accuracy in symptomatic and asymptomatic patients. This implies that we will look at both symptomatic and asymptomatic arteries. From a clinical point of view the unit of analysis should preferably be patients. If data are given on the sensitivity and specificity of an index test on the patient-level we will analyse these data in a separate analysis. We will resolve any disagreements by discussion.

Assessment of methodological quality

We will assess the quality of the diagnostic studies using the QUality Assessment of Diagnostic Accuracy Studies (QUADAS)-2 checklist (Whiting 2011). This checklist comprises four domains: patient selection, index test, reference standard, and flow and timing (Appendix 3). For all domains signalling questions are answered to assess the presence of bias and the applicability of the study. An 'appropriate spectrum' of patients is defined as: patients admitted to the hospital after a (posterior circulation) stroke or TIA for first evaluation of the carotid or vertebral arteries or both. An 'appropriate reference standard' is defined as the use of intra-arterial DSA and measurement of the degree of stenosis by applying the NASCET-criteria (Rothwell 2003). An 'acceptable delay' is defined as the use of intra-arterial DSA within one month after or before the index test. Relevant clinical data required for interpretation of the test results are: age, sex, and clinical diagnosis of TIA or stroke.

In addition to the QUADAS-2 checklist we will report the expertise of the clinician interpreting the imaging results (radiologist, neuroradiologist or neurologist). We will report whether the execution of the index test was described in sufficient detail to permit replication of the test and whether uninterpretable or intermediate test results were reported. We will resolve any disagreements by discussion. We will report the results of the quality assessment for each study separately in a table.

Statistical analysis and data synthesis

We will analyse data on a vessel-level. We will not mix studies that differ in the threshold to define the target condition (i.e. ≥ 50% or ≥ 70% stenosis) but will analyse these separately. We will draw forest plots of sensitivity and specificity with their corresponding 95% CIs for the data that were obtained from the different diagnostic tests, to show the degree of precision by which sensitivity and specificity have been measured in each study, and to get an impression of the amount of variability in estimates between studies. We will plot pairs of sensitivity and specificity data in receiver operating characteristic (ROC) space to visualise the variation between these values across studies and whether there is a pattern suggesting a negative correlation between them. We will use sensitivity to define the y-axis, specificity will define the x-axis. Each point on the plot will represent one investigated study. We will construct the summary ROC plots using Review Manager software (RevMan) (RevMan 2012).

We will construct linked ROC plots if studies comparing two or more of the index tests directly to the reference standard are done in the same population. We will do a separate analysis of these paired studies as it will be more robust methodologically.

If data from four or more studies are available for a particular index test, we will be using the bivariate random-effects approach to simultaneously model the sensitivities and specificities from each study. This bivariate model is a hierarchical model incorporating both the variation between studies (random-effects model for sensitivity and specificity) and the precision of sensitivity/specificity within each study. The bivariate model can be used to estimate the mean (logit) sensitivity and specificity, the corresponding 95% random-effects CIs, as well as providing estimates for the amount of between-study variation in sensitivity and specificity, and their correlation. The latter enables the estimation of confidence and prediction intervals. In addition, the bivariate model can be extended with covariates to examine their effect on sensitivity, specificity or both. Our main question - whether the accuracy differs between index tests - will be formally examined by including the data from different index tests into a single bivariate model and then adding indicator variables representing a specific type of index test. Such a model will allow us to examine whether there is a difference in sensitivity, in specificity, or both, for each threshold value ( ≥ 50% stenosis and ≥ 70% stenosis) with a covariate indicating the type of test. For fitting the bivariate models we plan to use the non-linear mixed-effects module (NLMIXED) of SAS.

Investigations of heterogeneity

We will analyse each of the study level characteristics (see Data extraction and management) as covariates in the bivariate model and use meta-regression analysis to conclude whether the accuracy is different between these characteristics. If there are only a few studies available and meta-regression analysis proves unfeasible, we will perform visual inspections of ROC and forest plots to investigate heterogeneity. Because the power for examining patient characteristics using summary indices on a study level (e.g. mean age of a study, percentage of males in a study) is often very low, we will not examine these patient characteristics separately.

Sensitivity analyses

We will undertake sensitivity analyses for each of the QUADAS-2 criteria described in four domains in Appendix 3 to determine the effect of poor study quality on the overall results.

Assessment of reporting bias

Given the uncertainty about the mechanisms behind and the best methods for assessing reporting bias, we will refrain from examining reporting bias within this review.

Appendices

Appendix 1. Search strategy

MEDLINE (Ovid) search strategy

1. Exp Vertebrobasilar Insufficiency/
2. Vertebral Artery/
3. Constriction, Pathologic/ or arterial occlusive diseases/ or arteriosclerosis/ or atherosclerosis/ or intracranial arteriosclerosis/
4. 2 and 3
5. (vertebral arter$ adj5 (insufficien$ or stenosis or stenoses or ischaemi$ or ischemi$ or atherosclero$ or arteriosclero$ or constrict$ or occlu$ or narrow$ or plaque$ or obstruct$)).tw.
6. ((vertebrobasilar or vertebro-basilar or vertebro basilar) adj5 (insufficien$ or stenosis or stenoses or ischaemi$ or ischemi$ or atherosclero$ or arteriosclero$ or constrict$ or occlu$ or narrow$ or plaque$ or obstruct$)).tw.
7. (posterior circulation$ adj5 (insufficien$ or stenosis or stenoses or ischaemi$ or ischemi$ or atherosclero$ or arteriosclero$ or constrict$ or occlu$ or narrow$ or plaque$ or obstruct$)).tw.
8. ((vertebral artery or vertebrobasilar or vertebro-basilar or vertebro basilar or posterior circulation) adj10 ((reduc$ or decreas$) adj3 blood adj3 flow)).tw.
9. VBI.tw.
10. vertebral artery/ra, us
11. 1 or 4 or 5 or 6 or 7 or 8 or 9 or 10
12. Magnetic Resonance Angiography/
13. angiography/ or cerebral angiography/
14. Magnetic Resonance Imaging/
15. 13 and 14
16. ((magnetic resonance or MR or MRI or NMR) adj5 (angiogra$ or arteriogra$)).tw.
17. MRA.tw.
18. exp Tomography, X-Ray Computed/
19. angiography/ or cerebral angiography/
20. 18 and 19
21. ((Compute$ tomograph$ or CT or CAT) adj5 (angiogra$ or arteriogra$)).tw.
22. CTA.tw.
23. (non-invasive adj5 imaging).tw.
24. ultrasonography/ or exp ultrasonography, doppler/
25. ((doppler or duplex) adj5 (ultrasound or ultrasonograph$ or ultrasonic)).tw.
26. DUS.tw.
27. ((doppler or transcranial) adj5 sonograph$).tw.
28. ((transcranial or trans cranial) adj5 (doppler or echodoppler)).tw.
29. TCD.tw.
30. 12 or 15 or 16 or 17 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29
31. 11 and 30
32. Vertebrobasilar Insufficiency/di, ra, us [Diagnosis, Radiography, Ultrasonography]
33. 31 or 32 

Appendix 2. Handsearched journals

1. Stroke
2. Lancet Neurology
3. Radiology
4. American Journal of Neuroradiology

Appendix 3. QUADAS tool

DOMAIN 1: patient selection   
  Bias   
 Describe methods of patient selection   
 

Was a consecutive or random sample of patients enrolled?

If the study enrolled all consecutive patients or a random sample of eligible patients then this item should be scored as “yes”. If study selection criteria are not clearly reported then this item should be scored as "no". In situations where patient enrolment procedures are only partially reported and you feel that you do not have enough information to score this item as "yes", then it should be scored as "unclear".

YesNoUnclear
 

Was a case-control design avoided?

If the study did not enrol patients with known disease and a control group of patients without the condition then this item should be scored as “yes”. If it appears that in the study a case-control design was applied, then this item should be scored as "no". If the design of the patient enrolment has not sufficiently been reported then this item should be scored as "unclear".

YesNoUnclear
 

Did the study avoid inappropriate exclusions?

If the study included all eligible patients and did not exclude patients with ‘red flags’ for the diagnosis, then this item should be scored as “yes”. If the study excluded patients with ‘red flags’ for the diagnosis or vertebral artery stenosis (e.g. due to previous testing) then this item should be scored as “no”. If the design of patient enrolment has not been sufficiently reported then this item should be scored as “unclear”.

YesNoUnclear
  Could the selection of patients have introduced bias? High Low Unclear
  Applicability   
 Describe included patients (prior testing, presentation, intended use of index test and setting)   
 

Is there concern that the included patients do not match the review question?

If the patients included in the study were admitted to the hospital for first evaluation of the carotid and/ or vertebral arteries then this item should be scored as "low". If the patients included in the study do not fit the prespecified patient spectrum then this item should be scored as "high". If the spectrum of patients has not been sufficiently reported then this item should be scored as "unclear".

HighLowUnclear
     

DOMAIN 2: index test(s)

*if more than one index test was used, please complete for each test

   
  Bias   
 Describe the index test and how it was conducted and interpreted:   
 

Were the index test results interpreted without knowledge of the results of the reference standard?

If the study clearly states that the index test results were interpreted blind to the results of the other test then these items should be scored as "yes". If this does not appear to be the case they should be scored as "no". If this information is not reported by the study then it should be scored as "unclear".

YesNoUnclear
 

If a threshold was used, was it pre-specified?   

If the threshold for diagnosing the presence of stenosis (both the degree of stenosis and the presence of stenosis) was pre-specified then this item should be scored as "yes". If this does not appear to be the case they should be scored as "no". If this information is not reported by the study then it should be scored as "unclear".         

YesNoUnclear
 

Were the results of the tests interpreted by individuals with adequate expertise?

If the clinician interpreting the imaging results of the index test is either a neurologist or radiologist (e.g. neurologist, radiologist) then this item should be scored “yes”. If the expertise has not been reported this item should be scored “no”.  In situations where details of expertise are partially reported and you feel that you do not have enough information to score this item as "yes", then it should be scored as "unclear".

YesNoUnclear
 

Could the conduct or interpretation of the index test have introduced bias? 

If the index test consists of either duplex ultrasound (DUS) with or without colour coding, magnetic resonance angiography (MRA) with or without contrast enhancement, computed tomographic angiography (CTA) or transcranial Doppler and the index test reports the presence of stenosis in the vertebral artery, then this item should be scored as "low". If another index test was applied then this item should be scored as "high". When the index test or its conduct have not been sufficiently reported then this item should be scored as "unclear".

High Low Unclear
  Applicability   
 Is there concern that the index test, its conduct, or interpretation differ from the review question?HighLowUnclear
     
Domain 3: reference standard   
  Bias   
 Describe the reference standard and how it was conducted and interpreted:   
 

Is the reference standard likely to correctly classify the target condition?

If intra-arterial DSA was used as the reference standard then this item should be scored “yes”. If the reference standard was different from intra-arterial DSA then this item should be scored “no”. If there is insufficient information to make a judgement then this item should be scored “unclear”.

YesNoUnclear
 

Were the reference standard results interpreted without knowledge of the results of the index test?

If the study clearly states that the test results (index or reference standard) were interpreted blind to the results of the other test then these items should be scored as "yes". If this does not appear to be the case they should be scored as "no". If this information is not reported by the study then it should be scored as “unclear”.

YesNoUnclear
 

Was the expertise of the clinician interpreting the imaging results of the reference test reported?

If the expertise of the clinician interpreting the imaging results of the reference test is reported (e.g. neurologist, radiologist) then this item should be scored “yes”. If the expertise has not been reported this item should be scored “no”.  In situations where details of expertise are partially reported and you feel that you do not have enough information to score this item as "yes", then it should be scored as "unclear".

YesNoUnclear
  Could the reference test, its conduct or interpretation have introduced bias? High Low Unclear
  Applicability   
 

Is there concern that the target condition as defined by the reference standard does not match the review question?

If the target condition in the study is vertebral artery stenosis50% or70% as measured by the NASCET ( Rothwell 2003 ) criteria with intra-arterial DSA, then the item should be scored "low". If another target condition was assessed or a different measurement was applied, then this item should be scored "high". If insufficient information is given on the target condition or measurement, then this item should be scored "unclear".

HighLowUnclear
     
Domain 4: flow and timing   
  Bias   
 Describe any patients who did not receive the index test(s) and/or reference standard or who were excluded from the 2x2 table (refer to flow diagram):   
 Describe the time-interval and any interventions between index test(s) and reference standard:   
 

Were all patients included in the analysis?

If all patients recruited in the study were included in the analysis, then this item should be scored as "yes". If it appears that some of the participants who entered the study did not complete the study, i.e. did not receive both the index test and reference standard, and these patients were not accounted for then this item should be scored as "no". If it is not clear whether all patients who entered the study were accounted for then this item should be scored as "unclear".

YesNoUnclear
 

Did patients receive the same reference standard?

If it is clear that patients received verification of their true disease status using the same reference standard then this item should be scored as "yes". If some patients received verification using a different reference standard this item should be scored as "no". If this information is not reported by the study then it should be scored as "unclear".

YesNoUnclear
 

Was there an appropriate interval between index test(s) and reference standard?

If intra-arterial DSA was performed within one month after or before the index test this item should be scored “yes”. If the time period between DSA and the index test was longer than one month, this item should be scored “no”. If insufficient information is provided this should be scored as "unclear".

YesNoUnclear
 

Did all patients receive a reference standard?

If it is clear from the study that all patients, or a random selection of patients, who received the index test went on to receive verification of their disease status using a reference standard then this item should be scored as "yes". If some of the patients who received the index test did not receive verification of their true disease state, and the selection of patients to receive the reference standard was not random, then this item should be scored as "no". If this information is not reported by the study then it should be scored as "unclear".          

YesNoUnclear
  Could the patient flow have introduced bias? High Low Unclear
Whiting P, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Annals of Internal Medicine. 2011;155(8):529-36

  

 

Contributions of authors

PJN, AGK and AC are responsible for the overall planning and conducting of the systematic review. JMW and YG will evaluate and comment on the pre-submission draft of the review.

Declarations of interest

AC has received research funding from the Netherlands Heart Foundation (grant number 2007BO45) for the Vertebral Artery Stenting Trial.

Sources of support

Internal sources

  • No sources of support supplied

External sources

  • Scottish Funding Council and the Chief Scientist Office of the Scottish Government through the Scottish Imaging Network, A Platform for Scientific Excellence (SINAPSE), UK.

Ancillary