Clevert 2008b.

Clinical features and settings	People who received a stent for AAA and were included in surveillance programme for endoleak detection. Type of stents received: not reported. Aneurysm diameter: not reported. Setting: radiology division.
Participants	43 consecutive participants; 43 males; mean age: 63 years. Comorbidities: not reported. Geography: Germany.
Study design	Prospective recruitment of participants ("patients who were referred to our radiology division to have a MS‐CT using commercially available equipment, a supplementary ultrasound investigation with a 4 MHz multi‐frequency probe was conducted with the modalities of colour duplex ultrasound and CEUS").
Target condition and reference standard(s)	Target condition: endoleak. Definition of endoleak: persistence of blood flow outside lumen of endoluminal graft but within an aneurysm sac or adjacent vascular segment being treated by the graft. Endoleak type I: If flow into aneurysm sac originated from around stent graft attachment site, it was classified as type I. Further categorization of type I endoleaks suggested as type IA (proximal), type IB (distal), or type IC, indicating persistent flow lateral to an iliac occlusion stent graft in contralateral limb after implantation of a mono‐iliac stent graft. Endoleak type II: Type II endoleaks represent retrograde blood flow through aortic branch vessels into aneurysm sac. They occur when blood travels through branches from unstented portion of aorta or iliac arteries that anastomose with vessels in direct communication with aneurysm sac. Typical sources include inferior mesenteric and lumbar arteries and Riolan's anastomosis. Endoleak type III: Type III endoleaks occur when there is structural failure with the stent graft such as stent‐graft fractures or holes that develop in fabric of device. In addition, category includes junctional separations seen with modular devices. Although type III endoleaks are currently rare, they will probably become more prevalent as stent grafts begin to age and long‐term follow‐up of these participants accrues. Endoleak (absolute n): 15. Prevalence of endoleak: 34.9% (15/43). Reference standard: 16‐ or 64‐detector CT scanner. Image acquisition: imaged volume included entire abdominal aorta from lower thoracic portion and common and external iliac arteries to upper femoral arteries. Acquisition direction cranio‐caudal; thin‐slice maximum‐intensity projections with increments of 0.6 mm and slice of 0.75 mm in coronal planes; for Sensation 64, collimation and table feed were 64 × 0.6 mm, rotation time 0.33, pitch 0.9, slice thickness 0.3 mm, reconstruction interval 0.3 mm; for Sensation 16, collimation and table feed were 16 × 0.75 mm, rotation time 0.5 s, pitch 1, slice thickness 3 mm, reconstruction interval 3 mm. Type of CT scanner: 16‐ or 64‐detector CT scanner (Somatom Sensation 16 or 64, Siemens Medical Systems, Forchheim, Germany). Use of contrast: 100‐120 mL Imeron (Altana Pharma, Germany) with iodine concentration 350 mg/mL.
Index and comparator tests	Index test: CDUS/CE‐CDUS. Image production: transverse and sagittal imaging. In CDUS, colour gain selected just as high as necessary to avoid overwriting artefacts (i.e. colour pixels outside perfused lumen of vessel). Additionally, an automatic image gain optimisation could be selected. For CEUS examinations, Sequoia systems were equipped with CPS software which detects the microbubbles' fundamental non‐linear response. CEUS employed continuous low mechanical index (0.15‐0.19) real‐time tissue harmonic imaging (Cadence) CPS imaging. The US device featured a high‐performance processor and allowed documentation of dynamic image sequences in cine mode by a digital frame buffer. Type of US: 4‐MHz multifrequency transducer (curved array). Use of contrast: yes; contrast agent (SonoVue, Bracco, Milan, Italy) consisting of stabilised microbubbles of sulphur hexafluoride administered into antecubital vein through 18‐G needle and followed by flush of 10 mL saline solution (0.9% NaCl). Operator: 1 (sonographer was unaware of CT scan).
Follow‐up	No loss to follow‐up, missing data, or adverse events.
Notes
Table of Methodological Quality
Item	Authors' judgement	Description
Representative spectrum? All tests	Yes	Participants consecutively enrolled for follow‐up to detect potential endoleaks.
Acceptable reference standard? All tests	Yes	Reference standard was CT scan.
Acceptable delay between tests? All tests	Yes	"For the whole study population, contrast‐enhanced CT examinations were performed within 1 day before contrast‐enhanced sonography."
Partial verification avoided? All tests	Yes	All participants who received US were subjected to CT scan.
Differential verification avoided? All tests	Yes	Index test not part of reference standard.
Incorporation avoided? All tests	Yes	Two tests were not incorporated.
Reference standard results blinded? All tests	Yes	All CT examinations performed and read by experienced radiologists who were blinded to results of both sonography and contrast‐enhanced sonography.
Index test results blinded? All tests	Yes	Sonographer unaware of CT scan results during the examination and reading of duplex and CEUS examination.
Relevant clinical information? All tests	Yes	Relevant clinical information reported.
Uninterpretable results reported? All tests	Yes	All data were interpreted. ("In the follow up the two false positive endoleaks (types I and II) in CE‐CDUS were confirmed as true positive endoleaks by CE‐CDUS and MS‐CT (Figs 7‐9). The sensitivity of CE‐CDUS was therefore 100% and its specificity 100%; the positive and negative predictive values were 1 and 1, respectively.")
Withdrawals explained? All tests	Yes	No withdrawals occurred.