Ultrasound measurement for abdominal aortic aneurysm screening: a direct comparison of the three leading methods.

OBJECTIVES Ultrasound (US) is non-invasive and cost-effective for screening abdominal aortic aneurysms (AAAs) but there is no universally accepted method to measure the aortic diameter. This study evaluates the accuracy, reproducibility, and repeatability of three methods: inner-to-inner (ITI), leading-to-leading edge (LTL), and outer-to-outer (OTO). The secondary objective of this study was to determine whether aneurysm size or grade of operator had any effect on either intra- or inter-observer variability. METHODS Fifty static US images were measured by six assessors (2 vascular radiologists, 2 interventional radiology trainees, and 2 sonographers) on two separate occasions 6 weeks apart. Repeatability and reproducibility were calculated and compared with computed tomography (CT) as the gold standard. RESULTS All three methods have high repeatability and reproducibility when static images are used. The inter-observer reproducibility coefficients between assessors were 0.48 cm, 0.35 cm, and 0.34 cm for ITI, LTL and OTO, respectively. The intra-observer repeatability coefficients between assessors were 0.30 cm, 0.20 cm, and 0.19 cm for ITI, LTL and OTO, respectively. The mean difference between CT and OTO, LTL, and ITI was 1 mm, 3 mm, and 5 mm, respectively (all underestimations) (p < .0001). CONCLUSIONS US consistently underestimates aortic size when compared with CT, with ITI demonstrating the greatest underestimation (on average 5 mm). In the UK, this underestimation by the NHS Abdominal Aortic Aneurysm screening programme reduces the sensitivity of the screening test and may impact on the way in which vascular specialists interpret the findings of the screening programme.