Comparison of abdominal aortic aneurysm diameter measurements obtained with ultrasound and computed tomography: Is...

Accurate diameter measurements of abdominal aortic aneurysm (AAA) with both computed tomography (CT) and ultrasound (US) are essential for screening, planning surgical intervention, and follow-up after endovascular repair. Often there is a discrepancy between measurements obtained with CT and US, and neither limit of agreement (LOA) nor correlation between the two imaging methods has been clearly established. The purpose of this study was to assess the paired differences in AAA diameter measurements obtained with CT and US in a large national endograft trial.

CT and US measurements were obtained from an independent core laboratory established to assess imaging data in a national endograft trial (Ancure; Guidant, Menlo Park, Calif). The study included only baseline examinations in which both CT and US measurements were available. Axial CT images and transverse US images were assessed for maximal AAA diameter and recorded as CT(max) and US(max), respectively. Correlations and LOA were performed between all image diameters, and differences in their means were assessed with paired t test.

A total of 334 concurrent measurements were available at baseline after endovascular repair. CT(max) was greater than US(max) in 95% (n = 312), and mean CT(max) (5.69 +/- 0.89 cm) was significantly larger (P <.001) than mean US(max) (4.74 +/- 0.91 cm). The correlation coefficient between CT(max) and US(max) was 0.705, but the difference between the two was less than 1.0 cm in only 51%. There was less discrepancy between CT(max) and US(max) for small AAA (0.7 cm, 15.3%) compared with medium (0.9 cm, 17.9%) and large (1.46 cm, 20.3%) AAA; however, the difference was not statistically significant. LOA between CT(max) and US(max) (-0.45-2.36 cm) exceeded the limits of clinical acceptability (-0.5-0.5 cm). Poor LOA was also found in each subgroup based on AAA size.

Maximal AAA diameter measured with CT is significantly and consistently larger than maximal AAA diameter measured with US. The clinical significance of this difference and its cause remains a subject for further investigation.