Diagnostic performance of computed tomography angiography in peripheral arterial disease: a systematic review and meta-analysis.

Abstract

Computed tomography angiography (CTA) is an increasingly attractive imaging modality for assessing lower extremity peripheral arterial disease (PAD). To determine the accuracy of CTA compared with intra-arterial digital subtraction angiography (DSA) in differentiating extent of disease in patients with PAD. Search of MEDLINE (January 1966-August 2008), EMBASE (January 1980-August 2008), and the Database of Abstracts of Reviews of Effectiveness for studies comparing CTA with intra-arterial DSA for PAD. Eligible studies compared multidetector CTA with intra-arterial DSA, included at least 10 patients with intermittent claudication or critical limb ischemia, aimed to detect more than 50% stenosis or arterial occlusion, and presented either 2 x 2 or 3 x 3 contingency tables (< or = 50% stenosis vs > 50% stenosis or occlusion), or provided data allowing their construction. Two reviewers screened potential studies for inclusion and independently extracted study data. Methodological quality was assessed by using the QUADAS instrument. Of 909 studies identified, 20 (2.2%) met the inclusion criteria. These 20 studies had a median sample size of 33 (range, 16-279) and included 957 patients, predominantly with intermittent claudication (68%). Methodological quality was moderate. Overall, the sensitivity of CTA for detecting more than 50% stenosis or occlusion was 95% (95% confidence interval [CI], 92%-97%) and specificity was 96% (95% CI, 93%-97%). Computed tomography angiography correctly identified occlusions in 94% of segments, the presence of more than 50% stenosis in 87% of segments, and absence of significant stenosis in 96% of segments. Overstaging occurred in 8% of segments and understaging in 15%. Computed tomography angiography is an accurate modality to assess presence and extent of PAD in patients with intermittent claudication; however, methodological weaknesses of examined studies prevent definitive conclusions from these data.