Diagnostic Performance of Ankle-Brachial Index: Building on a Foundation of Sand?

Background: The ankle-brachial index (ABI) is the mainstay of peripheral arterial disease (PAD) diagnosis. A recent Cochrane review found little evidence for the diagnostic accuracy of the test and recommended further cross-sectional studies of ABI in the diagnosis of PAD. Methods: A prospective database of patients attending a regional vascular laboratory was interrogated to identify patients who underwent an ABI and lower limb arterial duplex in the same sitting. The duplex acted as the reference standard with a stenosis 50% in any of the iliac, femoropopliteal or infra-popliteal considered diagnostic for PAD. The diagnostic performance of ABI was evaluated. Results: ABI plus arterial duplex was performed in 226 legs (118 patients) over a one-year period. The ABI was normal (0.9 to 1.4) in 78 limbs, 32 of which were noted to have a significant stenosis on arterial duplex. Of the 148 limbs with an abnormal ABI (0.9 or 1.4), 106 were found to have a significant stenosis. The area under the ROC curve for ABI was 0.71 (95% CI 0.58 to 0.84), indicating moderate predictive ability. ABI had a sensitivity for PAD of 76.8% and a specificity of 52.3%. 41% of patients with a negative test had significant PAD on duplex ultrasound. Conclusion: In this series, ABI has only moderate predictive ability as a diagnostic tool for PAD. In particular, a negative ABI cannot be taken to infer absence of PAD. The inclusion of waveform analysis may improve the tests performance.