Abstract
BACKGROUND: Non-invasive clinic-based tools for assessing PAD are not without limitations. Therefore, costly tests like Doppler study, CT angiography and MR angiography are often required to make a diagnosis. Ankle brachial index (ABI), commonly used for assessment of PAD, has high false positivity rates in sclerosed, calcified arteries which render them non-compressible. Toe brachial index (TBI) can be an alternative, as digital arteries are relatively unaffected by these changes. AIM: To compare the reliability of ABI and TBI in diagnosing PAD in type 2 diabetes using CT angiography (CTA) as the reference. METHODS: 175 adults with T2D were selected. ABI &TBI were measured with an automated vascular Doppler XT 6 ports bilaterally for all subjects. For any subject, the limb with lower ABI and TBI was included for analysis. ABI < 0.9 & TBI < 0.6 were taken as evidence of PAD. CTA showing > 50% narrowing was taken as evidence of PAD. RESULTS: 24% of our study subjects had CTA confirmed PAD. ABI has low sensitivity of 35.29% (95% CI 0.21-0.52) compared to TBI being 82.35% (95% CI 0.66-0.92). The specificity however was similar. ABI < 0.9 was able to detect CTA confirmed PAD, but ABI > 0.9, including the so-called normal ABI (0.9-1.3) was unable to detect PAD. ROC showed ABI at 1.005 has sensitivity 64.71% (95% CI 0.48- 0.79) and specificity 61.7% (95% CI 0.53-0.69) and TBI at 0.6 has sensitivity 82.35% (95% CI 0.66-0.92) & specificity 92% (95% CI 0.87-0.96). Utilizing Cohen's Kappa, the reliability of ABI with respect to CTA showed fair agreement (K = 0.225, p = 0.001), whereas the reliability of TBI with respect to CTA showed substantial agreement (K = 0.759, p < 0.0001). CONCLUSION: ABI < 0.9 detects PAD reliably, but presence of PAD in patients with ABI > 9.0 including the normal of ABI (0.9-1.3) can be confirmed with TBI, which correlated strongly with CTA. TBI is also non-inferior for PAD detection, when ABI < 0.9. TBI and not ABI can be utilized for initial assessment of PAD in subjects with T2D.