Accuracy of Ankle-Brachial Index, Toe-Brachial Index, and Risk Classification Score in Discriminating Peripheral Artery Disease in Patients With Chronic Kidney Disease

Abstract

The accuracy of ankle-brachial index (ABI) and toe-brachial index (TBI) in discriminating lower extremity peripheral artery disease (PAD) has not been evaluated in patients with chronic kidney disease (CKD). We measured ABI, TBI, and Doppler ultrasound in 100 predialysis patients with CKD without revascularization or amputation. Leg-specific ABI was calculated using higher systolic blood pressure (SBP) in posterior tibial or dorsalis pedis artery divided by higher brachial SBP; alternative ABI was calculated using lower SBP in posterior tibial or dorsalis pedis artery. PAD was defined as ≥50% stenosis detected by Doppler ultrasound. PAD risk classification score was calculated using cardiovascular disease risk factors. The area under the curve (AUC, 95% confidence interval [CI]) for discriminating ultrasound-diagnosed PAD was 0.78 (0.69 to 0.87) by ABI, 0.80 (0.71 to 0.89) by alternative ABI, and 0.74 (0.63 to 0.86) by TBI. Sensitivity and specificity were 25% and 97% for ABI ≤0.9, 41% and 95% for alternative ABI ≤0.9, and 45% and 93% for TBI ≤0.7, respectively. AUC (95% CI) of PAD risk classification score was 0.86 (0.78 to 0.94) with sensitivity and specificity of 95% and 60% for risk score ≥0.10, 76% and 76% for risk score ≥0.25, and 43% and 95% for risk score ≥0.55. Combining risk score with ABI, alternative ABI, and TBI increased AUC (95% CI) to 0.89 (0.82 to 0.96), 0.89 (0.80 to 0.98), and 0.87 (0.78 to 0.96), respectively. In conclusion, current ABI and TBI diagnostic criteria have high specificity but low sensitivity for classifying PAD in patients with CKD. PAD classification risk score based on cardiovascular disease risk factors improves the accuracy of PAD classification.