Abstract 3118: Different Calculations Of Ankle Brachial Index And Cardiovascular Risk Prediction

Abstract

The quotient between ankle and brachial blood pressure (ankle brachial index =ABI) is widely used for peripheral arterial disease (PAD) screening. Guidelines on ABI measurement do not uniformly define whether the higher or the lower ankle pressure should be applied to the calculation. We hypothesized that considering the lowest ankle pressure as numerator and the highest brachial pressure as denominator would increase the prognostic impact. Methods: In 831 patients preceding coronary angiography, ABI was calculated for both legs according to the current definition of the American Heart Association (highest ankle pressure/ highest arm pressure) and according to an extended definition (lowest ankle pressure/ highest arm pressure). For each patient the lowest ABI from both legs was used for further evaluation. 15 patients with an ABI >1.5 were excluded. PAD was defined, if an ABI of less than 0.9 could be detected in one leg with each ABI definition. Results: Using the current ABI definition 204 patients (25.0%) and using the extended definition 292 patients (35.8%) with PAD could be identified. Follow-up data (median 6.6 years) was available for 812 patients (99.5%) and 157 patients (19.3%) suffered from cardiovascular events (death, infarction, stroke). Event rate was significantly higher in patients with PAD (current ABI definition: PAD yes / no=28.1% / 16.3%, P<0.0001; extended ABI definition: PAD yes / no=27.4% / 14.8%, P<0.0001). Event rate was 25.0% in 88 patients with an ABI >0.9 using the current and >0.9 using the extended ABI definition. Hazard ratio (95% confidence interval) was 1.5 (1.1–2.1) for patients with PAD using the current ABI definition (P=0.02) and 1.7 (1.2–2.3) for patients with PAD using the extended ABI definition (P=0.004) compared to patients without PAD. In addition, including ABI as continuous variable improved the area under the ROC curve from 0.68 (0.63– 0.73) for a basic model to 0.70 (0.65– 0.75) for a basic model plus ABI based on current definition and 0.71 (0.66 – 0.76) for a basic model plus ABI based on extended definition. Conclusion: Using the highest ankle pressure the real PAD prevalence is underestimated and a group of patients at risk is overlooked. Therefore we recommend the use of the lowest ankle pressure for ABI calculation.