Accuracy of ankle–brachial index obtained by automated blood pressure measuring devices in patients with diabetes mellitus

Abstract Background Upper arm automated blood pressure devices are widely available and could be used to estimate the ankle–brachial index. Methods We conducted a trial to determine the equivalence of ankle–brachial index estimated by an upper arm blood pressure measuring device as index method compared to the handheld Doppler method as the reference standard. A total of 110 patients with diabetes mellitus were sequentially examined by two methods. Results The prevalence of peripheral arterial disease was 32%. The index method obtained valid measurements in 104 (95%) patients. Ankle–brachial index was lower with the index method compared to the reference standard (mean difference: −0.05; 95% confidence interval [CI]: −0.50 to 0.39). This confidence interval was above the boundaries clinically established as equivalence margins in our study. The kappa agreement between two methods was 0.45. The performance of the index method was: sensitivity: 67%; specificity: 87%; positive likelihood ratio: 5.25; negative likelihood ratio: 0.18; positive predictive value: 71%; negative predictive value: 85%; and the area under the receiving operating characteristic curve: 0.87 (95% CI: 0.78–0.93). Conclusion Upper arm automated blood pressure measuring devices cannot replace the handheld Doppler method to estimate the ankle–brachial index in patients with diabetes mellitus.