Automated photoplethysmography-based determination of ankle-brachial index: a validation study against Doppler sonography

Introduction Determination of ankle-brachial-index (ABI) by manual Doppler is well established to screen for lower extremity arterial disease (LEAD) and to predict cardiovascular risk. A new generation of digital-controlled devices promises automated ABI determination. The aim of this study was to determine comparability of automated photoplethysmography (PPG)-derived ABI calculation with the Doppler-ABI algorithm commonly used in cohort studies. Methods Automated PPG-based ABI measurements [Vascular Explorer (VE) and Vicorder (VI)] were recorded from 112 limbs of healthy subjects and 22 limbs of patients with confirmed LEAD. Validity was evaluated on the basis of receiver-operating characteristic (ROC) analysis of clinical status and concordance with Doppler-ABI. Differences between cuff inflation [inf]- and deflation [def]-based method were studied in VE. Results PPG-based ABI values were higher compared to Doppler-ABI (VI +0.06, VEinf +0.15, VEdef +0.09, p < 0.001, respectively). The difference was pronounced in pathological (<0.9), borderline (0.9–0.99) and low normal (1.0–1.09) ABI, but less in ABI ≥1.1. However, ROC analysis revealed excellent diagnostic value for LEAD (sensitivity/specificity) and comparable area under the curve at method-adapted ABI thresholds for all methods: Doppler (95/90 %, 0.95), VI (75/96 %, 0.91), VEinf (85/89 %, 0.93) and VEdef (80/98 %, 0.94). Conclusions Digital-controlled PPG-based ABI determination is a useful diagnostic application for LEAD. However, the systematic higher ABI in PPG-based measurement compared to Doppler and remarkable differences between the deflationary and inflationary method are critical for the interpretation of borderline and low normal ABI values where precise reading is essential to detect mild LEAD and subclinical disease and to predict cardiovascular risk.