Validation of ambulatory cuff-based devices for central aortic blood pressure and arterial stiffness measurement

Background: As the utility of non-invasive measurement of central aortic blood pressure parameters grows, the logical next step is for out-of-clinic measurement of aortic blood pressure parameters. Recently, ambulatory blood pressure monitors (ABPMs) have come on to the market with this ability. Aim: To investigate the accuracy of two ABPM devices that estimate central aortic blood pressure (BPLab* and AtCor/SunTech Oscar 2) by comparing them in-clinic to existing validated devices, the SphygmoCor XCEL (cuff-based) and tonometer-based devices. Methods: 45 subjects were recruited for the study, with 1 subject being excluded due to arrhythmia. The remaining subjects were 46±17 years old (30 male) with brachial systolic pressure ranging from 104 to 153 mm Hg, diastolic pressures from 63–106 mm Hg, heart rate from 48–94 bpm. Measurements were taken in triplicate with each device and averaged for repeated-measures statistical comparison of calculated central aortic systolic, diastolic, and augmentation index (AIx) between devices. Devices were also compared by Bland-Altman representation. Calculated aortic systolic pressure was 4.1±3.3 mm Hg higher in the Oscar 2 device than the tonometer-based device (tonometer-based 112±2, XCEL 113±2, Oscar 2 116±2 mm Hg; P=0.20). Aortic diastolic pressures were consistent between devices (tonometer-based 79±1, XCEL 80±1, Oscar 2 79±1 mm Hg; P=0.067). Aortic AIx was greater in the XCEL, and greater still in the Oscar 2 device compared to the tonometer-based device (tonometer-based 13±2%, XCEL 18±2%, Oscar 2 25±2%; P=0.003). There was little bias seen in the Oscar 2 device compared to the tonometer-based device for either aortic systolic (slope=0.92, intercept=12.7 mm Hg, R²=0.92, P<0.001) or aortic diastolic (slope=1.00, intercept=0.3 mm Hg, R²=0.99, P<0.001). However, AIx had a slope between devices less than unity (slope=0.74, intercept=15.7%, R²=0.36, P<0.001). Taking international guidelines for brachial blood pressure measurement in the absence of guidelines for central aortic blood pressure calculation, the Oscar 2 device has a Grade A rating. However, calculated augmentation index was considerably different between devices. (BPLab data has been collected, but the analysis is currently unavailable due to a software license issue that will be resolved before the conference date.) Conclusions: Our findings suggest that the cuff-based ABPM calculation of aortic systolic and diastolic pressure out of clinic settings appears feasible. However, the calculation of waveform shape parameters, such as augmentation index, requires further investigation.