Aortic systolic pressure values but not indices derived from waveform features are consistent between brachial cuff-based devices used for estimation of central aortic pressure

Objective:
For ease of measurement, and the utility of ambulatory central aortic blood pressure measurement, there has been a move toward brachial cuff-based devices for non-invasive computation of central aortic blood pressure quantities based on waveform features. However, waveforms detected by volumetric cuff displacement techniques are inherently more damped than signals obtained by applanation tonometry, potentially impacting on parameters reliant on higher frequency components of the pulse waveform.

Design and method:
In 45 subjects (age 46 ± 17 years, 30 male), in-clinic, seated measurements taken in triplicate using three brachial cuff-based devices (BPLab, [Petr Telgin]; Oscar2, [SunTech/AtCor Medical]; SphygmoCor XCEL, [AtCor Medical]) were compared using repeated measures ANOVA and Bland-Altman statistics against radial tonometric assessment of central aortic pressure (SphygmoCor CvMS, [AtCor Medical]). Results are expressed as means ± standard error.

Results:
There was good agreement between devices for aortic systolic pressure (aSP) and aortic diastolic pressure (aDP). There was great variability in aortic augmentation index (aAIx), ejection duration (ED) and subendocardial viability ratio (SEVR, Table). Cuff-based device regression slopes against the tonometer-based method varied markedly for aSP (BPLab, 0.76; Oscar2, 0.92; XCEL, 0.77), aAIx (BPLab, 0.32; Oscar2, 0.74; XCEL, 0.88), ED (BPLab, 1.07; Oscar2 does not report; XCEL, 0.83), and SEVR (BPLab, 0.16; Oscar2 does not report; XCEL, 0.81).
Conclusions:
Parameters relying on the low frequency components of the peripheral waveform have better agreement between cuff-based devices than parameters that rely on higher frequency waveform components. Further research is required for quantitative assessment of filtering methods utilised in cuff-based devices, as well as the cuff-based approach itself for use in measuring AIx, ED and SEVR.