Automated Noninvasive Central Blood Pressure Measurements by Oscillometric Radial Pulse Wave Analysis: Results of the MEASURE-cBP Validation Studies

BACKGROUND Central blood pressure becomes increasingly accepted as an important diagnostic and therapeutic parameter. Accuracy of widespread applanation tonometry can be affected by calibration and operator training. To overcome this, we aimed to evaluate novel VascAssist 2 using automated oscillometric radial pulse wave analysis and a refined multi-compartment model of the arterial tree. METHODS Two hundred and twenty-five patients were prospectively enrolled. Invasive aortic root measurements served as reference in MEASURE-cBP 1 (n = 106) whereas applanation tonometry (SphygmoCor) was used in MEASURE-cBP 2 (n = 119). RESULTS In MEASURE-cBP 1, we found a mean overestimation for systolic values of 4 +/- 12 mmHg (3 +/- 10%) and 6 +/- 10 mmHg (9 +/- 14%) for diastolic values. Diabetes mellitus and low blood pressure were associated with larger variation. In MEASURE-cBP 2, mean overestimation of systolic values was 4 +/- 4 mmHg (4 +/- 4%) and 1 +/- 4 mmHg (1 +/- 7%) of diastolic values. Arrhythmia was significantly more frequent in invalid measurements (61 vs. 18%, P < 0.0001) which were most often due to a low quality index of SphygmoCor. CONCLUSIONS Central blood pressure estimates using VascAssist 2 can be considered at least as accurate as available techniques, even including diabetic patients. In direct comparison, automated measurement considerably facilitates application not requiring operator training and can be reliably applied even in patients with arrhythmias.