Clinical utility of aortic pulses and pressures calculated from applanated radial-artery pulses

Brachial artery cuff blood pressures are but approximations of central aortic pressures. The actual pressures against which the left ventricle must pump would be useful clinical information if obtained noninvasively. Our aim was to determine the clinical utility of aortic pulses and pressures calculated from noninvasively obtained radial-artery pulses. Radial-arterial pulses were recorded by applanation and calibrated with arm/cuff oscillometric pressures. Aortic pulses and pressures were calculated from the radial pulses by Fourier analysis and transfer functions. These calculated aortic pulses were compared with directly recorded aortic pulses by a transducer-tip catheter in a series of 50 patients undergoing cardiac catheterization. The correlation coefficient (r) of the measured versus the calculated aortic systolic blood pressure was + 0.89, but the scatter was large ( standard deviation of the differences = +/- 11.3 mm Hg). The pulse pressure correlations were less good ( r = + 0.79) and also had a large scatter ( +/- 13.6 mm Hg). The average calculated pulse pressure was 11.5 mm Hg lower than the measured value because the cuff diastolic blood pressures, used to calibrate the radial pulses, were systematically higher than those in the aorta ( 8.9 mm Hg). Multivariable analysis incorporating height, age, heart rate, and ejection fraction as additional, independent variables eliminated mean differences between the new "predicted" and measured pressures, significantly improved correlation coefficients, and reduced the scatter. However, the improvements were small. The inaccuracy of the oscillometric cuff method for measuring arm blood pressure appears to be the limiting factor in the prediction of clinically useful, noninvasive aortic pressures.