Estimation of central aortic systolic pressure from the second systolic peak of the peripheral upper limb pulse depends on central aortic pressure waveform morphology

Background: Direct identification of second systolic peaks of peripheral upper limb pulses (pSBP2) has been used to represent central systolic blood pressure (cSBP), but its accuracy at low SBP was questioned. Objectives: We investigated the relationship of pSBP2 with characteristics of central pressure waveforms. Methods: High-fidelity central aortic and right brachial pressure waveforms were simultaneously recorded using a custom-made dual pressure sensor catheter in 78 patients (65.9 +/- 12.9 years) during catheterization for 285 measurements. Results: Overall agreement between cSBP and pSBP2 was good (mean difference -0.9 +/- 4.8, r = 0.98), with a systematic bias at low SBP. We examined agreements of different waveform types according to the relationship of the second systolic peak of aortic pressure waveforms (cSBP2) to cSBP. Of type A (positive late systolic augmentation) and type B (zero augmentation) aortic pressure waveforms, in which cSBP cSBP2, agreement between pSBP2 and cSBP was excellent (mean difference -0.4 +/- 4.1, r = 0.99). There were 40 type C aortic pressure waveforms (negative augmentation; cSBP > cSBP2) with cSBP 107.2 +/- 13.9 mmHg. Their cSBP2, compared with cSBP, showed closer agreement (mean difference -0.6 +/- 3.2 vs. -4.0 +/- 7.2 mmHg) and better correlation (r = 0.97 vs. 0.85, P = 0.03) with pSBP2. Conclusion: pSBP2 can be used with type A and B aortic pressure waveforms for estimation of cSBP. However, it should not be used with type C aortic pressure waveforms, typically at low SBP, because pSBP2 is closer to cSBP2 than cSBP. This explains why pSBP2 underestimates cSBP at low SBP.