Premium
Comparing the Accuracy of Physiologic and Triangular Pseudoflow Waveforms for Estimating Wave Reflection Indices in Adults
Author(s) -
Shenouda Ninette,
Stock Joseph M.,
Patik Jordan C.,
Chirinos Julio A.,
Edwards David G.
Publication year - 2020
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.2020.34.s1.09271
Subject(s) - waveform , aortic pressure , blood pressure , reflection (computer programming) , mathematics , aorta , cardiology , medicine , physics , computer science , quantum mechanics , voltage , programming language
Central aortic pressure waveforms contain valuable prognostic information in addition to central systolic pressure. Using wave separation analysis, aortic pressure‐flow relations can be used to decompose aortic pressure waveforms into forward (Pf) and backward‐travelling (Pb) components. Pf and Pb are used to calculate reflection magnitude (RM=Pb/Pf) which is predictive of heart failure and all‐cause mortality. Aortic blood flow can be measured non‐invasively via ultrasonography or estimated using a triangular pseudoflow waveform; however, the latter can underestimate true flow and thus overestimate Pb and RM. Recently, a physiologic pseudoflow waveform has been developed that improves on the assumptions of triangulation and so may provide a more accurate estimation of wave reflection. PURPOSE The purpose of this study was to compare estimates of Pf, Pb, and RM derived from physiologic and triangular pseudoflow waveforms relative to measured flow. We hypothesized that the physiologic waveform would provide a more accurate estimate of wave reflection indices compared to the triangular waveform. METHODS In 75 adults (age 18–75 yrs), radial artery pressure waveforms were acquired with applanation tonometry and a generalized transfer function was used to synthesize central pressure waveforms. Aortic blood flow velocities were measured via echocardiography and estimated using physiologic (in‐house software) and triangular (SphygmoCor CvMS) pseudoflow waveforms. Pressure‐flow relations were established offline and wave separation analysis was performed. Accuracy of the physiologic and triangular‐derived estimates of Pf, Pb, and RM were assessed using 95% equivalence testing, paired t‐tests comparing the mean absolute biases, and Bland‐Altman plots. RESULTS Using a 10% equivalence zone as the maximum acceptable difference, the physiologic and triangular pseudoflow waveforms were both comparable to the measured flow waveform in deriving Pf. However, only the physiologic waveform yielded similar estimates of Pb and RM compared to the measured waveform. Further, the physiologic waveform yielded lower mean absolute biases compared to the triangular waveform for Pf (1.7 ± 2.1 vs. 2.8 ± 2.6 mmHg, p<0.001), Pb (1.0 ± 1.1 vs. 3.3 ± 2.7 mmHg, p<0.001), and RM (3 ± 3 vs. 14 ± 13%, p<0.001). Lastly, Bland‐Altman biases were smaller, and 95% limits of agreement were narrower, for all wave reflection indices derived from the physiologic waveform than from the triangular waveform. CONCLUSION These data suggest that the physiologic pseudoflow waveform provides a more accurate estimate of wave reflection indices compared to the triangular waveform, and thus may offer a superior alternative when true flow cannot be measured.