Radial-femoral concordance in time and frequency domain-based estimates of systemic arterial respiratory variation.

Commonly used arterial respiratory variation metrics are based on mathematical analysis of arterial waveforms in the time domain. Because the shape of the arterial waveform is dependent on the site at which it is measured, we hypothesized that analysis of the arterial waveform in the frequency domain might provide a relatively site-independent means of measuring arterial respiratory variation. Radial and femoral arterial blood pressures were measured in nineteen patients undergoing liver transplantation. Systolic pressure variation (SPV), pulse pressure variation (PPV), area under the curve variation (AUCV), and mean arterial pressure variation (MAPV) at radial and femoral sites were calculated off-line. Two metrics, "Spectral Peak Ratio" (SPeR) and "Spectral Power Ratio" (SPoR) based on ratios of the spectral peak and spectral area (power) at the respiratory and cardiac frequencies, were calculated at both radial and femoral sites. Variance among radial-femoral differences was compared and correlation coefficients describing the relationship between respiratory variation at the radial and femoral sites were developed. The variance in radial-femoral differences were significantly different (p < 0.001). The correlation between radial and femoral estimates of respiratory variation were 0.746, 0.658, 0.858, 0.882, 0.941, and 0.925 for SPV, PPV, AUCV, MAPV, SPeR, and SPoR, respectively. Assuming a PPV treatment threshold of 12 % (or equivalent), differences in treatment decisions based on radial or femoral estimates would arise in 12, 14, 5.4, 5.7, 4.8, and 5.5 % of minutes for SPV, PPV, AUCV, MAPV, spectral peak ratio, and spectral power ratio, respectively. As compared to frequency domain-based estimates of respiratory variation, SPV and PPV are relatively dependent on the anatomic site at which they are measured. Spectral peak and power ratios are relatively site-independent means of measuring respiratory variation, and may offer a useful alternative to time domain-based techniques.