Frequency response evaluation of radial artery catheter-manometer systems: Sinusoidal frequency analysis versus flush method

Abstract

It is well recognized that catheter-manometer systems significantly distort direct radial artery pressure measurements. Sinusoidal frequency analysis and the flush method of assessing the degree of distortion caused by the monitoring system were compared to determine whether these two methods agree in the estimation of natural frequency and damping coefficient. The frequency response of 30 radial artery catheter-manometer systems used for intensive-care unit patients was measured by the flush method and sinusoidal frequency analysis. The monitoring system consisted of a 20-gauge cannula, 150-cm pressure tubing, two plastic stopcocks, a continuous infusion device with fast flush valve, an American Edwards dome, a Hewlett-Packard quartz transducer, and a Hewlett-Packard blood pressure amplifier. Sinusoidal frequency analysis demonstrated second-order underdamped response for all 30 catheter-manometer systems. No secondary resonance peaks were observed up to a frequency of 200 Hz. The measured frequency response demonstrated that the average catheter-manometer system in use in our intensive care unit would cause significant distortion of the radial artery pressure, with the mean natural frequency (fn) of 14.7 ± 3.7 Hz and the mean damping coefficient (ζ) of 0.24 ± 0.07. Although the 30 monitoring systems had identical configurations and visible bubbles were carefully removed, a wide range of frequency responses was found (fn=10.2 to 25.3; ζ=0.15 to 0.44). Linear regression analysis demonstrated excellent correlation (r=0.96) for the natural frequency measured by the flush method and sinusoidal frequency analysis. Linear regression for the damping coefficient, on the other hand, demonstrated poor correlation (r=−0.24) for the two measurement methods. Four of the 30 catheter-manometer systems exhibited overdamped behavior by the flush method; sinusoidal frequency analysis demonstrated an underdamped response. The difference in damping coefficient measurements between the two techniques occurs because sinusoidal frequency analysis does not include the effect of the cannula within the radial artery. Because of the position of the tip, this effect can significantly influence the damping coefficient. Since there is a wide variation in responses for debubbled identical systems, it is necessary to test each individual monitoring system used to determine its frequency response. The flush method is more reliable than sinusoidal frequency analysis since it includes the effect of the intraarterial cannula.