Implications for clinical monitoring of intra-arterial blood pressure based on the frequency content of worst-case pressure waveforms.

There is no consensus regarding the minimum acceptable frequency response for a system intended for routine monitoring of invasive blood pressure in the operating rooms or intensive care units. Part of the problem stems from the diverse criteria used to validate faithful reproduction of the pressure waveform and waveform parameters. Another part of the problem is related to the site of pressure measurement: some anatomic sites have waveforms that are more difficult to reproduce than others. The most often quoted criteria, around which many clinical blood pressure monitoring systems have been constructed, are based on reproduction of all details of the pressure waveform from the most demanding anatomic sites. However, the routine clinical setting is primarily concerned with obtaining accurate values for systolic and diastolic blood pressures as recorded from peripheral arterial sites. This study was designed to obtain worst-case, in the sense of most difficult to measure, blood pressure waveforms from the femoral artery and the left ventricle using a canine model. The purpose of worst-case analysis was to create pressure waveforms that were at both extremes of frequency content, i.e., a rapidly beating heart with extraordinary contractility and a more slowly beating heart with little contractility. The premise of this type of analysis is that all naturally occurring pressure waveforms within a very broad physiologic range would fall within these extremes; therefore, the ability to measure the extreme waveforms would imply the ability to measure all those in between. Worst-case waveforms chosen from among 360 pressure waveforms were analyzed to determine the minimum frequency content of each waveform that would be required for faithful reproduction of systolic and diastolic pressures and the dP/dt. The criterion for faithful reproduction was chosen to be +/- 5% or 1 mmHg, whichever is greater. Data from the worst-case canine study, with supporting data collected intraoperatively from three patients, demonstrate that systolic and diastolic pressures can be obtained from a peripheral measurement site within 5%, requiring only two harmonics in conjunction with an appropriately low-pass-filtered catheter manometer system. Left ventricular pressure waveforms require five harmonics to reproduce the systolic and diastolic pressures within 5% or 1.0 mmHg, whichever is greater. Determination of the maximum dP/dt within +/- 5% from the worst-case peripheral and left ventricular waveforms required 20 and 22 harmonics, respectively.