Evaluation of stroke volume via arterial pulse pressure waveforms in neonatal lambs

Arterial pulse waveforms contain information about stroke volume (SV) as an integral of pulsatile flow. SV estimation is accurate in adults with proper pulse pressure measurement technique. It is unclear whether the same methods are suitable in critically ill infants in the neonatal clinical setting where the fidelity of pulse pressure measurements are uncertain. We compared three pulse waveform SV methods with three systolic area SV methods in neonatal lambs in order to identify the most accurate and precise approach. Six newborn lambs were studied. Each lamb had a ligated ductus arteriosus and was instrumented to record high-fidelity pulsatile waveforms of arterial blood pressure using a transducer-tipped catheter and pulsatile flow via calibrated ultrasonic flow probe, respectively. Three steady-state hemodynamic conditions were induced experimentally: control, hypertension via infusion of angiotensin II, and hypotension by phlebotomy. Recordings of a range of SVs were made during a steady state that was interrupted by a transient period of decreasing SV, induced by momentarily increasing preload by pulmonary artery occlusion. Modification of pulse wave pressure measurement conditions, simulating an overdamped fluid-filled catheter system, were achieved by low-pass digital filtering of the original high-fidelity waveforms (high) to an 8-Hz cut-off (medium) and to a 2-Hz cut-off (low). The six SV estimates were then calibrated against flowmeter-derived SV and their accuracy and precision evaluated. Based on 6,479 waveforms, a systolic area method with pulse contour integration was the most accurate and precise. We conclude that neonatal pulse arterial waveforms embed SV information under a wide variety of hemodynamic and pressure waveform measurement conditions, and thus may be of potential clinical value in the assessment of newborn cardiovascular status.