P056: Non-invasive measurement of the central venous pressure using near-infrared spectroscopy versus point-of-care ultrasound

Abstract

Introduction: A fundamental hemodynamic parameter, the central venous pressure (CVP) is rarely available in the emergency patient due the delay and risks inherent to central vein cannulation. Recently, two non-invasive strategies have emerged: a) point-of-care ultrasound to supplement traditional inspection the internal jugular waveform ; or b) near-infrared spectroscopy (NIRS) of the external jugular vein. Methods: Five medical students underwent standardized training on both NIRS device (Venus 2000 CVP; Mespere Life Sciences, Waterloo ON) and ultrasound-assisted CVP assessment. During prescheduled, randomly permuted and balanced shifts, a pair of students obtained blinded independent measurements using each device within 10 minutes of each other. High priority subjects likely to have abnormal CVP (e.g. vomiting, dehydrated, heart failure, sepsis) were approached preferentially, followed by a convenience sample of other eligible patients in the emergency department. Secondary outcomes were stopwatch-recorded time from device ready to stable measurement, as well as operator ease, operator confidence and patient discomfort. The blinded treating physician rated each subjects volume status on an ordered scale: depleted, neutral and overloaded. Results: We enrolled 104 patients (median [IQR] age 68 [53, 78] years; 50% male; BMI 27.6 [17.0, 47.7] kg/m2; admission rate 27%) in June-August 2017. Treating physicians classified 17 as volume depleted and 12 overloaded. CVP measurements differed widely between techniques: ultrasound 8 [7, 9] cmH2O (3 cases unobtainable) vs NIRS 12 [8, 17] cmH2O (13 unobtainable). Agreement and correlation between the two devices was extremely low (R2=0.04). While neither technique demonstrated a strong association with the treating physicians estimate of volume status, only the ultrasound values increased monotonically with physician estimate. With regards to secondary outcomes, ultrasound measurements took less time (paired difference 50 seconds [95% CI 7, 93]), and operators were more confident (0.63 [0.02, 1.23] out of 10) and at ease (0.78, [0.13, 1.43]) with ultrasound; patients rated discomfort equally (-0.06 [-0.30, 0.18]). Conclusion: Non-invasive measurement of CVP remains a challenge in the emergency department. The external jugular pressure by NIRS has very high variability and poor agreement with ultrasound-enhanced inspection of the internal jugular, suggesting that this technique is not yet practical for use by non-experts.