Does a simple bedside sonographic measurement of the inferior vena cava correlate to central venous pressure?

BACKGROUND Bedside ultrasound has been suggested as a non-invasive modality to estimate central venous pressure (CVP). OBJECTIVE Evaluate a simple bedside ultrasound technique to measure the diameter of the inferior vena cava (IVC) and correlate to simultaneously measured CVP. Secondary comparisons include anatomic location, probe orientation, and phase of respiration. METHODS An unblinded prospective observation study was performed in an emergency department and critical care unit. Subjects were a convenience sample of adult patients with a central line at the superior venocaval-atrial junction. Ultrasound measured transverse and longitudinal diameters of the IVC at the subxiphoid, suprailiac, and mid-abdomen, each measured at end-inspiration and end-expiration. Correlation and regression analysis were used to relate CVP and IVC diameters. RESULTS There were 72 subjects with a mean age of 67 years (range 21-94 years), 37 (53%) male, enrolled over 9 months. Seven subjects were excluded for tricuspid valvulopathy. Primary diagnoses were: respiratory failure 12 (18%), sepsis 11 (17%), and pancreatitis 3 (5%). There were 28 (43%) patients mechanically ventilated. Adequate measurements were obtainable in 57 (89%) using the subxiphoid, in 44 (68%) using the mid-abdomen, and in 28 (43%) using the suprailiac views. The correlation coefficients were statistically significant at 0.49 (95% confidence interval [CI] 0.26-0.66), 0.51 (95% CI 0.23-0.71), and 0.50 (95% CI 0.14-0.74) for end-inspiratory longitudinal subxiphoid, midpoint, and suprailiac views, respectively. Transverse values were statistically significant at 0.42 (95% CI 0.18-0.61), 0.38 (95% CI 0.09-0.61), and 0.67 (95% CI 0.40-0.84), respectively. End-expiratory measurements gave similar or slightly less significant values. CONCLUSION The subxiphoid was the most reliably viewed of the three anatomic locations; however, the suprailiac view produced superior correlations to the CVP. Longitudinal views generally outperformed transverse views. A simple ultrasound measure of the IVC yields weak correlation to the CVP.