COMPARISON OF INFERIOR VENA CAVA ASSESSMENT BY 2-DIMENSIONAL AND M-MODE ECHOCARDIOGRAPHY FOR THE EVALUATION OF RIGHT ATRIAL PRESSURE

Abstract

Echocardiographic estimation of right atrial pressure (RAP) is used to assess intravascular volume, is a marker of prognosis in patients with cardiovascular disease, and is an essential component of the Doppler echo measurement of pulmonary artery pressure. Current guidelines recommend measuring the diameter and collapsibility of the inferior vena cava (IVC) to assign a value for RAP but do not explicitly state whether 2-dimensional (2D) or M-mode (MM) echocardiography should be used. This has resulted in variations in clinical practice among centres. We sought to determine if MM and 2D techniques differ in their evaluation of RAP using current recommendations. 2D and MM images of the IVC were prospectively acquired by sonographers in 100 clinical echocardiograms performed on non-ventilated patients over a four-week period. Two reviewers independently assessed the 2D and MM images of the IVC and made guideline-based measurements of the diameter and collapsibility. Quality of MM cursor positioning was documented. RAP was classified as normal (3 mmHg), intermediate (8 mmHg), or high (15 mmHg) based on IVC diameter and collapsibility for the MM and 2D techniques. IVC measurements were larger by MM than 2D (IVC diameter 1.82±0.51 vs. 1.70±0.49 cm, p<0.001), and agreement between MM and 2D on the presence of IVC dilation was significantly influenced by the quality of M-mode cursor positioning (χ2=9.6582, p=0.002). IVC collapsibility was also greater by MM than 2D (61±20% vs. 55±22%, p=0.001). The agreement of MM- and 2D-derived RAP values was only moderate-good (74% agreement, weighted kappa 0.61) with no indication that one technique consistently overestimated or underestimated RAP relative to the other. In the subset of patients with MM RAP derived from an improperly positioned MM cursor (n=24), agreement between 2D and MM RAP was worse (58% agreement, weighted kappa 0.39, p=0.055 for interaction between MM cursor positioning and RAP agreement). Interobserver agreement for RAP was similar for 2D (weighted kappa=0.67) and MM (weighted kappa=0.69). IVC size and collapsibility differed when evaluated by MM and 2D techniques and resulted in discrepant assessments of RAP in 26% of patients. Quality of the MM cursor positioning was a predictor of the agreement between MM and 2D IVC size and tended to affect the RAP ultimately assigned. This study suggests that discrepant measurements of RAP using MM and 2D techniques are common in clinical practice, but can be minimized by ensuring proper MM techniques are employed.