Feasibility of biventricular 3D transthoracic echocardiography in the critically ill and comparison with conventional parameters

Background: Transthoracic 3D cardiac analysis is enticing in its potential simplicity and wealth of data available. It has been suggested to be accurate vs magnetic resonance imaging in relatively stable patients, but feasibility and agreement with conventional echocardiographic assessment of stroke volume (SV) have not been thoroughly assessed in critically ill patients, who are traditionally harder to image. The objectives of this study were to compare 3D transthoracic volumetric analysis vs Doppler assessment of SV (which is suggested to be accurate in the critically ill) and Simpson's biplane assessment in a cohort typical of the intensive care unit (ICU), where accurate assessment is important: mechanically ventilated patients with a significant ventilation/perfusion (V/Q) mismatch. We hypothesised that it would be feasible but might lack agreement. Methods: Patients were imaged within 24 hours of admission. Inclusion criteria were adult patients, V/Q mismatch present (defined as a ratio of arterial oxygen partial pressure to fractional inspired oxygen < 300), and mechanically ventilated with Doppler SV assessment possible. Biventricular echocardiographic volumetric analysis was performed using Siemens SC2000 along with standard Simpson's biplane and Doppler SV assessment. 3D images were unacceptable if two segments or more were unable to be seen in two volumetric planes. 3D left ventricular (3DLV) and 3D right ventricular (3DRV) analyses were performed with the Tomtec Imaging and Siemens Acuson platforms, respectively. Results: Ninety-two patients were included (83 in sinus, 9 in atrial fibrillation). 3DLV and 3DRV analyses were feasible in 72% and 55% of patients, respectively; however, they underestimated SV compared with Doppler by 2.6 ml (+/- 10.4) and 4.1 ml (+/- 15.4), respectively. Limits of agreement for 2D, 3DLV and 3DRV volumetric analysis techniques were large. Conclusions: 3DLV and 3DRV volumetric analyses appear feasible (obtainable) in the majority of mechanically ventilated ICU patients. Compared with the Doppler method, 3DLV and 3DRV volumetric analyses underestimate SV. The large limits of agreement between the methods also cast doubt on their comparability. Given the scenarios in which SV analysis is required (e.g., assessment of cardiac performance), our study cautions against the use of 3D SV clinically.