Accuracy and precision of the ultrasound cardiac output monitor (USCOM 1A) in pregnancy: comparison with three-dimensional transthoracic echocardiography

Background. Cardiac output (CO) monitoring is helpful in the assessment of critically ill pregnant women, but invasive monitors are often unsuitable for use. We aimed to measure agreement between the non-invasive ultrasound cardiac output monitor (USCOM) and three-dimensional transthoracic echocardiography (3D-TTE) in pregnant women. Methods. Healthy pregnant women from 25 weeks gestation onwards participated. In the left lateral position at rest, CO was measured with the USCOM and 3D-TTE. A single operator performed all USCOM measurements, with a different operator performing all echocardiography. Both were blinded to results from the other device. Each USCOM trace was analysed using two modes: flow trace (FT) and touchpoint (TP). A second, blinded USCOM reading was taken to assess reproducibility. Results. USCOM readings were obtained in 92, and 3D-TTE images in 85 participants. The mean CO was 5.7, 7.7, and 6.2 litre min(-1) measured by 3D-TTE, USCOM FT, and USCOM TP, respectively. USCOM bias was +2.0 litre min(-1) (FT) and +0.4 litre min(-1) (TP). Limits of agreement were -0.2 to +4.2 litre min(-1) (FT) and -1.4 to +2.3 litre min(-1) (TP). The mean percentage difference was 32.6% (FT) and 31.4% (TP) for CO and 27.0% (FT) and 27.5% (TP) for stroke volume. Intraclass correlation between repeated USCOM readings was 0.9 (FT) and 0.86 (TP). Conclusions. USCOM has acceptable agreement with 3D-TTE for the measurement of CO in pregnancy. The positive bias of the USCOM, particularly in the FT mode, may be due to the hyperdynamic cardiovascular state in pregnancy. We suggest using the TP mode in this patient population.