Accuracy and precision of USCOM versus transthoracic echocardiography before and during pregnancy

Objective: Monitoring hemodynamic status throughout pregnancy may help in identifying women with maladaptation predisposing to hypertensive complications. The Ultrasonic Cardiac Output Monitor (USCOM) is an easy-to-operate device for measuring cardiac output (CO) quickly. Our aim was to assess agreement between USCOM and transthoracic echocardiography (TTE) in: 1) non-pregnant women to correct for possible sources of discrepancy; 2) women longitudinally over the course of the pregnancy. Study design: High-risk women admitted for cardiovascular risk factor evaluation before pregnancy and multiple times during pregnancy, were included. CO was measured by TTE directly followed by USCOM measurements. Main outcome measures: Bias, limits of agreement (LOA) and percentage error between the two methods by Bland-Altman analysis. Results: Despite comparable non-pregnant CO levels (4.6 L/min), LOA and percentage error between the two methods improved moderately by optimizing the measurements using only the highest quality USCOM recordings in 132 non-pregnant women (percentage error of 39% and 30%, respectively). During pregnancy, in total 83, 106, 96 and 77 measurements were evaluated at respectively 12, 16, 20 and 30 weeks gestational age. Mean CO in USCOM was about 0.6 L/min higher compared to TTE in all trimesters; percentage error ranged from 35% to 45%. Linear mixed model analysis showed no association between bias and moment of measurement. Conclusion: Agreement between USCOM and TTE in pregnancy was outside our a priori determined level of acceptability and therefore absolute values of USCOM and TTE cannot be used interchangeably. Future research should focus on the agreement of USCOM and TTE in clinical decision-making.