OBJECTIVETo assess agreement between the ultrasonic cardiac output monitor (USCOM) and conventional echocardiography (ECHO) in the measurement of cardiac output in newborn infants, investigate the accuracy and clinical utility of the USCOM in healthy neonates. To explore a more convenient, faster, more accurate hemodynamic monitoring method, for improving the outcome of the critically ill neonates.

METHODFrom October 1(st), 2011 to March 31(st), 2012, a total of 49 infants were included, 20 were term infants, 29 were preterm infants. Cardiac outputs were measured by both ultrasonic cardiac output monitor and echocardiography in all the infants, 60 times measurements were done in both the term infants the preterm infants. The cardiac output of the left and right ventricles, heart rate, diameter and velocity time integral of the aortic valve and pulmonary artery valve of each infant were recorded. The consistency of two methods was analyzed as described by Bland-Altman.

RESULTTerm the term infant group includea 20 term infants, 11 were male and 9 were female, the mean gestational age were (38.1 ± 0.56) weeks, mean age were (2 ± 1) days, mean weight were (3.2 ± 0.29) kg, mean Apgar score were 10. The mean left ventricular output measured by Echo was (242.3 ± 38.9) ml/(kg·min), measured by USCOM was (211.7 ± 38.5) ml/(kg·min); The mean right ventricular output measured by ECHO was (318.9 ± 47.0) ml/(kg·min), measured by USCOM was (340.7 ± 76) ml/(kg·min). Agreement between Echo and USCOM for left ventricular output (LVO) was (bias, ± limits of agreement, mean % error): (30.6 ± 51.1) ml/(kg·min), 21%, and for right ventricular output (RVO): (-21.8 ± 105) ml/(kg·min), 33.2%. The diameter of the aortic valve and pulmonary artery valve measured by conventional echocardiography were significantly larger than that estimated by ultrasonic cardiac output monitor (P < 0.001). The velocity time integral of the pulmonary artery valve measured by ultrasonic cardiac output monitor were significantly larger than measured by conventional echocardiography (P < 0.001). The heart rate, velocity time integral of the aortic valve measured by two methods had no significant differences (P > 0.05). The preterm neonates group included 29 preterm infants, 18 were male and 11 were female, the mean gestational age were (32.6 ± 2.8) weeks, mean age were (2 ± 1) days, mean weight were (1.88 ± 0.57) kg. All the infants were diagnosis as preterm infant, low birth weight. The mean left ventricular output measured by ECHO was (259.8 ± 70) ml/(kg·min), measured by USCOM was (235.6 ± 61.8) ml/(kg·min), the mean right ventricular output measured by ECHO was (318.9 ± 47.0) ml/(kg·min), measured by USCOM was (340.7 ± 76) ml/(kg·min). Agreement between Echo and USCOM for left ventricular output (LVO) was (bias, ± limits of agreement, mean % error): (24.1 ± 71.2) ml/(kg·min), 27.4%, and for right ventricular output (RVO): (-29.5 ± 192.9) ml/(kg·min), 51.8%. The diameter of the aortic valve and pulmonary artery valve measured by conventional echocardiography were significantly larger than estimated by ultrasonic cardiac output monitor (P < 0.001). The velocity time integral of the pulmonary artery valve measured by USCOM were significantly larger than that measured by conventional echocardiography (P < 0.001). The heart rate, velocity time integral of the aortic valve measured by two methods had no significant differences (P > 0.05).

CONCLUSIONAgreement between USCOM and conventional ECHO in the LVO measurement is acceptable, both in the term group and the preterm group. LVO measurement measured by USCOM is recommended. The accuracy and clinical utility of the USCOM in neonates is acceptable. USCOM is a convenient, fast and accurate hemodynamic monitoring method in neonates. While the agreement between USCOM and conventional ECHO in the RVO measurement is poor, especially in the preterm group, the results of the RVO cannot be considered interchangeable in the two methods.