Comparison of Cardiac Output Measurement by Electrical Velocimetry with Echocardiography in Extremely Low Birth Weight Neonates

Abstract

<b><i>Introduction:</i></b> Electrical velocimetry (EV) offers a noninvasive tool for continuous cardiac output (CO) measurements which might facilitate hemodynamic monitoring and targeted therapy in low birth neonates, in whom other methods of CO measurement are not practicably feasible. <b><i>Methods:</i></b> This prospective observational study compared simultaneous cardiac output measurements by electrical velocimetry (CO<sub>EV</sub>) with transthoracic echocardiography (CO<sub>TTE</sub>) in extremely low birth weight (ELBW) neonates in the neonatal intensive care unit (NICU). Echocardiography was performed by 1 single examiner. Data were analyzed by Bland-Altman analysis and independent-samples analysis of variance. A mean percentage error (MPE) of <30% and limits of agreement (LOA) up to ±30% were considered clinically acceptable. <b><i>Results:</i></b> Thirty-eight ELBW neonates were studied and yielded 85 pairs of CO<sub>EV</sub> and CO<sub>TTE</sub> measurements. Bland-Altman analysis showed an overall bias (the mean difference) and LOA of −126 and −305 to +52 mL min<sup>−1</sup>, respectively, and an MPE of 66%. Patients with patent ductus arteriosus had a higher bias with LOA and MPE of −166.8, −370.7 to +37 mL min<sup>−1</sup>, and 69%, respectively. The overall true precision was 58%. <b><i>Conclusion:</i></b> This study showed high bias and lack of agreement between EV and TTE for measurement of CO in ELBW infants in NICU, limiting applicability of EV to monitor absolute values.

Accuracy and Trending Ability of Electrical Biosensing Technology for Non-invasive Cardiac Output Monitoring in Neonates: A Systematic Qualitative Review

“…Blohm et al ( 26 ) showed that PDA status showed a trend toward significance when TEBT was compared to TTE measured VTI mean bias ( p = 0.077) but not when TEBT was compared to TTE m-mode measurements. Hassan et al ( 35 ) showed a near doubling in mean CO bias between PDA and non-PDA measurements (−167 ml/min and −89 ml/min, respectively, p < 0.001).…”

“…Increased complexity of respiratory support intervention caused an increase in PE. Hassan et al ( 35 ) showed that HFV only had a significantly lower bias ( p = 0.002) as compared to IMV and CPAP.…”

“…From an initial search, 295 studies were identified, which were assessed for eligibility. After full-text review, 15 studies were included, of which 14 were agreement studies ( 22 – 35 ) and 1 was a trending analysis study ( 36 ) ( Figure 1 and Table 2 ).…”

“…PE in the 4 studies varied between 21 to 74.4%. Hemodynamic significant PDA was defined in only 2 studies ( 30 , 35 ) ( Table 7 ).…”

“…However, the lack of other reference technologies attests to the difficulties of invasive testing for CO or SV in sick and small neonates and the difficulties in performing cMRI studies. In many of the included method comparison studies, a variety of TTE methods were used to measure SV or CO – aorta measurements at the annulus ( 24 , 32 , 33 , 35 ), valve hinge points ( 25 , 26 , 30 , 34 ), and Darmon’s triangular technique ( 23 , 27 ) or were not specified ( 28 , 29 ). This inconsistency in standardization of TTE measurements may also contribute to the range of agreement and trending data found in this review.…”