The present study investigated the agreement between CO estimation by EV and TTE in preterm infants during postnatal transition, reporting an overall good agreement between the two techniques and satisfactory EV accuracy.

As described in a recent systematic review [9], EV proved better than other thoracic electrical biosensing technologies in terms of agreement with COECHO in the neonatal population; nevertheless, the reported bias, either negative or positive, varies significantly among the available studies, half of which reported a MPE > 30% [2,3,4,5,6].

EV would represent a useful tool for non-invasive CO monitoring in preterm infants, who are prone to significant hemodynamic instability. To date, however, the evaluation of the agreement between COEV and COECHO in this population has yielded variable results. While a relatively good consistency (i.e., low bias, narrow LOA) and a MPE < 30% was reported by several studies [2, 10, 11], others described a poorer agreement and accuracy [3, 4, 6]. Methodological factors, such as the inclusion of infants with different baseline characteristics (e.g., gestational and postnatal age, body size) or the occurrence of technological advancements (i.e., more precise algorithms, introduction of neonatal sensors) over the period during which these studies were performed, may underlie these heterogeneous findings.

Since the need for arterial catheterization limits the applicability of transpulmonary thermodilution for CO assessment in neonates, TTE is considered the clinical gold-standard for non-invasive CO estimation in this population. Nevertheless, echocardiographic CO measurements are not exempt from a significant inter- and intra-operator variability, and a MPE around 30% compared to thermodilution-derived measurements has been reported [1]. Hence, TTE may not represent the best reference method for COEV validation, and an increase of MPE threshold up to 45% has been suggested to compensate for COECHO variability.

According to our results, hsPDA was associated with a significant COEV overestimation compared to COECHO, consistently with previous data [12]. A significant hsPDA impact on both COEV and COECHO, although with a negative bias, was also reported by other studies [3, 10]. The interference of transductal shunt on volumetric changes and on the aortic alignment of erythrocytes during the cardiac cycle may underlie this finding.

To our knowledge, this is the first study investigating the influence of cardiovascular drugs on COEV accuracy. While no significant effect was observed with vasopressor agents, such as dopamine, a slight but significant COEV overestimation occurred during inotropic treatment with dobutamine; however, further validation is required to confirm this result and hypothesize potential underlying mechanisms.

In the present study, neither conventional nor high-flow oscillatory ventilation were associated with a significant proportional bias. Our results are in line with previous data reporting no significant effects of ventilatory modalities [5, 10]. Conversely, Hassan et al. described a lower bias in association with HFOV [3], whereas opposite evidence of higher bias and PE was reported by two studies [2, 4]. These variable findings may be ascribable to the different characteristics of infants requiring HFOV (gestational age, hemodynamic instability, hsPDA) and to the noticeably low number of HFOV measurements.

In the presence of systemic hypoperfusion, EV may play a potentially relevant role for CO monitoring; however, the limited number of infants with a left ventricular output < 150 ml/kg/min, which defines a low-flow state, limits the generalizability of the present results to this condition, which therefore requires targeted investigations.

Our data overall support the role for COEV monitoring in preterm infants during postnatal transition. The use of neonatal sensors, avoidance of inter-operator bias for COECHO and the relatively homogeneous characteristics of the study population may have contributed to the low bias and MPE. However, a slight COEV overestimation was observed in association with hsPDA and during dobutamine treatment, highlighting the importance of complementary echocardiographic assessments for clinical decision-making, especially in these conditions. Large and well-designed studies allowing to adequately analyse population subsets (e.g., different gestational and weight ranges) and the impact of clinical and hemodynamic factors are needed to better define COEV accuracy and precision.