Even with clinical recognition and targeted hemodynamic resuscitation [11,12,13], the mortality rate for neonates with septic shock was as high as 34% [14]. The high burden was likely due to the non-precise clinical evaluation [15, 16] and invasive risk from the hemodynamic monitoring [17]. On the other hand, ECHO showed good consistency with invasive hemodynamic methods such as pulse index contour cardiac output (PiCCO) [18] and central venous pressure (CVP) [19].

According to the guidelines from the American Echocardiography Association, during forced breathing for adults, the width of the IVC and the IVC-CI should correspond to CVP, indicating the amount of fluid should guide clinical decision-making [20]. Therefore, the combination of IVC and IVC-CI can improve the accuracy of blood volume assessment. In a study of 70 children aged 1 month to 12 years, the results show a negative correlation between IVC-CI and CVP. IVC-CI > 50% corresponded to CVP < 8 cmH2O, indicating insufficient blood volume and usefulness of IVC-CI to evaluate the blood volume status of critically ill pediatric patients [21]. These indicators show their application for management of pediatric shock patients [22, 23], but without evidence for neonates. Our study fulfills the knowledge gap by providing reference values for IVC parameters across gestational ages and birth weights among stable neonates.

In this study, IVCmin, IVCmax, and AO were collected by a cardiac ultrasound doctor and a NICU doctor, and there were no statistically significant differences between them. In a study using adults, there was intra- and inter-observer accuracy in measuring IVC by ECHO [24]. This indicates that the parameters can serve as important indicators for evaluating hemodynamics in critically ill patients. Our results provide normal reference ranges for IVCmin, IVCmax, IVC-CI, IVCmin/AO, and IVCmax/AO stratified by gestational maturity and birth weight in hemodynamically stable neonates (the control group). IVC diameter and AO increased with greater maturity and size, while IVC-CI and IVC/AO were similar across groups. Another study in neonates shows that there was a good negative correlation between IVC-CI and CVP in mechanically ventilated patients, but no correlation with gestational age and weight. IVCmax and IVCmin were not correlated with CVP, but had a good positive correlation with gestational age and weight [25]. A study in healthy children also shows that age, height, and weight were positively correlated with IVCmax and IVCmin, while IVC-CI was not significantly correlated with age, height, and weigh [26]. These are consistent with the results of our study. In addition, IVC-CI, IVCmin/AO, and IVCmax/AO did not change with the days of increased age, and the presence of the PDA (Not hsPDA) did not affect them. In autonomous breathing, the IVC contracts during inhalation and expands during exhalation. However, during positive pressure mechanical ventilation, the intrathoracic pressure and right atrial pressure increase during inhalation, and the amount of blood flowing back from veins to the right atrium decreases, affecting the diameter of the IVC. The IVC expands during inhalation and contracts during exhalation. Therefore, the measurement of IVCmax and IVCmin in patients with positive pressure mechanical ventilation was opposite to the normal breathing state, that is, IVCmax was measured during inhalation and IVCmin was measured during exhalation [27]. Another study show that IVC-CI was not affected by whether patients were breathing spontaneously or were mechanically ventilated [28]. Intra-abdominal hypertension had no effect on IVC-CI but reduced IVCmax, and large IVCmax with no collapse, that is, not hypovolemic [29]. The above information indicates that the IVC-CI value did not depend on individual’s physical parameters and breathing patterns, but rather on blood volume status, making it a good indicator for evaluating blood volume. In summary, IVC-CI, IVCmin/AO, and IVCmax/AO were not affected by gender, gestational age, and birth weight. Therefore, newborns were organized into one age group.

The target range of IVC-CI was usually between 20% and 50% [28, 30], which is different from the approximately 18–40% range of IVC-CI in newborns in this study. This may be because neonates normally have faster respiratory rate and lower respiratory amplitude than adults or children. A report on 23 healthy premature infants with a gestational age of 30.9 ± 2.9 weeks and a birth weight of 1146 (966, 1460) g shows that IVC-CI fluctuated between 15% and 24%, with an average of 20% [31]. Another report shows measured IVC-CI values of 12–46% (P10 to P90) in 25 healthy full-term infants with an average weight of 3425 g [32]. There is a certain deviation from the results of this study, which may be explained by different sample sizes. Nonetheless, these values provide context for interpreting IVC measurements in hypotensive states.

Except for birth weight, the general characteristics of the EOS-shock group were not statistically significant compared to the control group. As shown above, IVC-CI, IVCmin/AO, and IVCmax/AO were not affected by weight, therefore the two groups were comparable. Our observations show that IVC and IVC/AO declined markedly while IVC-CI rose in neonates with EOS-shock compared to stable controls across etiologies. These findings concur with adult and pediatric studies showing IVC narrowed and collapsed to a greater degree in shock [33, 34]. A study was reported on comparing the levels of 5 preload indicators between a septic shock and a healthy control groups of 46 premature infants [31]. The results show that only IVC-CI was significantly increased in the septic shock group, which is consistent with the results of our study.

A report described IVC collapsibility > 40% as 93% sensitive and 100% specific for hypovolemic shock pediatric patients [23]. A prospective longitudinal study involved 66 newborns with low blood volume (CVP < 5 cmH2O), 22 newborns with normal blood volume (CVP 5–8 cmH2O), and 34 newborns with high blood volume (CVP > 8 cmH2O) [35]. The results show that the sensitivity of predicting low blood volume at a cut-off value of 55% for IVC-CI was 87.9%, the specificity was 82.0%, and the sensitivity of predicting high blood volume at a cut-off value of 20% was 91.1%, the specificity was 83.2%. IVC-CI was significantly and negatively correlated with CVP, suggesting that IVC-CI can be used to guide fluid resuscitation and the application of vasoactive drugs in neonatal shock. In our study, IVC-CI > 34.15% had a sensitivity of 82.35% and a specificity of 74.75% for EOS-septic shock. The lower IVC-CI may be explained by the shallower and faster breathing of newborns, while the lower sensitivity and specificity may be explained by the insufficient number of shock patients in newborns. A previous study shows the optimal cut-off value of IVC/AO was 0.675 in neonates [4], which was similar to IVCmax/AO < 66.11% in our study. Our data also provides an optimal cutoff value of IVCmin/AO < 47.58%, which can be combined with the cutoff values of IVC-CI and IVCmax/AO to jointly warn of EOS-septic shock. Our findings reinforce the potential diagnostic utility of IVC metrics for compromised neonatal circulation.