In our study, the use of RT-US to assess the position of the tip of umbilical catheters resulted in a significant increase in the accuracy of proper UVC and UAC placement. We also found a decreased requirement for catheter adjustment after the first X-ray, and a decreased requirement for a follow-up X-ray image after the catheter adjustment.

To the best of our knowledge, our study was one of the few RCT to demonstrate that RT-US improves the accuracy of umbilical catheter placement. Our findings are consistent with a small RCT by Fleming et al. that compared US-guided UVC and UAC insertion with the conventional approach (calculation based on anthropometric measurements) [10]. Although they did not find a difference in the number of UAC manipulations between the two groups, US-guided insertion decreased the number of UVC manipulations, time of catheter placement, and number of X-rays taken. Kaur et al. conducted a similar RCT, but only UVC placement was investigated with similar results to the other trial [11].

In our study, the accuracy of UVC placement significantly enhanced from 20.4% to 64.0% when RT-US was used instead of Shukla’s formula. Previous investigations reported an accuracy of Shukla’s formula ranging from 17.8% to 57.0% for UVC insertion [9, 11, 15,16,17,18,19]. The variation in accuracy could be attributed to different definitions of the optimal catheter position (e.g., inferior vena cava to right atrial junction, 0.5–1.0 cm above the diaphragm or at the 8–10th thoracic vertebrae). Our results were consistent with those previously report in the literature. Rossi et al. performed a retrospective cohort study and found improved accuracy of UVC placement from 46.2% to 58.8% when US was used instead of Shukla’s formula [9]. Additionally, Kaur et al. observed improved accuracy of UVC placement from 26.0% to 57.7% when US was used instead of typical formula-based insertion [11]. Similarly, the accuracy of UAC placement increased considerably from 33.3% to 83.8% when RT-US was used instead of Shukla’s formula. Previous investigations reported an accuracy of Shukla’s formula ranging from 52.4% to 90.6% for UAC insertion [15, 16, 18,19,20]. However, research on US-guided UAC placement is scarce. In the study of Rossi et al., only two neonates received US for UAC placement, and no analysis of UAC was performed [9].

In this study, we did not intend to use US to help mobilize the catheter through the inferior vena cava passing through the liver and ductus venosus. However, a recent study reported that US can be used for this purpose. In 2024, Misha et al. compared the failure rate of US-guided UVC insertion with the blind approach and found no significant difference between the two groups (20.6% vs. 29.6%, p = 0.27) [12]. However, when catheter insertion failed in the blind group, US-guided reinsertion was attempted, and this helped route the catheter into the proper position in 30% of cases. This possibility could explain the decreased rate of low catheter placement and fewer catheters removed in the RT-US group in our study.

The results of multiple investigations, including our own, indicate that the methods currently used for calculating the length of umbilical catheter insertion are not accurate [9, 11, 15,16,17,18,19,20] Therefore, we developed an equation (linear regression) using the data from our study on the optimum catheter length (not including the umbilical stump length) in relation to the neonate’s birth weight to predict the length of umbilical catheters. The equations for the UVC and UAC are as follows: UVC length (cm) = birth weight (kg) + 5.26 (R2 = 0.63, p < 0.001) and UAC length (cm) = 2 × birth weight (kg) + 9.60 (R2 = 0.79, p < 0.001). Interestingly, a previous study in northern Thailand that aimed to determine a new formula to calculate the UVC length using data of 135 Thai neonates reported closely comparable equation to ours [21].

In our study, the total number of catheter readjustments was considerably lower in the RT-US group than in the No-US group (RR: 0.32, 95% CI 0.19-0.54, p < 0.001 in UVC; RR 0.19, 95% CI 0.08-0.47, p < 0.001 in UAC). Our findings are consistent with a previous study of Fleming et al. who found that the number of UVC manipulations was lower in patients who had US than in those with standard methods (2.8 to 1.6 times per patient, p = 0.002) [10]. However, in their study, the number of UAC manipulations was not significantly different between the two methods.

In our study, the total number of X-rays performed in the RT-US group was significantly lower than that in the No-US group (RR 0.71, 95% CI 0.51–0.97, p = 0.034). A study by Fleming et al. found a reduction in the number of X-rays per patient from 4.1 in the control group to 2.3 in the US group [10]. Rossi et al. reported that the number of X-rays per patient was lower from 1.5 to 1.19 after US was used to assist in catheter placement [9]. Rubortone et al. showed a reduction in the number of X-rays performed in pre/post intervention when US was used to replace X-rays for confirming the UVC tip position (92.3% vs. 32.1%, p < 0.001) [22]. Kaur et al. also reported a 45% reduction in additional X-ray exposure in the group that used US-guided UVC insertion [11]. In our study, there was a trend toward lower radiation exposure in the RT-US group than in the No-US group, however this was not statistically significant. In contrast, Rossi et al. found that the effective radiation dosage (mSv) per episode was significantly lower in the US group [9].

The average X-ray expense per episode was considerably higher in the No-US group than in the RT-US group because of a reduction in the number of X-rays performed in each patient. However, the expense of an US examination was not included in this study. Rossi et al. also showed that US used was cost-effective with a cost saving of £19.33 per line [9].

In this study, the decision to perform a repeated X-ray after catheter adjustment was left to the physician who performed the procedure. Some physicians may not want to repeat X-rays following catheter manipulation, while others might demand them to verify catheter insertion in the proper position. Although the latter might be preferable from a medico-legal standpoint, the final X-ray confirmation of proper placement was performed only in 52.9% of neonates in the No-US group and in 86.0% of neonates in the RT-US group. If final X-ray confirmation of proper catheter placement became mandatory, there would be a considerably higher disparity in the number of X-rays obtained, X-ray expense, and possibly radiation exposure between the two groups.

In the RT-US group, the average time to perform an US to access the tip of the catheter position was approximately 8 minutes, which is surprisingly similar to that reported in the study of Mishra et al. in 2024 [12]. In our NICUs, the routine procedure is to contact the radiology service and use a portable digital X-ray machine to verify the correct positioning of the UVC and UAC after successfully placing them in vessels with an expected insertion length and adequate blood flow. This practice provides an interval (varying from 10 to 45 minutes) in which the RT-US procedure can be performed to accurately locate and adjust the catheter before the X-ray technician’s arrival. Consequently, the use of RT-US had no effect on the total duration of the procedure or the workflow. In contrast, neonates in the RT-US group were more likely to have a suitable catheter location confirmed in the initial X-ray. This finding could account for the trend toward a slightly reduced total procedure duration in the RT-US group. Several authors reported that US-guided UVC placement reduces the total procedure duration because it decreases the requirement for repeated adjustment or eliminates the requirement for an X-ray [11, 12, 22]. Increasing evidence suggests that US may be more accurate than X-rays in determining the position of umbilical catheters. Therefore, if US had been the primary method utilized to evaluate the placement in our study, the overall procedure time may have been reduced even further.

Our study used AP chest X-ray as the gold standard for determining the umbilical catheter tip position because it is the standard of care in our NICUs and the most commonly used modality to confirm tip placement worldwide [3, 4]. However, the optimum method to determine the umbilical catheter position is subject to debate. This optimum method is controversial because, in studies where US was performed on an appropriately positioned UVC documented on an X-ray, a large percentage of these catheters were shown to be malpositioned (likely to be too high, e.g., located in the right atrium) [23, 24]. A recent systematic review that compared X-ray with US to verify the UVC tip position reported an X-ray sensitivity of 0.90 (95% CI 0.7-0.97) and a specificity of 0.82 (95% CI 0.53–0.95) [25]. In our study, more than half of the catheters in the RT-US group that were evaluated as being in an improper position and needed to be adjusted after an X-ray were those in which the radiologist was unable to comment on their position in an US image (10/18 in UVCs and 3/6 in UACs). The overall inter-rater reliability (k values) of an US image specifically for these catheters was relatively low (UVC: 0.220; UAC: 0.217). Poor US visualization of the catheter tip is likely to be the main explanation of these findings, emphasizing the importance of the abilities of the provider, which may be addressed by standardized protocols and training [22, 26].

There were no significant differences in short- and long-term complications, such as hypothermia, hypotension, brady-tachycardia, and desaturation, during the procedure between the two groups. Despite the theoretical concern that an US scan may contaminate the sterile field and increase the risk of central line infection or sepsis, there was no significant difference in the infection rate between the two groups in this study. The group B Streptococcus in blood culture of a neonate in the RT-US group could have been vertically transmitted. Similar to our study, Mishra et al. identified central line-associated bloodstream infection in 9 of 98 patients in their study [12]. However, there was no difference in these incidents between the US-guided UVC insertion and blind groups.

The limitations of our study include a small sample size, which may lead to an overestimation of results. Additionally, we have limited long-term outcome data and did not evaluate the financial expenses of US usage. The open-label design of the study may have introduced bias, although we implemented several measures to minimize this. All catheters in the controls were strictly inserted according to the standard formula. The definition of a catheter tip’s proper positioning on an X-ray was predetermined and adhered to rigorously. The physician responsible for the catheter insertion decided on the manipulation of the catheter and the necessity of additional X-rays, with the expectation of adhering to high standards of practice. To ensure quality, all X-ray and US images were reviewed by a radiologist (NR) who was not involved in patient care. Lastly, our study was conducted across three university hospital NICUs operated by the same group of physicians with consistent training and practices, which may limit the generalizability of our findings to other hospital settings. Despite that, the results of our study may help in developing comprehensive training programs that promote ongoing education and proficiency, thereby supporting wider adoption and effective utilization of this technology.

Growing evidence worldwide underscores the benefits of US as a point-of-care tool in NICUs [5]. As the new generation of neonatologists gains increasing expertise in US, its application for assessing catheter tip placement in neonates is expected to gain more popularity and could potentially be integrated into routine practice. Future research should confirm our findings and address gaps in knowledge, including implementation and training challenges as well as cost-effectiveness and long-term outcomes. Finally, alternative approaches to facilitate the correct positioning of the catheter should be explored and evaluated in comparison to existing techniques.