This study retrospectively analysed the medical records of patients admitted to a tertiary neonatal unit between July 2023 and January 2025. The study was conducted at a neonatal center in Kraków, Poland, where approximately 1,700 infants are born annually including premature babies and children with congenital defects. The unit also admits neonates referred from other hospitals in the region and serves as a certified training site for specialization in neonatology Inclusion criteria included an attempt to insert a UVC catheter and a minimum hospital stay of three days. The primary indication for UVC placement was the anticipated need for parenteral nutrition for three or more days. Contraindications included omphalocele, gastroschisis, a planned abdominal surgical procedure within the first days of life, and the absence of a ductus venosus, which was either reported prenatally or identified postnatally during neonatal ultrasound prior to the procedure.

Before UVC insertion, ultrasound was performed to assess the anatomy of the portal tract and the ductus venosus, in order to determine the trajectory of the ductus venosus as it branches from the left portal vein. A wider angle between the ductus venosus and the left portal vein, combined with a wider ductus venosus, facilitated catheter advancement (Figure 1). In this study, a difficult ductus venosus was defined as one with a smaller angle of origin and/or a narrower lumen (Figure 2). Although this was considered a relative contraindication, all patients were included regardless of ductus venosus anatomy.

Two clinicians were involved in each insertion– either a consultant neonatologist and a trained resident, or two trained residents. Residents were considered adequately trained after at least two months of daily ultrasound-guided catheter assessments during procedures or routine monitoring, and after performing a minimum of 25 UVC insertions. After incising the umbilical vein and advancing the catheter approximately 4–5 cm, further positioning was guided by ultrasound. The catheter was first visualized in the umbilical vein and then gradually advanced into the portal sinus under continuous ultrasound guidance, until reaching the ductus venosus. Correct catheter tip placement was defined as positioning in the subdiaphragmatic vestibule. The procedure was completed by securing the catheter. If difficulty arose in directing the catheter through the venous system into the subdiaphragmatic vestibule various manoeuvres were employed to achieve proper placement, such as posterior liver mobilization (5), gentle compression of the upper abdomen—which involved applying titrated pressure to the liver with a transducer in order to compress the left portal vein, accentuate the caudal angulation of the right portal vein, and straighten the “S-turn” from the portal sinus to the ductus venosus, a manoeuvre most frequently used in our practice (6)—or slight rotation of the patient to the right (7). All manoeuvres were performed under continuous ultrasound visualization, with the choice determined by whether the catheter was seen entering the liver parenchyma or deviating into the right portal vein—information available only with real-time ultrasound. For imaging, either a microconvex transducer or a 12–15 MHz linear probe was used. The choice of probe depended primarily on the patient's body weight: the smaller microconvex transducer was used mainly in patients weighing less than 1,500 g. The entire procedure, including the final catheter placement, was recorded. A longitudinal subcostal view was employed to visualize the catheter. Figure 3 shows the UVC during the procedure, prior to entering the ductus venosus, whereas Figure 4 demonstrates the catheter after successful advancement into the ductus venosus.

In the Supplementary Files, we have included a video showing the placement of a UVC under ultrasound guidance.

If initial placement was unsuccessful, the catheter was withdrawn, and an alternative vascular access was required. Ultrasound assessments were conducted every 24 h to monitor for complications until catheter removal. The study's primary outcome was establishing the number of successfully placed catheters. The secondary outcome was the occurrence of mechanical liver injury related to catheter insertion and manipulation excluding thrombosis.

The calculations were performed using TIBCO Statistica 13 and PQStat from PQStat Software. The level of significance was set to α = 0.05, and the result was considered statistically significant when p < α.

To examine the relationship between categorical variables, the chi-square independence test or Fisher-Freeman-Halton test was calculated.

The normality of the distribution of variables was tested with the Shapiro–Wilk test. To compare two groups of variables with a normal distribution and equal variance, the Student's t-test for unrelated samples was calculated. In the case of non-compliance with the normal distribution or for variables measured on an ordinal scale, the Mann–Whitney test was calculated.

During the study period, ultrasound guided UVC placement was attempted in 310 neonates. Four patients were transferred to another hospital with surgical facilities, and one patient died. Of the transferred cases, three had congenital heart defects requiring cardiac surgery, and one had oesophageal atresia. The infant who died was an extremely premature neonate born at 24 weeks of gestation, who succumbed to congenital sepsis. These cases were excluded from the final analysis due to the inability to assess procedure-related complications. The final analysis included 305 patients.

The statistical analysis showed no differences between the group in which the catheter was successfully inserted and the group in which the catheter was not successfully inserted in terms of gender, body weight, Apgar score at 5 min. The only statistically significant difference was the gestational age. The statistical results and group characteristics are summarized in Table 1.

The analysis of the reasons for catheter insertion took into account the most common ones, i.e., respiratory failure, prematurity, low body weight, hypoglycaemia and congenital anomalies (Table 2) and showed no statistical difference between the groups: p = 0.130.

The overall success rate for correct placement of UVC was 86.56% and additional analysis over a six-month period revealed an increasing trend in insertion efficiency (Table 3).

However, using the Cochran–Armitage trend test, this difference did not reach statistical significance (p = 0.2645). Analysis of catheter insertion complications identified one hepatic hematoma around the left portal vein branch, detected via ultrasound 24 h post-insertion. In this case, the catheter tip was correctly positioned. No other liver complications were observed.