Abstract
Umbilical venous catheters (UVCs) are commonly used in neonatal units. Abdominal radiograph, taken soon after the insertion, is used to confirm the correct placement of the catheter. However, as demonstrated by our case series, complacency when an initial UVC is normally positioned can lead to detecting UVC-related complications very late. We describe cases of three neonates where, despite the initial UVC being confirmed to be well positioned, the patients developed life-threatening complications, including liver haematoma, left atrial thrombus and pericardial effusion causing cardiac tamponade. The routine use of point of care echocardiography helped in the early diagnosis of these complications. We suggest repeated echocardiographic screening at regular intervals in all newborn babies requiring UVCs, to ensure that the catheter is maintained in the proper position and for the early detection of catheter-related complications.
Background
Umbilical venous catheters (UVCs) are commonly used in neonatal intensive care units (NICUs). Though UVC provides an easy and reliable source of central venous access, it is not free of complications. Abdominal radiograph taken soon after the insertion of UVC is usually the only investigation performed to ensure the appropriate position of the catheter in most NICUs worldwide. Hence many complications of the UVC are diagnosed late. We report three cases of UVC-related complications that were diagnosed early and serendipitously.
Case presentation
Case 1
The first case is of a baby boy born at 31 weeks of gestation, with a birth weight of 1400 g. An UVC was inserted on day 1 of life, for total parenteral nutrition (TPN). In the abdominal radiograph, the catheter tip was found at the level of the diaphragm and T8 vertebra (figure 1).
Figure 1.
Abdominal radiograph (case 1, day 1) showing appropriate position of the catheter tip at the level of the diaphragm and T8 vertebra.
On day 4 of life, the baby developed signs of poor perfusion and severe metabolic acidosis. Septic shock was considered. He was ventilated and started on inotropes. When bedside echocardiography was carried out for assessing cardiac contractility and inferior vena cava (IVC) filling, a hypoechoic 3×2 cm mass with septations and hyperechoic rim was found in the liver parenchyma next to the ductus venous. The tip of the UVC was found within the mass (figure 2 and video 1). Repeat X-ray showed the UVC tip at the level of the T11 vertebra within the liver shadow (figure 3).
Figure 2.
Ultrasonography (case 1, day 4) showing a hypoechoic mass with septations and hyperechoic rim in the liver parenchyma next to the ductus venous and the tip of the umbilical venous catheter within the mass.
Figure 3.
Radiograph (case 1, day 4) showing the umbilical venous catheter tip at the level of the T11 vertebra within the liver shadow.
As the clinical course was suggestive of sepsis, ultrasound-guided needle aspiration was performed in order to rule out an abscess. One millilitre of haemorrhagic fluid was aspirated, the culture of which was sterile. The UVC was removed and the baby was given inotropic and ventilator support for 48 h. Antibiotics were given for 5 days as the blood culture was sterile, after which the C reactive protein was negative and the baby improved clinically. Repeat ultrasound carried out after 2 weeks showed only a small remnant 8×6 mm mass (figure 4). The further course of hospital stay was uneventful and the baby was discharged on day 45 of life.
Figure 4.
Ultrasonography (case 1, after 2 weeks) showing a small remnant mass (marked with an asterisk).
Case 2
The second case is of a preterm infant born at 30 weeks of gestation, with a birth weight of 1020 g. The baby was started on nasal continuous positive airway pressure soon after birth, for respiratory distress syndrome. An UVC was inserted on day 1 of life and the tip of the catheter was confirmed to be at the appropriate position at the level of the diaphragm by abdominal radiography (figure 5). On day 2 of life, the baby developed tachycardia and signs of poor perfusion. Bedside echocardiography showed only a haemodynamically significant patent ductus arteriosus, which was treated with oral Ibuprofen.
Figure 5.
Radiograph (case 2, day 1) showing appropriate placement of the umbilical venous catheter with tip at the level of the diaphragm.
Repeat echocardiography carried out on day 4 of life showed that the ductus was closed, but an echogenic 7×5 mm mass was found inside the left atrium. The mass was a thrombus formed around the tip of the UVC; it was confirmed by the cardiologist, using echocardiography. After removal of the UVC, a track formed by the UVC was found within the thrombotic mass (video 2). Repeat X-ray showed that the catheter had migrated upwards, and the tip was found above the diaphragm at the level of the T7 vertebra (figure 6).
Figure 6.
Radiograph (case 2, day 5) showing high position of the catheter tip above the diaphragm at the level of the T7 vertebra.
The baby was started on low-molecular-weight heparin after consulting the haematologist. Factor Xa levels were monitored and heparin dose was adjusted accordingly. The plan was to continue heparin for 6–12 weeks. However, the parents did not want to continue their baby's treatment. They were counselled about the baby's need for intensive care and the risk of discontinuing treatment. However, they insisted on discharge, which occurred against medical advice on day 7 of the child's life.
When the baby was presented for follow-up at 10 weeks, it was found that the baby had been given treatment for 3 weeks in a hospital proximate to the family's home, but further doses of heparin had not been given. The baby was active and thriving well. Repeat echocardiography showed complete resolution of the thrombus (video 3).
Case 3
The final case is of a baby girl born at 34 weeks of gestation, with a birth weight of 1500 g. The baby was started on gavage feeds initially. As she was not tolerating feeds well, an UVC was inserted on day 1 of life, and she was started on total parenteral nutrition (TPN). X-ray of the abdomen showed that the UVC tip was in an acceptable position below the diaphragm at the level of the T10 vertebra (figure 7).
Figure 7.
Radiograph (case 3, day 1) showing umbilical venous catheter tip below the diaphragm at the level of the T10 vertebra.
On day 4 of life, the baby suddenly worsened—she developed signs of poor perfusion and had gasping respiration. She was intubated, and started on ventilator support and inotropes. However, she continued to deteriorate rapidly. As the heart sounds were inaudible, cardiac arrest was suspected and the baby was given cardiopulmonary resuscitation with two doses of epinephrine.
Bedside echocardiogram showed a huge pericardial effusion causing cardiac tamponade. Emergency pericardiocentesis was performed and 40 mL of milky white fluid aspirated. Chest radiograph, which had been taken before the CPR, showed a huge cardiomegaly and right-sided pleural effusion (figure 8).
Figure 8.
Radiograph (case 3, day 5) showing huge cardiomegaly and right-sided pleural effusion.
After pericardiocentesis, the baby started improving. Repeat chest X-ray carried out after a few hours showed a normal sized heart shadow and persisting right pleural effusion (figure 9), which was left to resolve spontaneously. Pericardial fluid analysis showed glucose of 1324 mg/dL, triglyceride of 197 mg/dL, protein of 0.2 g/dL and 40 cells/mm3, suggesting that it was extravasated TPN fluid. The UVC was removed; inotropes and ventilator support were given for 48 h and stopped. The baby's further hospital stay was uneventful and she was discharged on day 20 of life.
Figure 9.
X-ray (case 3, day 5) after pericardiocentesis showing normal sized heart shadow and persisting right pleural effusion.
Video 1.
Ultrasonography (case 1, day 4) showing a hypoechoic mass with septations and hyperechoic rim in the liver parenchyma next to the ductus venous and the tip of the umbilical venous catheter within the mass.
Video 2.
Echocardiography (case 2, day 5) showing a thrombotic 7×5 mm mass inside the left atrium. A track formed by the catheter is seen within the mass after removal of the umbilical venous catheter.
Video 3.
Repeat echocardiography (case 2, after 10 weeks) showing complete resolution of the thrombus.
Discussion
UVCs have been used in NICUs for more than 60 years. Though UVC insertion is a relatively safe procedure, there are many reports of cardiac, pulmonary and hepatic complications.1–3 The first step to avoid such complications is to ensure appropriate placement of the tip of the UVC—either at the IVC-right atrial junction or in the intrathoracic portion of the IVC.4 However, there are several obstacles in ensuring the correct placement.
UVC insertion is a blind procedure, performed, in most centres, by junior residents. The length of insertion is usually decided based on the birth weight or the shoulder umbilical length.5 6 Often, this estimate is incorrect, resulting in mal-positioning of the catheter. Moreover, at the time of insertion, absence of free backflow of blood at the correct length of insertion results in the catheter being pulled back and fixed at the length, where good backflow of blood is present. This results in inappropriate low positions of the catheters in the vessels inside the hepatic parenchyma. Sometimes, the catheter fails to enter the ductus venosus from the left portal vein, the most difficult site for negotiation. This results in the catheter tip being placed below the level of the ductus venosus, which in turn can cause the tip, in due course, to migrate into one of the portal vein branches, resulting in extravasation and hepatic necrosis because of the blood flow towards the liver in the portal system.
If the UVC tip is positioned inside the liver, the vessel wall and hepatic parenchyma may, with the infusion of hyperosmolar fluids and with repeated flushing of the catheter, get eroded, resulting in extravasation of TPN or haematoma formation. Then the accumulated blood and/or TPN fluid takes the path of least resistance, tracks through the liver parenchyma and reaches the surface, to form a subcapsular collection, which then ruptures to produce ascites.
In most of the previous cases reported in the literature, the hepatic lesions presented at 6–9 days after UVC insertion, with signs of clinical deterioration.3 7 The common signs are abdominal distension, anaemia, shock, hypotension and abdominal compartment syndrome, secondary to subcapsular collections and ascites. Some cases have required paracentesis or even surgical drainage. In our case, the intrahepatic haematoma was found incidentally and serendipitously on day 3 after catheter insertion, when the haematoma was small and localised. As the catheter was removed, further complications were avoided.
The hepatic lesion seen in our case was similar to that described in the previous series—a hypoechoic mass with an echogenic rim.7 In our case, as the catheter was in situ, the tip of the catheter could be demonstrated within the lesion. In each of the cases reported in the literature, the UVC tip was found to be low, below the diaphragm and within the liver shadow in the radiograph, as in our cases.3 7 This proves that a low position of the catheter tip inside the liver is an important contributing factor for hepatic complications of UVC.
The Dunn and Shukla methods used for calculating the length of insertion have not been validated in preterm and low birthweight infants. Studies have shown that these methods result in intracardiac positions of the UVC tip especially in extremely low birthweight infants.8 Moreover, with lung expansion and diaphragmatic descent, the IVC-right atrial junction moves downward, resulting in the catheter tip moving upward. The direction and force of blood flow may push the catheter tip into the right atrium and further into the left atrium through the foramen ovale. There are some reports that an UVC tip in the right atrium may be acceptable, though this is not uniformly accepted. However, the catheter tip in the left atrium is definitely contraindicated. In a previous study, of the 31 cases with UVC tip in the left atrium, 8 babies (26%) developed thrombosis.9 The catheter tip seen inside the left atrium in the echocardiography was correlated with the tip above the T7 vertebra in the chest radiograph in the previous study. In the case of our second baby as well, the UVC tip was found at the T7 vertebra in the radiograph taken when the left atrial thrombus was diagnosed (figure 3).
Factors contributing to thrombosis are mechanical trauma to the wall of the blood vessel or the cardiac chamber, damage caused by the infused hypertonic fluids, disrupted blood flow and thrombogenic catheter materials. The risk of thrombus formation increases significantly with the UVC tip inside the heart. This is due to increased movement of the catheter tip secondary to cardiac contractions and flow disturbances around the tip inside the cardiac chambers, unlike the laminar blood flow in the superior vena cava or IVC.
Management of catheter-related thrombosis includes removal of the catheter and a course of anticoagulant, with thrombolytic therapy in exceptional cases. Left atrial thrombus is a good source of systemic emboli, which may be catastrophic. In our second case, the serendipitous diagnosis led to prompt removal of the catheter and institution of anticoagulant therapy. Even though the anticoagulant therapy was not completed, in follow-up, the thrombus was found to have resolved spontaneously.
The most common presentation of pericardial effusion due to UVC is acute cardiac collapse requiring resuscitation, as in our case.1 In almost all the cases, the diagnosis was retrospective, as the pericardial fluid analysis confirmed that it was the extravasated TPN fluid. In most cases, the effusion occurred later, in 3–12 days, probably secondary to erosion by the infused hyperosmolar fluid or migration of the catheter through the vessel wall or myocardium. However, the effusion may occur on the same day of insertion due to a mal-positioned catheter.10
Echocardiography has been found to be superior to abdominal radiograph in confirming the position of the UVC tip in all the studies.11 12 Michel et al11 found that the sensitivity and specificity of radiography were only around 60%, whereas those of echocardiography were more than 90%. In another study, echocardiography showed that the UVC tip was located in the appropriate position in only 23% of the cases.12 Hence, the recent consensus is that echocardiography should be considered to confirm correct placement of UVCs. A recent study reported that echocardiography carried out by trained house-staff paediatricians to determine the UVC tip position had a high accuracy rate with an area under the receiver operating characteristic curve of 0.81, when compared with scans performed by a cardiologist.13 Hence training treating physicians and house-staff paediatricians in—and making them aware of the need for—echocardiography to confirm UVC placement is not only more feasible but is also reliable.
Our cases and all the previous reports emphasise the fact that a single radiograph taken soon after the insertion of the UVC is not adequate to ensure the safety of the catheter. The confirmation of the catheter tip position by echocardiography is important. Moreover, repeated echocardiographic examinations at regular intervals, say once every 2–3 days for as long as the UVC is in place, can be made routine in NICUs. Though it sounds impractical in a busy unit, with training and experience, the screening can be carried out faster and it can be incorporated into daily NICU work.
Most of the delay-induced complications of UVCs reported in the literature have been shown to occur in the second week after catheter insertion. Hence, even though UVC use of up to 14 days is recommended, limiting this duration may help to decrease complications.
To conclude, the three important steps taken to minimise UVC-related complications are: ensuring appropriate position of the catheter tip at the IVC-right atrial junction or in the intrathoracic IVC at the time of insertion, by radiography and echocardiography; limiting the duration of UVC use; and repeated screening by echocardiography at regular intervals until removal of the catheter, to ensure that the UVC is maintained in the proper position and for the early detection of complications.
Learning points.
Since in most centres abdominal radiography soon after the insertion of an umbilical venous catheter (UVC) is the only investigation carried out to ensure the appropriate position of the catheter, many complications of the UVCs are diagnosed late.
Even if the placement of UVCs is appropriate at the time of insertion, the catheter tip tends to migrate in due course, resulting in mal-positioning and life-threatening complications.
Screening by echocardiography at regular intervals until removal of the catheter to ensure that the UVC is maintained in the proper position can be performed for the early detection of catheter-related complications.
Footnotes
Contributors: TA and MPS were involved in the conceptualisation of the manuscript, collecting patient data, conducting the literature search and drafting the manuscript. MK and NT supervised the data collection, helped in the literature search and revised the manuscript for scientific content. All the authors were involved in clinical management of the patient.
Competing interests: None declared.
Patient consent: Obtained.
Provenance and peer review: Not commissioned; externally peer reviewed.
References
- 1.Nowlen TT, Rosenthal GL, Johnson GL et al. Pericardial effusion and tamponade in infants with central catheters. Pediatrics 2002;110:137–42. 10.1542/peds.110.1.137 [DOI] [PubMed] [Google Scholar]
- 2.Björklund LJ, Malmgren N, Lindroth M. Pulmonary complications of umbilical venous catheters. Pediatr Radiol 1995;25:149–52. 10.1007/BF02010332 [DOI] [PubMed] [Google Scholar]
- 3.Hagerott HE, Kulkarni S, Restrepo R et al. Clinical-radiologic features and treatment of hepatic lesions caused by inadvertent infusion of parenteral nutrition in liver parenchyma due to malposition of umbilical vein catheters. Pediatr Radiol 2014;44:810–15. 10.1007/s00247-014-2895-2 [DOI] [PubMed] [Google Scholar]
- 4.Paster S, Middleton P. Roentgenographic evaluation of umbilical artery and vein catheters. JAMA 1975;231:742–6. 10.1001/jama.1975.03240190046020 [DOI] [PubMed] [Google Scholar]
- 5.Dunn PM. Localization of the umbilical catheter by post-mortem measurement. Arch Dis Child 1966;41:69–75. 10.1136/adc.41.215.69 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Shukla H, Ferrara A. Rapid estimation of insertional length of umbilical catheters in newborns. Am J Dis Child 1986;140:786–8. [DOI] [PubMed] [Google Scholar]
- 7.Lim-Dunham JE, Vade A, Capitano HN et al. Characteristic sonographic findings of hepatic erosion by umbilical vein catheters. J Ultrasound Med 2007;26:661–6. [DOI] [PubMed] [Google Scholar]
- 8.Harabor A, Soraisham A. Rates of intracardiac umbilical venous catheter placement in neonates. J Ultrasound Med 2014;33:1557–61. 10.7863/ultra.33.9.1557 [DOI] [PubMed] [Google Scholar]
- 9.Raval NC, Gonzalez E, Bhat AM et al. Umbilical venous catheters: evaluation of radiographs to determine position and associated complications of malpositioned umbilical venous catheters. Am J Perinatol 1995;12:201–4. 10.1055/s-2007-994452 [DOI] [PubMed] [Google Scholar]
- 10.Abiramalatha T, Kumar M, Shabeer MP. Pleural effusion caused by a malpositioned umbilical venous catheter in a neonate. BMJ Case Rep 2015;2015:pii: bcr2015212705 10.1136/bcr-2015-212705 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Michel F, Brevaut-Malaty V, Pasquali R et al. Comparison of ultrasound and X-ray in determining the position of umbilical venous catheters. Resuscitation 2012;83:705–9. 10.1016/j.resuscitation.2011.11.026 [DOI] [PubMed] [Google Scholar]
- 12.Ades A, Sable C, Cummings S et al. Echocardiographic evaluation of umbilical venous catheter placement. J Perinatol 2003;23:24–8. 10.1038/sj.jp.7210851 [DOI] [PubMed] [Google Scholar]
- 13.Pulickal AS, Charlagorla PK, Tume SC et al. Superiority of targeted neonatal echocardiography for umbilical venous catheter tip localization: accuracy of a clinician performance model. J Perinatol. 2013;33:950–3. 10.1038/jp.2013.96 [DOI] [PubMed] [Google Scholar]