Umbilical venous catheterisation: emergency central venous access which saves lives in coarctation of the aorta

Abstract

We describe a 9-day-old baby with coarctation of the aorta who required urgent resuscitation including intubation and cardiac compressions. Despite the commencement of prostaglandin E1 (PGE₁) to reopen the ductus arteriosus via the intraosseous route, postductal saturations remained unrecordable for a further 45 min. Within 3 min of administration of PGE₁ via an umbilical venous catheter (UVC), saturations were recordable at 92%. UVC access was the sentinel intervention that irrevocably altered the clinical prognosis. This baby boy has survived with excellent neurodevelopmental outcome. Clinicians are less familiar with UVCs outside of the newborn period. Our data demonstrate successful placement in neonates up to 28 days of age. We hope this case encourages clinicians to consider the UVC as first-line central venous access in collapsed neonates. In cases of suspected left heart obstruction, we argue that UVCs are the optimal route.

Background

The collapsed neonate represents one of the highest stakes resuscitation situations. The potential underlying pathologies are broad and prompt, appropriate clinical management can be the difference between life and death. In situations where systemic circulation is compromised, central vascular access is required to enable rapid administration of intravenous life-saving medication, including inotropes and prostaglandin E₁ (PGE₁) for ductus-dependent cardiac lesions. Although peripheral venous access can be used, poor systemic circulation limits successful peripheral cannulation. Intraosseous access (IO) provides alternative access, delivering therapy via the bone marrow.

In the context of critical duct-dependent cardiac lesions such as coarctation of the aorta, where the ductus has closed and the infant presents in cardiogenic shock, the perfusion to postductal limbs (lower limbs±left upper limb) is severely impaired. The distribution of a drug administered via these limbs—whether by peripheral venous access or via IO in the bone marrow—can be severely limited, impairing the drug delivery and its life-preserving potential. There is no literature exploring efficacy of drug distribution in this context; however, if the flow in the arteries supplying the postductal limb is obstructed it is biologically plausible that any drug administered postductally would be poorly distributed. In the context of a duct-dependent critical coarctation of the aorta, it is essential to promptly re-establish patency of the ductus arteriosus by administering PGE₁ to re-establish systemic perfusion. Administration of PGE₁ to postductal ischaemic limbs can be expected to have little, to no, effect. In contrast, catheterisation of the umbilical vein and thence the ductus venosus enables direct delivery of medication to the right atrium. Functional closure of the ductus venosus occurs within minutes of birth. Permanent structural closure starts within days after birth and is completed by one to 1–3 months of age and this delay allows for life-saving reopening up to one month of postnatal age.

We share this case report to illustrate the benefit of umbilical venous catheters (UVCs) for central access in the collapsed neonate. In this case, the establishment of UVC access for PGE₁ administration dramatically changed the clinical course and the outcome for this 8-day-old neonate who presented in extremis with critical duct-dependent coarctation of the aorta and cardiogenic shock.

UVC placement is common in a neonatal setting and is a procedural skill which is taught on the advanced paediatric life support course. We recognise that clinicians working in an emergency medicine or paediatric intensive care setting are less familiar with this procedure. In addition, most clinicians will consider UVCs to be an option in a newly born infant only and are not necessarily aware that UVC placement can be successful in an older neonate (up to 28 days of age).

We hope that this case encourages clinicians to consider UVC central access in the collapsed neonate.

Case presentation

Baby A was born at full term via spontaneous vaginal delivery after an uncomplicated pregnancy with normal first trimester screening and normal 20-week fetal anomaly scan. Birth weight was 3755 g (79th centile). Day 2 newborn check revealed effortless tachypnoea and baby A was admitted to the neonatal high-dependency unit for observation and empiric antibiotic therapy for 48 hours; he was discharged home on day 5 of postnatal age. On day 9, his parents sought medical review via the emergency department for a 2-day history of poor feeding and progressive lethargy.

At triage, he was noted to be in shock, triggering immediate assembly of a resuscitation team led by an experienced senior emergency physician. He was spontaneously breathing, dusky and pale. Central capillary refill time was 6 seconds with weak femoral pulses, postductal oxygen saturation (SpO₂) was 91%. Right upper limb blood pressure was hypertensive 104/70 mm Hg (mean arterial pressure 81 mm Hg) and blood pressures on remaining limbs were unrecordable. He was floppy and responded only to a painful stimulus. Bedside glucometer blood sugar measurement was‘unrecordable’. Peripheral intravenous (24 gauge) cannulation was unsuccessful and bilateral tibial IO (15 gauge) lines were inserted. Blood cultures were collected. Initial resuscitation included an intravenous 2 mL/kg bolus of 10% glucose for hypoglycaemia, intravenous 50 mg/kg cefotaxime and a total of 30 mL/kg of 0.9% sodium chloride fluid resuscitation. Chest X-ray demonstrated significant cardiomegaly and plethoric lung fields (figure 1). Femoral pulses became impalpable and he was assessed as being in cardiogenic shock with a working diagnosis of a ductal dependent cardiac lesion.

Figure 1.

Chest X-ray on day 8 demonstrating cardiomegaly and plethoric lung fields.

The retrieval team was activated and assistance sought from the onsite neonatal unit.

While the team were preparing to intubate and commence epinephrine and PGE₁ infusions, baby A suffered pulseless electrical activity/bradycardic arrest. He was intubated and ventilated in FiO₂ 100%, chest compressions were commenced and four doses of epinephrine (10 µg/kg) were administered via the IO line, followed by an infusion of epinephrine. CPR was continued for 26 minutes until return of spontaneous circulation was achieved. Baby A was grey and lifeless.

The retrieval team commenced PGE₁ at 50 ng/kg/min via the IO. Antibacterial coverage was broadened with the addition of meropenem and gentamicin. Femoral pulses remained impalpable despite commencement of PGE₁ via the IO. Preductal and postductal oxygen saturations were unrecordable.

The retrieval consultant encouraged the onsite neonatologist to insert an urgent UVC. A 4Fr (equivalent to 18 gauge) UVC was easily passed into the umbilical vein, with aspiration of blood and flush without resistance. A PGE₁ infusion was commenced via the UVC with cessation of the IO infusion, with an initial bolus to account for catheter dead space. Within 3 minutes of commencement of the UVC PGE₁ infusion, postductal oxygen saturations became detectable as 92%, 70 min after arrest.

Initial blood gas collected from the UVC demonstrated a profoundly acidotic pH of 6.45 with severe mixed metabolic and respiratory acidosis: pCO₂ 92 mm Hg, HCO₃ 6 mmol/L, BE −28.7 and lactate 20 mmol/L. He was hyperkalaemic with a K⁺ of 8.3 mmol/L. He was coagulopathic with an INR of 9.6, PT 128 s, APTT>200 s and fibrinogen 0.3 units/L.

He received intravenous calcium gluconate to treat hyperkalaemia, sodium bicarbonate to treat acidosis and hyperkalaemia, fresh frozen plasma (20 mL/kg) and cryoprecipitate (1 unit). He was moved by the retrieval team to the cardiac neonatal intensive care unit. Venous gas prior to transfer was improved: pH 6.98, pCO₂ 52 mm Hg, HCO₃ 12 mmol/L, BE −19 and lactate 15.7 mmol/L. Oxygen saturations deteriorated en route and preductal and postductal oxygen saturations were 83% and 60%, respectively, in 100% inspired oxygen on admission. He was immediately commenced on inhaled nitric oxide at 20 parts per million with improvement in his preductal oxygen saturations to 100%.

He had evidence of multiorgan failure on admission: anuria with a creatine of 136 μmol/L, urea of 6.7 mmol/L and K⁺ of 5.0 mmol/L, liver impairment with AST 1215 U/L and ALT 532 U/L, improving coagulopathy (INR 1.8, fibrinogen 1.2 g/L, APTT 56.4 s), thrombocytopaenia (platelets 83×10⁹/L) and electrolyte disturbances.

Investigation

Echocardiogram performed on admission revealed severe coarctation of the aorta (figure 2A, B) and dampened velocity in the postcoarctation descending aorta (figure 2) with a bicuspid aortic valve and biventricular dysfunction with pulmonary hypertension.

Figure 2.

(A) High parasternal view demonstrating the pulmonary artery (PA) and descending aorta (DAo). The white arrow depicts the juxtaductal coarctation with posterior indentation of the DAo. (B) The same parasternal view with colour flow Doppler. Aliasing of flow can be seen at the coarctation site. (C) A significantly dampened pulsed wave Doppler trace in the DAo below the level of the coarctation.

Amplitude integrated electroencephalography on day 11 demonstrated evidence of electrical seizures with a burst-suppression/discontinuous background. A formal EEG on day 12 demonstrated reduced amplitude bilaterally with slowing of the waveforms; suggestive of global encephalopathy, but no definite epileptiform discharges were seen. Preoperative MRI on day 13 demonstrated small bilateral infarcts and intraventricular and bilateral extra-axial intracranial haemorrhage.

Treatment

Baby A remained intubated and ventilated on iNO, sildenafil, inotropic support and PGE₁ at 50 ng/kg/min. He received broad empiric antibiotic treatment, electrolyte replacement, blood products and diuretics stabilise until operative repair. Creatinine peaked at 257 μmol/L on day 14 and improved to 168 μmol/L by day 16. The transaminitis peaked on day 10 with AST 3540 U/L and ALT 1091 U/L and resolved completely by day 14.

After stabilisation and management of his secondary multiorgan failure, baby A underwent repair of long segment coarctation involving the isthmus via thoracotomy on day 16 of postnatal life. He developed significant hypertension postoperatively, managed with sodium nitroprusside infusion and beta blockade once on enteral feeds. The multiorgan failure completely resolved in the postoperative period. He was able to be extubated on day 20. A progress echocardiogram on day 26 demonstrated a good repair with normal ventricular function. Diuretics were ceased on day 28 of age.

His postoperative course was complicated by a left vocal cord palsy secondary to coarctation repair and left-sided thoracotomy. At the time of discharge on day 34, small suck feeds via bottle were introduced with the remainder of the feeds via nasogastric tube. A hearing screen test was normal.

Outcome and follow-up

At discharge, his weight was 3.54 kg (4th centile), head circumference was 36.2 cm (16th centile) and length was 51.5 cm (4th centile).

At formal developmental assessment at 3 months of age, he was on full suck feeds and had been since 1 month post discharge. The Bayley III Scales of Infant Development scaled scores were within normal range for all domains: fine and gross motor, receptive and expressive language and cognitive skills.

Developmental assessment at 2 years of age using the Bayley Intravenous Scales of Infant Development found normal scores across all domains except for expressive language. He has therefore been referred for hearing assessment and speech therapy intervention.

Discussion

The literature around UVC access has concentrated on insertion technique, safe placement, insertion depth measurement and how to avoid infections. The role of the UVC in an emergency has been reported in case reports only. A recent review of the literature has revealed no studies comparing the efficacy of different routes of medication administration in neonatal resuscitation¹ and a German survey revealed a preference for IO access in the newborn in non-neonatologists² attributed to lack of experience with UVC placement. A simulation-based study demonstrated faster placement of IO compared with UVC in a neonatal resuscitation scenario and factors delaying UVC placement were noted to be perceived complexity of the procedure, lack of appropriate equipment and uncertainty regarding appropriate asepsis.³ There is a paucity of published research and audits evaluating the postnatal age limit for UVC access. Some protocols and guidelines state that the umbilical vein is available for access up to 7 days after birth or until the cord separates and many are silent on postnatal age limit. Step by step instructions on insertion of UVCs are addressed in local guidelines.⁴ Figure 3A illustrates the vascular anatomy of the umbilical cord and figure 3B illustrates a UVC inserted into the umbilical vein on day 9 of postnatal age. All treatments including inotropes and blood products may be administered via a UVC.

Figure 3.

(A) The vascular anatomy of the umbilical cord comprising two closely located umbilical arteries and one larger calibre umbilical vein. (B) A venous catheter inserted into the umbilical vein on Day 9 of postnatal age.

The Grace Centre for Newborn Intensive Care (GCNIC) is a referral unit for newborns with complex medical and surgical needs in New South Wales, Australia. Central access is often necessary in this patient cohort for life-saving medications such as antibiotics, glucose, inotropes or PGE₁.

An audit of UVC placements in neonates admitted to the GCNIC for the period 1 January 2008 to 30 June 2018 found that UVCs were sited in 658 (12.5%) of 5282 neonates. In 46.4% (n=305), the umbilical vein was accessed on the day of birth (day 0), in 38.7% (n=255) on days 1–3, in 10.5% (n=69) on days 4–7, in 3.6% (n=24) on days 8–13 and in 0.8% (n=5) UVC placements occurred between days 14 and 28 of life. The mortality of patients with umbilical access was 14% (n=93) compared with total overall mortality of 5.5% (n=292), which reflects the acuity of the underlying condition necessitating central UVC access. Two-thirds of the patients who received a UVC had a congenital cardiac defect (CCD) (441/658).

The underlying condition in patients who received a UVC after day 3 of life was predominantly cardiac, including CoA, arrhythmias, aortic stenosis and pulmonary stenosis. The proportion of patients with congenital cardiac conditions who were duct dependent and who had a UVC was 40% (261/658), 48% (33/69) of patients with a UVC sited between days 4 and 7, 75% of patients with a UVC sited between days 8 and 13 (18/24) and two out of five patients with a UVC sited beyond day 14 of life.

The majority of the group of infants who received a UVC in the second week of life were postnatally diagnosed with a duct dependent cardiac lesion after an out-of-hospital collapse. Extensive resuscitation was necessary in most of these patients. Survival of all patients with duct dependent CCD managed using a UVC was 87.4% and the survival rate of infants with late insertion of UVC (after day 3 of life) with a duct dependent cardiac lesion requiring PGE₁ was 88.7% (47/53). This is comparable to reported survival data for critical CCD in the USA of 82.5%.⁵

There are neither animal or human data comparing the relative effect of PGE₁ administered via a UVC and an IO nor is there likely to be any head-to-head pharmacokinetic trial. However, there is no doubt in the minds of the clinicians who cared for baby A that UVC administration of PGE₁ resulted in a prompt clinical response: establishing patency of the functionally closed ductus arteriosus and re-establishing systemic perfusion. With the initial postarrest gas demonstrating a pH of 6.45 and lactate of 20 mmol/L, cardiogenic shock and multiorgan failure, his prognosis was very guarded and one would be forgiven to think unsurvivable. But he did survive and is now a thriving 2 year old.

We hope that this case and our experience demonstrating successful placement of UVCs up to 28 days postnatal age encourages clinicians to consider UVC access in the emergency and retrieval setting. We recognise that the lack of experience of non-neonatal clinicians with UVC placement may pose a barrier to the use of UVCs as first-line venous access. Use of UVCs will require a commitment to upskilling and having preassembled UVC equipment kits readily available. We believe UVC access has the potential to save lives in critically unwell neonates.

Patient’s perspective.

To have waited for the arrival of our baby boy for 9 months, to hold him in our arms and then for him to be separated from us to battle between death and life has definitely been the worst thing we have ever experienced.

The first time our baby boy was taken away from us was on day 2, when he was admitted to the neonatal low dependency unit. I dreamt of a voice which whispered in my ear… ‘It’s the heart!’ Dismissing my nightmare, the doctors told us that our baby boy was well and he was discharged home. We did not know then that the real nightmare was going to start a few days later.

The second time our baby boy was taken away from us was the following Friday (at just 8 days). He had had little interest in breast feeding or bottles for the previous 2 days. We thought it was due to the milk formula.

Early on that Friday, he seemed lethargic and pale. We gave him a different formula, bought at 5:00. He finished the bottle and even had a little more and we were reassured. But suddenly, he started to breath rapidly as if he had run a marathon. I looked into his eyes and his eyes told me he was dying, so we rushed to the hospital.

We arrived to the emergency department and as soon as our baby boy was triaged an alarm was activated. Eighteen clinicians converged around him, so many doctors and nurses…. you can tell something is definitely wrong when you see so many people helping one little baby.

After a few hours, the team explained that they had done all they could there and that a specialised retrieval team, the newborn emergency transport service (NETS), was going to transfer our baby boy to the children’s hospital for specialist paediatric cardiac assessment and surgery.

The journey from day 1 was very hard and uncertain; it felt totally overwhelming. He even had to wait to have his surgery done because he was so ill. For every milestone our baby boy reached, there were more obstacles he had to face. And when those were achieved, there were further challenges he had to overcome, even once home.

We remember two very special moments of hope and comfort: seeing our baby boy open his eyes when he heard our voice after a long period of sedation and the second one, holding him in our arms skin to skin after being in intensive care for a few weeks.

We met several doctors in all their specialties, nurses and hospital staff, whose faces we still remember. We remember their empathy and compassion for us, except one. These clinicians mirrored our pain while trying to do their best, doctors with watery eyes or with a lump in their throat saying our baby boy was very ill. To all of these clinicians, we are really grateful for your empathy.

An example of this empathy was when one of the team gave us a bit of hope saying, ‘Which sports team will he be playing for… when he gets older?’, while another was as cold as ice and said several times ‘he might die’. A very special thanks to the first one because he really tried to cheer us up even when we knew he was telling us this only to comfort us and to not lose hope in our hearts when everything indicated otherwise. To the second one, we would suggest enrolling in a course on ‘how to show compassion when dealing with patients battling between life and death’. Do not get us wrong, we cannot tell how many times during the journey we heard ‘he has little chance of survival’, ‘he might die’ and that ‘he might not make it through the night’, but the rest of the team were compassionate and empathic, even though they did not know we were suffering and dying inside from so much pain.

Having said the above, we want to thank God and all the doctors, nurses and staff from the ED at the referring hospital, the NETS retrieval team and the Grace Centre for Newborn Intensive Care Unit at the Children’s Hospital at Westmead who intervened to save our baby boy’s life. Their collective knowledge, skills and techniques, such as the use of the umbilical venous catheter, brought him back from death to life.

We would like to thank all for not giving up on him even when the chances of survival were minimal and for trying everything to save his life, because you all worked together as a team and turned this distressing part of our lives into the joy of having our little one with us. To all of you, we will be eternally thankful. May God bless you for striving to save the lives of other babies.

Lastly, we would like to acknowledge our baby boy, who fought bravely. He is a little warrior and a beautiful, strong, healthy, intelligent and cheeky boy who completes our family. We love his smile. He is adored by his parents, siblings, grandparents and extended family and friends… and he knows it. We cannot be happier without him.

What we learnt is: life is so fragile and it can change from 1 day to another and that those little things and precious moments in life is what we need to treasure the most. We do not need to wait a whole year to celebrate life. Now, we celebrate life… DAILY.

Baby boy’s family

Learning points.

• Umbilical venous catheters (UVCs) have been successfully placed in neonates up to 28 days of age.

• UVC access enables large bore central access for the delivery of life-saving treatment.

• In the collapsed neonate with suspected coarctation of the aorta/left heart obstruction, UVC central access offers direct delivery of PGE₁ to the heart and ductus arteriosus.

• PGE₁ delivery via the intraosseous route to postductal limbs can be expected to be less effective in coarctation of the aorta as these limbs are poorly perfused.

• UVCs should be considered as first-line venous access in the collapsed neonate.

Ethics statements

Patient consent for publication

Consent obtained from parent(s)/guardian(s).