Sudden death from cardiac tamponade in an extremely low birth weight neonate with an umbilical venous catheter in situ

Introduction

The use of central venous catheters for intravenous feeding of neonates has become common practice in neonatal intensive care units [3]. It is of particular value in neonates considered to be “high risk” for feeding complications, such as necrotising enterocolitis (NEC), and those of extremely low birth weight (ELBW). Risks of central line insertion, such as infection and, more rarely, pericardial effusion, must be weighed against risks of alternative feeding methods and suboptimal nutrition. Indeed, there is evidence that early nutrition and growth influence long-term outcomes for ELBW neonates, making early central venous access and parenteral nutrition essential for these babies [5].

Pericardial effusion and cardiac tamponade are known to be rare but potentially fatal complications of umbilical venous catheterisation even when the line tip is appropriately placed, proximal to the right atrium [1, 2]. This case highlights the need for prompt recognition and management of this complication.

Presentation of the case

A baby was delivered at 30 weeks’ gestation because of intrauterine growth restriction and absent end diastolic flow on umbilical Doppler imaging. The baby was intubated at 10 min because of poor respiratory effort and received surfactant immediately. Birth weight was 860 g. The baby was routinely commenced on total parenteral nutrition (TPN) in 10% dextrose with 20% Intralipid via a Kendal Argyle 4 Fr double-lumen umbilical venous catheter (UVC) (Dublin, Ireland), according to unit protocol. An umbilical artery catheter (UAC) was inserted later. Line tip positions were assessed radiographically. It was noted that the UVC and UAC appeared to cross unusually, with the UVC ascending to the left of the UAC. No other imaging modalities were available at the time of insertion. UVC tip position was considered satisfactory at T8, and the UAC, thought to be too low, was withdrawn into a low position (Figure 1). Initial radiographs confirmed appropriate endotracheal tube (ETT) and nasogastric tube positions and revealed atelectasis/consolidation of the left lung field (Figure 1). An infection screen had been performed and antibiotics were commenced. Arterial blood gas parameters were within normal limits and adequate ventilation was maintained on pressure-controlled synchronised intermittent mandatory ventilation. The UAC sampled well but did not provide reliable blood pressure monitoring. Perfusion and urine output were within normal parameters.

Figure 1

Chest and abdomen radiograph at 4 h of age.

At 18 h of age the baby became acutely unwell and was poorly perfused with a capillary refill time longer than 5 s despite a normal arterial blood gas. A fluid bolus was given and dopamine was commenced. The baby soon became severely bradycardic and full cardiorespiratory resuscitation was initiated. It was noted that the baby’s abdomen was tense and distended at the time of deterioration. After responding to initial resuscitation, the baby underwent an urgent cranial ultrasound scan that did not reveal any large intracranial haemorrhage. Abdominal X-ray showed no evidence of bowel perforation or other signs of acute NEC (Figure 2). A radiograph of the chest confirmed appropriate ETT position with bilateral diffusely opacified lung fields. UVC position did not appear to have changed (Figure 2). The baby continued to deteriorate and, despite full resuscitation, died.

Figure 2

Chest and abdomen radiograph during resuscitation.

Postmortem examination revealed an appropriate UAC position. The UVC tip was noted to be impacted in the blind end of the orifice of the coronary sinus. The inner wall of the inferior vena cava, right atrium, and the orifice and blind end of the closed coronary sinus had no disruption, discontinuity or perforation. However, several U-shaped abrasions were found at the bottom of the closed coronary sinus with no obvious perforation, bleeding or tear. The UVC was well secured and could not move. Approximately 5 mL of lipid-rich fluid was found in the pericardial space (Figure 3), with a further 6–7 mL of the same fluid present in the pleural space surrounding the left lung (Figure 4). The abdomen was unremarkable.

Figure 3

Post-mortem image of pericardial space. Arrow indicates presence of lipid rich fluid.

Figure 4

Post-mortem image of left pleural space. Arrow indicates presence of lipid rich fluid.

Discussion

It is hypothesised that diffusion of hyperosmolar infusates can cause endothelial damage and may be the cause of TPN-related pericardial effusion where the central venous catheter is properly positioned and no evidence of trauma is found [2]. Although cardiac tamponade is a known complication of central venous lines, it is rare (incidence of 0.5%–2%) [3], and therefore, even in larger neonatal units, this complication will not often be seen. Prompt recognition and treatment using pericardiocentesis with/without chest drain placement is crucial [1, 7].

Pleural effusion may also result from transudation of hyperosmolar fluid from an appropriately placed UVC [1]. Alabsi [1] reported one previous case where prompt insertion of a chest drain was sufficient to treat both pleural and pericardial effusions. Other cases of survival after urgent pericardiocentesis are reported [7]. The presence of bilateral pleural effusions in association with central-line-related pericardial effusion has been rarely seen [4]. The presence of pleural effusions, in addition to pericardial effusion, in this case supports the hypothesis of transudation of fluid.

Decisions about use and type of UVC must be based on the balance of risks and benefits to individual patients. Studies suggest that using multiple-lumen UVCs reduces the need for further peripheral intravenous access in the first week of life, but further trials are required to study other aspects of safety [6]. Ultrasound has been shown to be more accurate than X-ray in assessment of UVC placement [8]. In this case, UVC tip position was satisfactory on X-ray but the apparent “crossing” of the UAC and UVC was difficult to explain. Ultrasound should be considered in clarifying UVC tip position where doubt exists on radiographic imaging [3, 8].

There are many potential causes of acute deterioration of a previously well neonate. In this case, pericardiocentesis was not performed. Emergency echocardiography may aid in identifying pericardial effusion in an acutely unwell neonate if this can be performed without delay. Clinicians should be aware that the radiographic appearance of bilateral diffuse opacification of lung fields may be an indicator of pleural effusions and should raise suspicion of cardiac tamponade. Pericardiocentesis should be considered in a preterm baby who deteriorates suddenly with a central venous catheter in situ.