Secondary Logo

Journal Logo


Cardiac Arrest from Tension Pneumopericardium in a Premature Infant

Auden, Steve M., MD; Lapin, Sanford L., MD; Joseph-Reynolds, Ann Marie, MD

Author Information
doi: 10.1097/00000539-200211000-00025
  • Free

Pneumopericardium is a rare and potentially fatal event. We report a premature infant who sustained isolated tension pneumopericardium with cardiac arrest because of a preventable anesthetic mishap.

Case Report

A 2.03-kg, 68-day-old male twin infant, a product of 29 wk of gestation, came to the operating room (OR) for closure of an ileostomy. The ileostomy had been performed in conjunction with bowel resection 36 days earlier because of necrotizing enterocolitis. The perinatal course had also been significant for 13 days of mechanical ventilation and an additional 16 days of supplemental oxygen therapy for respiratory distress. There had been no significant barotrauma event. Before coming to the OR, the baby had been stable, with spontaneous respiration of room air for 4 weeks. Successful bowel reanastomosis and closure was performed under general anesthesia with fentanyl, isoflurane, and rocuronium. Endotracheal intubation had been uneventful, and intraoperative peak airway pressures had been between 20 and 26 cm H2O. The infant had tolerated the procedure well and had resumed spontaneous ventilation. The esophageal stethoscope had been removed, and the infant was being monitored with pulse oximetry and electrocardiogram on a transport monitor. During transfer from the OR table, and immediately after changing to cylinder oxygen supply, there was a sudden decrease in heart rate from approximately 160 to 65 bpm, and normal perfusion of the extremities decreased to little or no perfusion. Chest compressions were begun, and positive-pressure ventilation was resumed with 100% oxygen. Exhaled CO2 was <10 mm Hg. Epinephrine 20 μg IV was given, with no response. The 3.0-mm-internal-diameter endotracheal tube was removed and replaced, to no avail.

The anesthesiology resident who had been with the patient throughout these events suspected pneumothorax, and both hemithoraces were needled, with no result or response. Radiology was summoned. Epinephrine 100 μg IV and then 200 μg were given, still with no response from the patient. Chest compressions and ventilation continued, but without apparent perfusion of the baby. Intracardiac injection of epinephrine was elected. A left subxiphoid approach was made with a 21-gauge needle attached to a 10-mL syringe containing 4 mL of 1:10,000 epinephrine (400 μg). The needle was advanced toward the midline at approximately 30° from both horizontal and sagittal planes. Aspiration as the needle advanced suddenly yielded 6 mL of freely aspirated air. Instantly, the infant’s heart rate increased, and peripheral perfusion improved. Blood was never aspirated, and the epinephrine was not injected. A left pleural tube was placed, but no air returned. A chest radiograph was obtained and confirmed residual pneumopericardium without evidence of pneumothorax (Fig. 1). Oxygen saturation by oximetry was 100%, and heart rate and blood pressure recovered to normal. Within a few minutes, the infant opened his eyes, moved all extremities vigorously, and resumed spontaneous ventilation. No further inotropes were required.

Figure 1
Figure 1:
Chest radiograph reveals an endotracheal tube at the carina, a left pleural tube without evidence of pneumothorax, and pericardial air, best seen as a partial “halo” below and on the right side of the heart.

Additional fentanyl was given, and the infant was taken to the neonatal intensive care unit, where his trachea was extubated the next morning. Repeat head ultrasound was normal. The pneumopericardium resolved without further treatment over the next 2 days. The infant went home on postoperative Day 13 without apparent adverse sequelae.

A detailed review of this event was conducted as soon as the infant was stable. The anesthesia resident had planned to place a nasal cannula at the 15-mm endotracheal tube connector for the administration of oxygen en route to the recovery area (Fig. 2). This was common practice at that time in one of the adult hospitals in our teaching system, with the cannula serving as a source of “blow-by” oxygen. A nurse, also relatively new to the pediatric OR, began oxygen flow by cylinder at 4 L/min and placed the distal end, not of the nasal cannula, but of the oxygen tubing, into the 15-mm connector (Fig. 2). This fit all too snugly, and within seconds the described events had occurred. It was because of the incorrect use of the oxygen tubing and the temporal relationship to the patient’s deterioration that the anesthesia resident suspected a tension pneumothorax.

Figure 2
Figure 2:
The 15-mm connectors of two infant-sized endotracheal tubes are shown, one with the nasal cannula in place to provide blow-by oxygen and the other with oxygen supply tubing wedged into the connector.


Pneumopericardium has been reported in the anesthesia literature after tracheoplasty in a three-month-old infant (1), with high-frequency jet ventilation in a four-year-old child (2), and after severe coughing in a one-year-old child on a T piece (3). Other causes are more common at older ages. Trauma, infections, ulcers, and esophageal diseases are well described causes of pericardial air but rarely lead to cardiovascular compromise.

Pneumopericardium in an infant is an unusual event and usually occurs with other air leaks as a result of resuscitation, mechanical ventilation, severe lung pathology (4), or a combination of these. The etiology of this has been described by Hummler et al. In brief, “Dissection of air along perivascular and peribronchial sheaths leads to mediastinal … air entering into the pericardium … at the reflection of the pericardial membrane on the pulmonary vessels”(5). In the neonate, pneumopericardium can occur without pneumothorax (6), especially in the presence of significant lung disease.

Tension pneumopericardium, particularly isolated tension pneumopericardium with cardiac arrest, is a rare event with frequent mortality. Tension pneumopericardium in the neonate has mortality rates in the 72% to 83% range (7), but this mortality is multifactorial (8). Although the event is rare, it must be recalled in the differential diagnosis of neonatal cardiac arrest because survival requires prompt and specific treatment. Although radiograph often reveals the classic “halo” of air around the heart (8), waiting several minutes for such studies may not be acceptable. Without prompt evacuation of the pericardial air, prognosis is poor.

The decision to transport this neonate with the nasal cannula blow-by device was made by the anesthesia resident without attending physician input, applying a lesson learned in adult anesthesia methods to a small infant. Even had the device been applied as the resident desired, it would not have allowed for positive-pressure support of the airway. Small neonates with or without significant lung disease should be transported with the means to deliver oxygen via positive pressure. This neonate had both respiratory (history of chronic lung disease and major abdominal surgery) and central nervous system (apnea risk related to prematurity) indications to have positive-pressure airway support in place. Our usual practice with such infants had been and still is to transport either with an infant resuscitator bag (Ambu bag) or with a modified Jackson-Rees circuit.

Finally, makeshift medical devices are potentially dangerous and should be used only when no more proven option is available. Most medical devices are designed for a specific purpose. Even use of a well described modification, for example, using an embolectomy catheter as a bronchial blocker, is not without risk. Exceptions and innovations should be used only when necessary. Even then, they should ideally be discussed with colleagues and attempted only by the most experienced personnel. Such experiments are potentially catastrophic and should never be delegated to subordinates.


1. Sugimoto M, Shimaoka M, Hagihira S, et al. Tension pneumopericardium following tracheoplasty for congenital tracheal stenosis. Anaesth Intensive Care 1997; 25: 539–41.
2. Mikkelsen S, Knudsen KE. Pneumopericardium associated with high-frequency jet ventilation during laser surgery of the hypopharynx in a child. Eur J Anaesthesiol 1997; 14: 659–61.
3. Djaiani G, Major E. Pneumopericardium: an unusual cause for cardiac arrest. Anaesthesia 1998; 53: 580–8.
4. Itani MH, Mikati MA. Early onset neonatal spontaneous pneumopericardium. J Med Liban 1998; 46: 165–7.
5. Hummler HD, Bandstra ES, Abdenour GE. Neonatal pneumopericardium: successful treatment with nitrogen washout technique. J Perinatol 1996; 16: 490–3.
6. Varano LA, Maisels MJ. Pneumopericardium in the newborn: diagnosis and pathogenesis. Pediatrics 1974; 53: 941–5.
7. Heckmann M, Lindner W, Pohlandt F. Tension pneumopericardium in a preterm infant without mechanical ventilation: a rare cause of cardiac arrest. Acta Paediatr 1998; 87: 346–8.
8. Burt TB, Lester PD. Neonatal pneumopericardium. Radiology 1982; 142: 81–4.

Attention Authors!

Submit Your Papers Online

You can now have your paper processed and reviewed faster by sending it to us through our new, web-based Rapid Review System. Submitting your manuscript online will mean that the time and expense of sending papers through the mail can be eliminated. Moreover, because our reviewers will also be working online, the entire review process will be significantly faster. You can submit manuscripts electronically via There are links to this site from the Anesthesia & Analgesia website (, and the IARS website ( To find out more about Rapid Review, go to and click on “About Rapid Review.”

© 2002 International Anesthesia Research Society