Department of Anaesthesia, Critical Care and Pain Management, Leicester Royal Infirmary, University Hospitals Leicester, Leicester LE1 5WW, UK*Corresponding author
Accepted for publication: April 4, 2002
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
Br J Anaesth 2002; 89: 51719
Key words: airway, obstruction; complications, mediastinitis; gastrointestinal tract, oesophagus
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Case report |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
Her respiratory function deteriorated (PaO2 9.6 kPa, FIO2 1.0) and she was transferred to the intensive care unit for tracheal intubation and artificial ventilation. She developed systemic sepsis, which was considered to be caused by a hospital-acquired pneumonia. Epinephrine and dobutamine were started to maintain her arterial pressure and urine output. A pulmonary artery flotation catheter was sited, and the initial measurements were: cardiac output was 5.8 litre min1, systemic vascular resistance was 797 dyne s1 cm5, pulmonary artery capillary wedge pressure was 20 mm Hg, and central venous pressure was 10 mm Hg.
She received artificial ventilation with BIPAP (bilevel positive airway pressure) with an inspiratory pressure (Pinsp) of 24 cm H2O and PEEP (positive end expiratory pressure) of 5 cm H2O. Therapy with ciprofloxacin was commenced in addition to cefuroxime and metronidazole. Daily chest x-rays revealed persistent changes in both bases, but no suggestion of mediastinal collection.
On the tenth postoperative day a sudden decrease in minute volume with reduced delivered tidal volume was noted. Initially it was thought that the tracheal tube might have been dislodged or displaced against the tracheal wall or above the vocal cords. A bougie was inserted through the tracheal tube and the tube advanced further down to 24 cm from the lips. Because ventilation improved, a check laryngoscopy was not performed and other diagnoses, such as tension pneumothorax seemed unlikely. However, shortly afterwards difficulties recurred with inadequate ventilation despite increasing the inspiratory airway pressure.
An urgent fibreoptic bronchoscopy was done, where it was seen that the tracheal tube had moved back to the original position of 22 cm and the trachea was swollen and inflamed with complete circumferential occlusion at that level. The tracheal rings were not visible, the posterior tracheal wall was bulging at 24 cm, and the carina looked swollen with marked oedema. The tracheal tube was advanced to just above the carina, which allowed satisfactory ventilation with the previous airway pressures.
A chest x-ray now showed widening of the upper mediastinum. A contrast-enhanced CT scan (Fig. 1) showed a large low attenuation lesion containing pockets of gas seen in the prevertebral region of the neck and upper chest, extending on either side of the trachea from the superior cornua of the thyroid to the carina. The oesophagus and trachea were displaced anteriorly.
|
Her gas exchange remained poor.
A tracheostomy was performed on day 14 to improve weaning and over the next 3 weeks her condition improved. All the drains except one of the mediastinal drains were removed. After 24 days of intensive care she was transferred to the ENT ward. Total oesophagectomy was considered, but with conservative treatment a controlled oesophago-cutaneous fistula developed. Her condition gradually improved. After a further 2 months she was able to eat and mobilize, and was discharged home.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
Oesophageal perforation can occur spontaneously with vomiting or retching, or after instrumentation. It is estimated that the incidence is approximately 1 in 100 rigid oesophagoscopies.2 The commonest early symptoms are acute chest pain, epigastric pain, and dyspnoea. Signs include surgical emphysema, or signs of pneumothorax or mediastinal air. The diagnosis is usually confirmed with chest x-ray and contrast swallow.
The management of oesophageal perforation depends on the level and size of the perforation.2 With instrumental perforations this is usually conservative, with complete fasting, nasogastric suction, and broad-spectrum antibiotics, if the diagnosis is made before mediastinal contamination has occurred. 3 4 Surgery is indicated for large free lower oesophageal perforations4 or complications, for example mediastinitis.5 In the case reported here, the exact level of perforation was not clear but was in the upper oesophagus, and she initially progressed satisfactorily with conservative treatment. Signs of systemic sepsis developed, with respiratory failure and cardiovascular instability. At that stage, she was thought to have developed a nosocomial pneumonia, but we were concerned that a mediastinal infection may be developing.
The sudden onset of airway obstruction was first thought to be related to problems with the tracheal tube position. After recurrence of the obstruction, it became clear that this was caused by extrinsic tracheal compression, which occurred suddenly when the extrinsic intrathoracic pressure reached a level that caused the trachea to collapse. Artificial ventilation with BIPAP, where Pinsp was constant, meant that once the extrinsic pressure exceeded the Pinsp, the ventilation became ineffective. In addition, the mediastinal abscess had probably been developing for a few days, causing some inflammation and weakening of the tracheal wall. The chest x-ray appearances did not change during this period, and the tracheal tube was effectively stenting the upper trachea, so the possibility of sudden airway obstruction was unsuspected.
This case shows that life-threatening distal tracheal obstruction can occur suddenly, and may be undetected by serial chest x-rays. Furthermore, the obstruction developed despite an appropriately positioned tracheal tube. Conventional teaching is that if a tracheal tube is obstructed it should be removed, but in this case, removal of the tube is likely to have made ventilation more difficult or impossible.
During the conservative management of oesophageal perforation, the threshold for performing a CT scan of the chest should be low. In this case, earlier imaging would probably have shown the mediastinal collection, and allowed drainage before the subsequent airway obstruction occurred. However, the index of suspicion was not high and although CT scanning was considered, her poor cardio respiratory state made us reluctant to move her for a chest CT scan.
An alternative could have been regular inspection of the trachea and carina by fibreoptic bronchoscopy, although this carries risks in an unstable patient, and it is unclear at what stage early tracheal inflammation or compression would have been detectable.
We suggest that in a patient with mediastinal pathology, the possibility of airway obstruction should always be considered, even in the presence of an appropriately placed tracheal tube. Intensive care patients are very often dependent on PEEP and high FIO2, and investigation may be difficult. The axiom of changing a tracheal tube first in the setting of acute airway obstruction is not always safe. In this case, fibreoptic bronchoscopy was very useful to determine the cause, and plan subsequent management.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Smith CCK, Tanner NC. The complications of gastroscopy and oesophagoscopy. Br J Surg 1956; 43: 396403[ISI]
3 Shaffer HA, Valezuela G, Mittal RK. Esophageal perforation. A reassessment of the criteria for choosing medical or surgical therapy. Arch Int Med 1992; 152: 75761[ISI]
4 Lyons SL, Seremetis MG, deGuzman VC, Peabody JW. Ruptures and perforations of the esophagus: the case for conservative management. Ann Thoracic Surg 1978; 25: 34450
5 Triggiani E, Belsey R. Oesophageal trauma: incidence, diagnosis, and management. Thorax 1977; 32: 24149[Abstract]