Airway fire during awake tracheostomy using high‐flow nasal oxygen

Summary

High‐flow nasal oxygen is increasingly used in complex head and neck surgical procedures and difficult airway management. We describe a case where an operating room fire occurred while using high‐flow nasal oxygen during an awake tracheostomy for an obese patient in airway extremis due to supraglottitis. Shortly after the operation began, and before incision of the trachea, electrical diathermy applied to bleeding sub‐cutaneous vessels ignited a small flame. This was extinguished without harm to the patient and the procedure was completed without further complication. Fire requires three components: fuel; heat; and an oxidiser. We speculate that high‐flow oxygen channelled under the drapes and acted as the oxidiser; either tissue eschar or vapourised fat were the fuel; and the diathermy supplied a source of ignition to complete the fire triad. When using high flows of concentrated oxygen, practitioners should aim to minimise all of these factors and be alert for the risk of fire at every stage of the operation.

Introduction

High‐flow nasal oxygen has been used to facilitate oxygenation during emergency tracheostomy in awake and anaesthetised patients where other airway methods were deemed inappropriate or had failed ^{1, 2}. Although there have been warnings regarding the risk of operating room fires in airway surgery managed with high‐flow nasal oxygen ^{3, 4}, only one previous case report exists of an airway ignition whilst using high‐flow nasal oxygen ⁵, occurring in very different circumstances to this case. In this report, we present a case where an operating room fire occurred during an awake tracheostomy for an obese patient in airway extremis whilst oxygenation was provided using high‐flow nasal oxygen.

Report

A 49‐year‐old man with mild intellectual disability and a weight of 133 kg was brought to the emergency department after bystanders called for ambulance assistance. He had been found on a pavement in respiratory distress, only able to speak in single words and with marked tracheal tug. Paramedics made a provisional diagnosis of anaphylaxis, applied oxygen via facemask and administered intramuscular adrenaline, and nebulised salbutamol and adrenaline. In the emergency department the patient had stridor, a heart rate of 114 beats.min⁻¹, a blood pressure of 135/77 mmHg and arterial oxygen saturations of 92%. Airway assessment was unfavourable, with a Mallampati score of 4 and a bull neck. An ear, nose and throat (ENT) surgeon and consultant anaesthetist conducted a nasal endoscopy, showing crusted lesions on the posterior two‐thirds of the vocal cords and significant supraglottic erythema and oedema.

The glottic opening was deemed too small to intubate, and the patient was consented for an awake tracheostomy under local anaesthesia, with induction of anaesthesia planned after the surgical airway was secured. The anaesthetists involved in the case included a consultant experienced in ENT surgery and a fellow, both of whom had used high‐flow nasal oxygen for other ENT procedures, but not for awake tracheostomy.

The patient was transported to the operating table and positioned head‐up with standard monitoring applied. Optiflow^TM high‐flow nasal oxygen (Fisher & Paykel Healthcare, Auckland, New Zealand) was commenced at a flow rate of 30 l.min⁻¹ with an F_IO₂ of 1.0. The patient was lightly sedated with a total of 2 mg midazolam and 100 μg fentanyl. After the patient was appropriately sedated, oxygen flow rates were increased to 60 l.min⁻¹.

Before the operation began, the surgical plan and risks, including fire, were discussed with all theatre team members. To mitigate the fire risk, the surgeons would inform the anaesthetists before incising the trachea so that high‐flow nasal oxygen could be turned off.

Following routine procedures, the surgical field was prepared with alcohol‐free poviodine solution and draped. The single‐use adhesive drapes were elevated off the patient's face using drip stands. Surgery began with lignocaine 2% with adrenaline infiltration around the operative site and a scalpel incision to the skin. Electrical monopolar diathermy (ForceTriad^TM, Covidien, Lane Cove, NSW, Australia) was used to control bleeding vessels. Two minutes into the operation, during diathermy approximately 2 mm below the skin, a small 1 cm by 1 cm jet of flame ignited from the top of the incision, burning caudally. The surgeon ceased diathermy and simultaneously extinguished the flame with his hand. The anaesthestic fellow turned the oxygen off.

The patient was inspected for burn damage or inadvertent penetration of the trachea, with none found. He reported his respiratory distress was unchanged. After discussion amongst the team, high‐flow nasal oxygen was restarted at 30 l.min⁻¹ and surgery continued uneventfully. The oxygen was turned off before tracheal incision. Once placement of the tracheostomy tube was confirmed by capnography, general anaesthesia was induced with propofol and maintained with sevoflurane, with rocuronium administered for neuromuscular blockade. A panendoscopy showed severe supraglottitis and the patient was transferred to the intensive care unit for ongoing care.

In the intensive care unit, the patient was managed with broad spectrum antibiotics and dexamethasone. No underlying cause for the supraglottitis was found and it resolved over the following days.

Discussion

The prerequisites for any fire are the components of the fire triad: fuel; an oxidiser; and a source of ignition. Consequently, during high fire risk procedures such as head and neck surgery, it is the responsibility of all members of the theatre team to be alert to and minimise these components. Clear communication of potential fire risks and risk mitigation is a key aspect of patient safety when operating room fires may occur and should be discussed during the team brief and the ‘time out’.

Many possible fuels exist in operating theatres including: surgical drapes; sponges; gauze; the patient's hair or skin; plastic dressings; and instruments. Recognised common fuel sources in tracheostomy fires are tracheal tubes in intubated patients and skin preparation solutions in awake patients ^{6, 7}. It is important to ensure that as many potential fuel sources as possible are removed from the surgical site before operating. As the aqueous poviodine solution used by the surgeons in this case was alcohol‐free and non‐flammable, and there was no burn damage to the patient's skin or the surgical drapes, we suggest that vapourised subcutaneous fat may have acted as a fuel source for the fire. Alternatively, tissue debris could have accumulated on the end of the diathermy probe to become a flammable bead of eschar ⁸.

Potential ignition sources in surgical fires include laser and fibreoptic light sources, defibrillators and high‐speed drills. Diathermy is a well‐known ignition source, implicated in the majority of operating room and tracheostomy fires ^{6, 7}, and is the likely ignition source for the fire reported here. Diathermy types and modes affect fire risk, with cutting mode generating more heat than coagulation, and monopolar causing more electrical arcs than bipolar diathermy, but all have been implicated in tracheostomy fires ^{7, 9}. A dry surgical field is desired for a tracheostomy, so diathermy is usually a necessity, but some steps to decrease the ignition risk include: using the lowest energy possible; using diathermy as infrequently as possible; not using diathermy to incise the trachea; and using local anaesthetic with adrenaline to reduce bleeding and therefore the need for cautery. A harmonic scalpel may also be used instead of diathermy ⁹, which may negate the risk of ignition completely.

Oxygen delivered at high concentrations and high flows via high‐flow nasal oxygen is an obvious oxidising agent, completing the fire triad in this case. One other case report exists of a brief intra‐oral airway fire occurring during the use of both diathermy and high‐flow nasal oxygen, where an electrical arc connected the diathermy needle to titanium implants in the patient's mouth, resulting in a burn on the diathermy needle‐tip insulator, but no harm to the patient ⁵. Given that in our case the trachea was intact on inspection, we suspect that oxygen may have flowed under the drapes, channelled by a gap in the adhesive or fold in the material towards the surgical field. This can be seen in videos of mock operating room fires ¹⁰.

High‐flow nasal oxygen is a relatively new and increasingly popular tool in airway management, but having experienced an airway fire associated with its use we caution others to think carefully before using it in similar situations. In this case, we suspected that mask oxygenation would have been difficult due to the patient's facial features and thick neck, and that high‐flow nasal oxygen would be more successful. In other patients, we suggest first attempting oxygenation via facemask, as a good seal with lower flows will decrease the potential for oxygen to flow to the operating site. Furthermore, we offer the following additional recommendations to those seeking to use high‐flow nasal oxygen in awake tracheostomies:

1. While ignition can occur with F_IO₂ below 1.0 ⁹, lower oxygen concentrations will reduce the risk. Use an air‐oxygen blender and use the lowest fraction of oxygen to maintain acceptable oxygen saturations.

2. Use the minimum gas flow rates necessary.

3. If possible, cease the flow of oxygen for 1 min before ignition source activation.

4. Ensure drapes are meticulously sealed against the patient to direct oxygen flow away from the operating site.

With the increased use of high‐flow nasal oxygen, this case is a timely reminder of the risk of airway fires, the need to minimise their potential occurrence and the requirement for extreme vigilance and caution in high risk settings.