Multispecialty tracheostomy experience

Abstract

Introduction

Tracheostomy is a common surgical procedure used to create a secure airway in patients, now performed by a variety of specialties, with a notable rise in critical care environments. It is unclear whether this rise is seen in units with large head and neck surgery departments, and how practice in such units compares with the rest of the UK.

Methods

A three-year retrospective audit was carried out between anaesthetic, surgical and critical care departments. All tracheostomy procedures were recorded anonymously.

Results

A total of 523 tracheostomies were performed, 66% of which were in men. The mean patient age was 60 years. The majority (83%) were elective, performed for various indications, while the remaining 17% were emergency tracheostomies performed for pending airway obstruction. A fifth of the tracheostomies were percutaneous procedures. Most emergency tracheostomies (78%) were performed by otolaryngology. Three cricothyroidotomies were performed within critical care and theatres. Complications related to tracheostomy occurred in 47 cases (9%), most commonly lower respiratory tract infection. The mean time to decannulation was 12.8 days.

Conclusions

This paper discusses the findings of a comprehensive, multispecialty audit of tracheostomy experience in a large health board, with over 150 tracheostomies performed annually. Elective cases form the majority although there is a significant case series of emergency tracheostomies performed for a range of pathologies. Around a quarter of those requiring tracheostomy ultimately died, mostly as a result of advanced cancer.

Introduction

Tracheostomy is a commonly performed multispecialty procedure used to create a secure surgical airway in patients with upper airway obstruction, excessive respiratory secretions or neurological impairment of the airway. It is performed in a controlled elective general anaesthetic (GA) environment, or under local anaesthetic (LA) in response to immediately life threatening pathology, most typically following failed endotracheal intubation.

Cricothyroidotomy involves stoma creation using needle or scalpel, through the skin and cricothyroid membrane, performed as an emergency airway procedure. Emergency front of neck access (eFONA) is regularly taught and practised by anaesthetists. It is indicated in anaesthetised and paralysed patients where intubation and oxygenation have failed. Cricothyroidotomy is not considered a definitive airway.

A variety of techniques are employed to create a tracheostomy. A midline transverse or vertical incision is made superior to the sternal notch through skin and subcutaneous tissues, with retraction of the strap muscles and retraction or division of the thyroid isthmus. The trachea is entered by creating a window either through multiple tracheal cartilage rings or between two rings to create the stoma, through which a tracheostomy tube is sited. Experienced aftercare ensuring correct placement and humidification of the trachea is crucial.

LA techniques are used in complex cases with subtotal airway obstruction and/or for those in whom GA poses an unacceptable risk. The most common indications for surgical tracheostomy (ST) include head and neck malignancy, infection, trauma and non-infectious swelling.

Critical care physicians desiring a surgical, reversible airway to facilitate prolonged ventilation in the intensive care unit (ICU) classically favour percutaneous dilational tracheostomy (PDT), which may be performed at the bedside. The large multisite UK-based TracMan study found no significant reduction in 30-day or 2-year mortality, antibiotic use, ventilator associated events or duration of hospital stay when comparing early (day 1–4) and late (day ≥10) tracheostomy.¹ PDT rates in the UK are rising, with 97% of responding ICUs in one 2005 postal survey performing PDT compared with 78% in 1998.²

This audit was performed in Scotland’s largest health board, which contains a large head and neck surgery unit. The fourth National Audit Project (NAP4) of the Royal College of Anaesthetists and the Difficult Airway Society identified 29 anaesthetic cases of ST, all of which were successful. Cricothyroidotomy was the first approach in 29 cases (26 needle, 3 scalpel). Fifteen (58%) of the needle cricothyroidotomies failed. NAP4 identified eight needle cricothyroidotomies sited in the ICU or emergency department, of which six (75%) failed, requiring conversion to either ST or PDT.³ The Difficult Airway Society 2015 guidelines for management of unanticipated difficult intubation in adults describe the current widely accepted algorithm to managing adult patients with difficult airways, with preferred eFONA in a ‘can’t intubate, can’t oxygenate’ scenario being scalpel cricothyroidotomy.⁴ Much published controversy exists surrounding this procedure, particularly from otolaryngologists, who typically favour bedside emergency tracheostomy in this situation.^5,6

This case series reviews elective and emergency tracheostomy formation in adult patients over a three-year period (January 2015 – December 2017). In an era of changing emergency airway algorithms, and a decline in tracheostomy for major head and neck procedures, this audit was performed in a health board with a large head and neck unit. The aim of this audit was to assess tracheostomy rates and to examine the pathway of patients requiring emergency airway management.

Methods

A retrospective audit was conducted of all clinically coded tracheostomies performed in adult patients in NHS Greater Glasgow and Clyde between 1 January 2015 and 31 December 2017. No patient identifiable data were included. The data parameters audited are listed in Table 1.

Table 1.

Data parameters audited

General	Surgical	Anaesthetic
Patient demographics	Procedure	Drugs
Diagnosis	Securing	Intubation equipment
Indication	Location	Laryngoscopy grade
Decannulation	Proceduralist grade	Endotracheal tube size
Admitting specialty	Decannulation	High flow nasal oxygen use
Complications	Technique
Mortality

Results

A total of 523 tracheostomies were performed during the audit period, of which 346 (66%) were in men. The mean patient age was 60 years. The overall mean deprivation quintile was 2.1 (1 = most deprived, 5 = least deprived). For those aged <65 years, the mean deprivation quintile was 1.9. The responsible specialty was oral and maxillofacial surgery (OMFS) in 264 cases (50%), ICU in 103 cases (20%), otolaryngology in 97 cases (19%), neurosurgery in 20 cases (4%), respiratory medicine in 10 cases (2%) and neurology in 8 cases (2%). Twenty-one tracheostomies (4%) were performed under the care of other specialties.

Indications

Of the 523 tracheostomies performed in total, 435 (83%) were elective (333 GA, 6 LA, 96 PDT) and 88 (17%) were emergency (60 GA, 25 LA, 2 PDT, 1 bedside emergency), performed for pending airway obstruction. Of the elective tracheostomies, 266 (61%) were performed with oral resection and/or neck dissection.

Tracheostomy was performed by OMFS in 279 (53%), otolaryngology in 152 (29%) and anaesthetics in 92 patients (18%). Two hundred and seventy-two tracheostomies were performed as part of a larger procedure, typically oral cancer resection. The grade of proceduralist was consultant in 479 cases (92%), with senior grade non-consultants performing the remaining 44 tracheostomies (8%).

The mean time to decannulation was 12.8 days. There were 54 documented complications, most commonly lower respiratory tract infection (LRTI) (n=28).

There were 161 deaths overall, of which 139 (86%) were unrelated to tracheostomy. Deaths attributable to tracheostomy were documented as LRTI (n=12), laryngeal bleeding (n=3), blocked tracheostomy (n=1) and respiratory failure (n=1). The cause of death was unknown in five cases.

Elective tracheostomy

Of the 435 elective tracheostomies, 97 were PDTs performed to facilitate ventilation. There were six LA tracheostomies, performed for a variety of indications: head and neck malignancy, bilateral vocal cord palsy, traumatic brain injury, Guillain–Barré syndrome and infected metalwork. The mean time to decannulation was 12.9 days. The decannulation rate was 79% (13 undocumented); in 43% of these cases, decannulation occurred within seven days. Common indications for elective tracheostomy are listed in Table 2.

Table 2.

Common underlying diagnoses for elective tracheostomy

Diagnosis	Total tracheostomies	Elective tracheostomies
Oral or head and neck cancer	280	65%
Lower respiratory tract infection / respiratory failure	37	9%
Trauma	23	5%
Other infection/abscess	17	4%
Osteoradionecrosis	5	1%
Other acute medical pathology	72	17%

Almost two-thirds of the elective tracheostomies (n=260, 60%) were performed by OMFS. A fifth (n=93, 21%) were performed in the ICU and a similar number (n=82, 19%) were performed by otolaryngology. Presence of a consultant proceduralist was documented for 398 (91%) of the elective tracheostomies while the most senior proceduralist present was a non-consultant in 37 cases (9%). Elective tracheostomies performed by specialty and year are listed in Table 3.

Table 3.

Elective tracheostomies performed by specialty and year

	2015	2016	2017
Oral and maxillofacial surgery	93 (62%)	81 (57%)	86 (59%)
Otolaryngology	26 (17%)	31 (22%)	25 (17%)
Intensive care medicine	30 (20%)	29 (21%)	34 (23%)
Total	149 (100%)	141 (100%)	145 (100%)

There were 332 GA elective STs (not including PDTs). Of the 318 procedures where the anaesthetic chart was available, propofol was the induction agent of choice in all but one, in which volatile inhalational agent was used. Rocuronium was the most frequently documented neuromuscular blocking agent (n=269), followed by vecuronium (n=28). Atracurium was used in five instances. Documented intubation equipment for this cohort included: direct laryngoscopy (n=151, 45%), video laryngoscopy (n=113, 34%), awake fibreoptic intubation (n=33, 10%), laryngeal mask (n=5, 2%) and nasal tracheal intubation (n=2, 0.6%). The remaining cases were undocumented. Gum elastic bougie was documented as an adjunct to laryngoscopy in 13 cases (4%).

Complications relating to elective tracheostomy were documented in 40 patients (9%). LRTI complicated 25 procedures. Four PDT patients not undergoing oral resection as part of the initial procedure returned to theatre for washout of haematoma or collection. Other less common complications included mild bleeding (n=2), localised infection (n=2), dislodged tracheostomy tube (n=2), tracheo-oesophageal fistula (n=1), tracheitis (n=1), prolonged dysphagia (n=1), respiratory failure (n=1) and subglottic granuloma (n=1).

Of those undergoing elective tracheostomy, 116 (27%) died prior to the data collection period (May – June 2018). Forty-six (40%) of those deaths occurred in patients without a primary diagnosis of cancer. Fifty-nine deaths (51%) were attributable to locally advanced or metastatic malignancy and eighteen (16%) to LRTI. Other less common causes of mortality included sepsis, myocardial infarction, stroke and multiorgan failure.

Emergency tracheostomy

There were 88 emergency tracheostomies, 2 of which were PDTs. Sixty patients (68%) underwent attempt at GA. The anaesthetic chart was unavailable for three of these. Documented anaesthetic equipment in those attempting GA included: video laryngoscopy (n=26, 43%), awake fibreoptic intubation (n=16, 27%), standard (n=12, 20%) and flexible nasendoscopy assisted (n=2, 3%). In one case, the equipment used was not clearly documented on the chart. Fifty-five (63%) of the emergency tracheostomies were intubated, twenty-nine (33%) were not intubated (unable/decided against) and four were undocumented. The most commonly documented reasons for unsuccessful intubation in cases of emergency tracheostomy were no view on flexible nasendoscopy and awake fibreoptic intubation. High flow nasal oxygen (HFNO) was used in 8 (28%) of the 29 unsuccessfully intubated emergencies, 20 (23%) of the 88 total emergencies and 32 (6%) of the 523 total tracheostomies.

Head and neck malignancy was the most common indication for emergency tracheostomy (59%). This was followed by neck space infection (15%), trauma/non-infectious swelling (14%) and other causes (12%).

Of the 88 emergency tracheostomies, 28 (32%) were decannulated and 36 patients (41%) died of locally advanced or metastatic malignancy. Two patients (2%) had a long-term tracheostomy or conversion to laryngectomy. Complications attributable to emergency tracheostomy occurred in eight individuals (9%). Three patients developed LRTI, one minor bleed, one dislodged tracheostomy tube, one persistent cord oedema, one local infection and one surgical emphysema. Three-quarters of emergency tracheostomies (n=69, 78%) were performed by otolaryngology, the remaining cases (n=19, 22%) by OMFS.

Emergency cricothyroidotomy

There were three documented cases of eFONA. Two were following failed extubation and one was following failed intubation due to venous congestion after unsuccessful internal jugular central venous catheter insertion.

Discussion

This study demonstrates that tracheostomy is a commonly performed multispecialty procedure. Of the 523 tracheostomies performed during our 3-year audit period, 83% were elective. LA tracheostomy was performed infrequently. Two-thirds (65%) of elective tracheostomies were performed as part of oral or head and neck cancer resection. OMFS performed 60% of the elective tracheostomies. Over three-quarters (79%) of elective patients were decannulated, with a mean decannulation time of <14 days. Consultants were present in over 90% of elective procedures. Our data did not capture the lead proceduralist in most cases.

Over the three-year study period, there was no evidence of change in rate of elective tracheostomy performed among specialties. Owing to recent health board restructuring and centralisation of services, it was not possible to collect historical data on tracheostomies performed for a similar population.

The complication rate in elective cases was 9%, most commonly LRTI. Over a quarter (27%) of the elective patients had died by the time of data collection; in most cases, this was as a result of locally advanced/metastatic malignancy.

A meta-analysis by Dulguerov et al comparing ST with PDT showed that perioperative complications occurred in 3% of STs performed between 1985 and 1996, in contrast to 10% of PDTs during the same time period.⁷ A prospective randomised controlled trial conducted by Friedman et al found no statistically significant difference in perioperative complication rates between ST and PDT although postoperative complication rates did differ (41% for ST vs 12% for PDT, (p=0.008).⁸

Another meta-analysis of prospective trials comparing ST and PDT in long-term ventilated patients by Freeman et al demonstrated no difference in overall complication rates between the two techniques.⁹ However, PDT was associated with lower risk of peri and postoperative bleeding, stomal infection and overall postoperative complication rate when compared with ST.

Our complication rate for ST is slightly higher than in the literature. This could be due to multimorbidity, tracheostomy formation in the context of a larger procedure (eg neck dissection) in advanced malignant disease and inclusion of ‘palliative’ tracheostomy in patients with non-curative disease.

Postoperative complications relating to tracheostomy occurred in 8% of STs and 10% of PDTs. Over a third (38%) of all tracheostomies were decannulated within seven days. (For ST, this was generally earlier than for PDT.) The mean time to decannulation was 12.8 days. The overall decannulation rate was 71%. Eighty-one per cent of PDTs and sixty-nine per cent of STs were decannulated. This compares favourably with other published UK rates of 49% of critical care and 61% of ward patients, with approximately a quarter of ward-based decannulations occurring within seven days.¹⁰

Use of HFNO to oxygenate during tracheostomy was 6% overall. There are no published data on worldwide HFNO rates for this indication. The non-depolarising neuromuscular blocking agents rocuronium and vecuronium were the favoured muscle relaxants rather than the depolarising neuromuscular blocking agent suxamethonium (which has been favoured historically for use in rapid sequence inductions), reflecting the recent availability of rapid reversal with sugammadex. Video laryngoscopy was employed frequently, with good success. eFONA was performed in just three instances. It is noteworthy that otolaryngology performs fewer tracheostomies than OMFS. This may be explained by use of other endoscopic airway procedures by otolaryngologists (particularly in elective cases) to negate the need for tracheostomy.

Overall, STs and PDTs are performed for a wide variety of indications, most typically oral and head and neck cancer. Adverse outcomes are not uncommon although advanced malignant disease contributes significantly to morbidity and mortality.

Conclusions

This paper discusses the findings of a comprehensive anaesthetic and surgical audit of tracheostomy experience. Over 150 adult tracheostomies were performed annually. GA elective tracheostomy represents the most common form of tracheostomy although LA and PDT are used in complex cases. There was no evidence of change in rate of elective tracheostomy performed by specialty over the study period. Seventeen per cent are emergency procedures, of which two-thirds are intubated. Head and neck cancer is the most common pathology in adult patients undergoing both emergency and elective tracheostomy, which is often performed owing to anticipated postoperative swelling. LRTI is the most common complication while advanced malignancy is the most common cause of death. eFONA is rarely performed.

As a tertiary facility performing frequent tracheostomy (with close multispecialty contact between anaesthetists, radiologists, and head and neck surgeons), we have significant experience in performing both elective and emergency tracheostomy. Head and neck cancer represents the most significant disease burden on our service. Complications attributable to tracheostomy occur infrequently although one in four patients undergoing elective tracheostomy had died by the time of data collection, most typically because of advanced malignant disease. Given the life changing impact of tracheostomy, it is crucial that timely informed discussions take place with our patients to ensure that we practise realistic medicine and focus our efforts on enhancing quality of life.