Abstract
Despite the training and skills of airway managers, airway management complications still occur and may cause patient harm or death. The causes are multifactorial and may include patient, environment and clinician factors. Airway complications likely contribute to a significant proportion of deaths due to anaesthesia and are certainly more common outside the operating theatre and especially in the critical care unit. Reported incidences of failure and harm during airway management vary depending on the population studied and definitions used. Numbers may be of less value than understanding themes that help us improve care and reduce harm. The chapter emphasises that conventional research (e.g. device evaluation studies and randomised controlled trials) may be of little use in identifying low frequency events and complications because of their restricted inclusion and exclusion criteria, the use of devices only by experts and in conventional settings and because of their focus on efficacy rather than safety. The chapter highlights the important and growing role of registries and databases. Several are described in detail including the 4th National Audit Project and the Dutch ‘mini-NAP’. The value and limitations of litigation databases are explored. Specific complications of note are described at the end of the chapter.
Overview
Anaesthetists and others who manage the airway are trained airway specialists who strive to prevent harm to our patients, but airway management complications may still occur. Sometimes these happen because of patient factors, sometimes because management is suboptimal and, most often, because of a combination of these two. Recent developments and technology should have made anaesthesia safer but complications during airway management still lead to consequences such as cancelled operations, unplanned intensive care admission, airway trauma, brain damage or even death. Modern anaesthesia has become considerably safer with mortality attributable to anaesthesia falling at least 10-fold in the last few decades. Improvements in airway safety almost certainly contribute to this.
How much airway complications contribute to anaesthesia-attributable mortality and morbidity is unknown and will vary considerably according to the location. In well-resourced environments mortality is low: a Japanese study reported anaesthesia-related deaths in a low-risk population to be 10 per million, 40% of which were airway related. In contrast, in a low-resource environment, in Togo, post-operative mortality rate was 1 in 38 at 24 hours; 93% of deaths were judged avoidable with half attributed to anaesthesia including 30% due to respiratory management. It is likely that much airway-related morbidity and mortality is avoidable, but only with availability of trained personnel and sufficient equipment. Airway complications related to airway management performed outside the operating room (OR) environment are several-fold more common than in the OR. Consequently, although only a modest proportion of airway management takes place in the intensive care unit (ICU) up to a quarter of major events occur in that location.
Details of many specific complications, avoidance strategies and management plans are included in other chapters, so this chapter provides an overview of complications occurring during airway management, focussing particularly on epidemiology and patterns.
Despite almost universal concern amongst anaesthetists about avoidance of airway complications the reality is that serious complications are infrequent enough that individual practitioners will only encounter complications infrequently. For that reason, databases likely provide the most reliable sources of information on complications. In this chapter we will focus on information from clinical practice databases including
– the UK 4th National Audit Project: major complications of airway management (NAP4)
– a Dutch prospective database of airway complications (mini-NAP)
– data from litigation databases including that of the UK National Health Service NHS Litigation Authority (NHSLA) and the USA American Society of Anesthesiology Closed Claims Project (ASACCP).
Common Airway Complications – Difficulty and Failure
Incidences of airway management failure will vary depending on definitions, operator experience and the patient population examined: difficult laryngoscopy occurs in 1 in 16 unselected elective patients but this rises to 1 in 5 in patients having cervical spine surgery. Failure and complications occur disproportionately commonly in ICU and in the emergency department (ED) where failure may be 10-fold higher than during anaesthesia.
Common complications of airway management are listed in Table 3.1. Difficulty and failure of a primary airway procedure should be considered complications, especially as they inevitably precede the vast majority of airway complications leading to patient harm. Failed tracheal intubation likely occurs at some point in almost every airway-related fatality.
Complication | Location or setting | Approximate incidence |
---|---|---|
Difficulty | ||
Face mask ventilation | Anaesthesia | 1 in 50–100 |
Ventilation via SGA | Anaesthesia | 1 in 10 |
Intubation (low-risk group) | Anaesthesia | 1 in 18 |
Intubation (high-risk group) | Anaesthesia | 1 in 5 |
Face mask ventilation and laryngoscopy | Anaesthesia | 1 in 250 |
Intubation | ED | 1 in 12 |
Intubation | ICU | 1 in 3 |
Failure | ||
Face mask ventilation | Anaesthesia | 1 in 600 |
Ventilation via SGA | Anaesthesia | 1 in 50 |
Intubation | Anaesthesia | 1 in 200–1500 |
ICU | >1 in 100 | |
ED | >1 in 100 | |
CICO | Anaesthesia | 1 in 5000 |
Front of neck airway | Anaesthesia | 1 in 50,000 |
ED | 1 in 400 |
Risk factors for failure of airway management and complications are described in Table 3.2.
Factor | Note |
---|---|
Difficult airway | While seemingly obvious that a difficult airway is a cause of complications it is more complex. Approximately half of difficult intubations are not predicted. Tests have low sensitivity. Previous difficult intubation is the best predictor of future difficult intubation and should never be ignored. Most airway complications, however, occur in patients who are not predicted and may not have anatomically difficult airways |
Obesity | Obesity is repeatedly identified as a risk factor for difficulty for all types of airway management in all settings. Increased risk likely starts as low as BMI 35 kg m−2. Reduced safe apnoea time and progression to severe hypoxaemia is the greatest factor |
Emergency | Urgency of airway management and factors such as use of cricoid force increase risk of failure and complications up to 10-fold |
Outside OR | All locations outside the OR are associated with a marked increase in risk of failure and complications. Multiple reasons are described in the text |
Head and neck surgery | The combination of airway abnormality due to disease and treatment increases risk many-fold. The need for a shared airway and blood in the airway at extubation add further difficulty |
Reduced mouth opening | Decreases access for laryngoscope, SGA and airway adjuncts |
Reduced neck movement | Reduces mouth opening. Increases difficulty in optimal positioning for FMV and intubation. Hyperangulated videolaryngoscopy is useful to overcome the problem |
Previous radiotherapy | Increases difficulty in optimal positioning for FMV and intubation. Often associated with reduced neck movement. Hinders anatomical recognition and performance of eFONA |
Repetition of failing technique | Repetition of the same airway technique that has already failed is illogical – after a failed attempt at intubation subsequent attempts have an approximately 80% failure rate – but is consistently seen in airway disasters. Also termed ‘failure to transition’ to describe the failure to move to the next step of the airway algorithm |
Lack of strategy | Safe airway management requires a series of plans each consequent on the failure of the preceding plan. A lack of a strategy, communicated to all, leads to repetition of failing techniques and chaotic airway management |
Communication issues | Common in airway disasters. Failure to ensure a strategy is understood by all involved, including the transition points |
Poor decision making | This often involves choosing a poor primary plan and lacking a strategy for failure. During difficulty, a combination of repetition, missing algorithm steps and using unfamiliar techniques is often seen |
Untrained personnel | Training and knowledge are fundamental prerequisites for avoidance of airway complications. Training is not the same as seniority – seniors are more commonly involved in airway mismanagement than junior staff |
BMI, body mass index; eFONA, emergency front of neck airway; FMV, face mask ventilation.
Procedural difficulty is also associated with secondary complications – including airway swelling, trauma, pulmonary aspiration and development of airway obstruction and the cannot intubate, cannot ventilate/oxygenate (CICV/CICO) situation. Avoidance of primary difficulty is therefore at the heart of complication avoidance.
‘Composite Airway Failure’
It is essential to understand that in a patient in whom one airway technique fails, the risk of failure of other techniques is increased – so called ‘composite failures’.
After failed face mask ventilation intubation failure increases more than 10-fold.
After failed intubation, face mask ventilation fails in approximately 1 in 10.
After SGA insertion the risk of difficult mask ventilation rises threefold.
To minimise the risk of complications, when one technique is predicted to be difficult it is important to focus carefully on assessing the likely ease of other techniques that may be used for rescue.
National Level Complications: Lessons from NAP4
NAP4 examined major complications of airway management in the UK. This lengthy document cannot be adequately summarised here and is signposted in the Further Reading.
The aims of NAP4 were:
To examine the extent of major complications of airway management
To characterise these problems
To capture recurrent themes and causes
To make recommendations to improve airway management at national, organisational and individual levels. The former two are the basis of institutional preparedness and the latter of personal preparedness.
Key Points from NAP4
NAP4 was a 1-year national registry of major airway complications during anaesthesia and in ICU or ED. It only included cases leading to
death
brain damage
emergency front of neck airway (eFONA)
ICU admission or prolongation of ICU stay
As such it only captured the airway events with the worst outcomes, the very ‘tip of the iceberg’ – lesser complications or ‘rescued events’ were not captured.
A concurrent denominator survey enabled a national incidence of events to be calculated (Table 3.3). NAP4 studied the complications of approximately 3 million anaesthetics. There were 133 anaesthesia events, 36 in ICU and 15 in ED.
Risk of event | ||
---|---|---|
Included event | 46 per million | 1:22,000 |
Death | 5.6 per million | 1: 180,000 |
Death and brain damage | 6.6 per million | 1: 150,000 |
Amongst ≈3 million general anaesthetics, there were 16 airway-related deaths and 3 cases of persistent brain damage. Incidences are reported in Tables 3.3 and 3.4 according to the degree of injury. The rarity of such events is noteworthy and is a key reason why assessing safety in airway management is so difficult – no individual’s or department’s practice is likely to shed light on high level safety and this is also true for almost any randomised clinical trial (RCT).
Primary airway device | Events | Death and brain damage |
---|---|---|
Any | 1:22,000 | 1:150,000 |
Tracheal tube | 1:12,000 | 1:110,000 |
Supraglottic airway | 1:46,000 | 1:200,000 |
Face mask | 1:22,000 | 1:150,000 |
Important themes identified in NAP4 included:
Omitting an assessment of potential airway difficulty and risk of aspiration, and the failure to tailor the anaesthetic technique appropriately, contributed to poor outcomes.
Poor planning and ‘failure to plan for failure’ were common in events. Responses to unexpected difficulty and failure were unstructured. Airway managers started with only a single plan. Airway strategies were advocated; a logical sequence of plans, designed to manage failure at each step and to achieve oxygenation, ventilation and prevent aspiration.
Difficult or failing techniques were regularly managed with repeated attempts, especially tracheal intubation. This was associated with deterioration from a ‘cannot intubate, can oxygenate’ to a ‘cannot intubate, cannot oxygenate’ (CICO) situation. NAP4 strongly advocates adopting a limited number of attempts as part of any strategy.
Decisions made and techniques chosen were sometimes illogical, including using routine care in cases of known difficulty and avoiding awake intubation when strongly indicated. Lack of judgement, skills, experience, confidence and equipment all contributed.
Anaesthetists often used ‘their usual technique’ when this was not in the patient’s interest. Best care may require involvement of colleagues with other skill sets.
Quality of care was judged to be ‘poor’ or ‘good and poor’ in three quarters of cases. In a secondary study, human factors were identified in all cases (mean of four factors per case). Poor judgement, education and training were the most common contributory factors.
Delayed, difficult or failed intubation was the primary event in almost half of reports and intubation difficulty and failure likely occurred at some point in all cases.
SGAs (most often first generation) were used in the face of high aspiration risk or marked obesity and aspiration followed. Use by junior doctors and accepting a poorly functioning airway were themes in non-aspiration SGA events. Use of an SGA to avoid anticipated difficult tracheal intubation, without a rescue plan or strategy, was followed by problems: unstructured management of difficulty followed and some of these patients died. Fatalities would likely have been avoided by an awake intubation technique or tracheal intubation through the SGA (see Chapter 13).
Obese and morbidly obese patients were over-represented throughout NAP4. This finding has been replicated and reinforced in other important studies (see Chapter 24).
Head and neck cases accounted for 40% of all anaesthesia cases and the need for multidisciplinary communication and senior anaesthetic and surgical involvement was emphasised.
Many reported cases involved the obstructed airway. CICO was common in these. Human factors were plentiful including poor planning, communication, equipment, teamwork and situation awareness. Awake tracheostomy was too infrequently considered. When problems occurred transition to eFONA was often slow, even when part of the strategy.
Transition to eFONA when required was often delayed and eFONA often failed.
In anaesthesia events aspiration was the commonest cause of death (51% of reports of death or brain damage). Half of these cases involved tracheal intubation. Poor judgement and ignoring risk assessment were causative in many cases.
Unrecognised oesophageal intubation occurred in all locations and accounted for 1 in 16 cases. It was emphasised that harm from oesophageal intubation is preventable by capnography, even in cardiac arrest. (This is discussed further below.)
One quarter of events took place during emergence and recovery: all were associated with airway obstruction and many with post-obstructive pulmonary oedema. Blood in the airway and a suboptimal airway during maintenance were common precipitants.
ED and ICU
A quarter of airway events occurred in the ICU or ED. Estimates are that the rate of events leading to death or brain damage were, compared with anaesthesia, 35-fold higher in ED and 55-fold higher in ICU.
Permanent harm or death followed 61% of ICU reports, 33% of ED reports and 14% of anaesthesia reports.
In the ICU, much morbidity and mortality followed airway displacement, especially of tracheostomy and in obese patients. Delayed recognition and lack of a structured plan for such an event was prominent (see Chapter 28).
In the ED, most complications followed rapid sequence induction.
Suboptimal care, including preventable deaths, was especially common in ICU and ED reports. Issues included not recognising at-risk patients; poor planning; inadequate or inaccessible skilled staff and equipment; slow recognition of problems; unstructured responses; and poorly prepared institutional and individual strategies for managing predictable airway complications.
Failure to use capnography in ventilated patients or to interpret it correctly (and consequent failure to identify airway displacement or misplacement) contributed to more than 70% of ICU-related deaths.
Lessons from Databases and Registries
Airway-related databases and registries (whether permanent or short term as part of a trial) are sources of useful information but differ from routine databases that generally gather data about a large number of routine cases. Registries more often collect smaller datasets relating to a focussed area, e.g. patient populations, diseases, procedures or complications. There is a degree of overlap between databases and registries. Both can provide useful information about patterns of airway complications and in some cases detailed information.
Databases that include information from routine cases are useful as they generally create a complete dataset and this puts rarer complications into context and provides a denominator enabling calculation of incidences. However, they need to be very large to gather sufficient numbers of cases of interest to be useful. Limitations include: the effort required to collect such a volume of data; that the data is often collected for other purposes (e.g. for financial or administrative purposes) so that clinical information may be of secondary purpose, leading to omissions or perverse associations; databases from a single or atypical institution may not be generalisable. Selective databases and registries are more focussed, and a greater proportion of cases are likely relevant to those interrogating the dataset. This provides economy of effort in collection and analysis. Their limitations include that: they lack denominators so cannot (by themselves) provide incidences; and the methods by which cases are captured may lead to uncertainty over whether all cases are included.
A range of airway-related databases have been established in the last decade and they are now starting to provide important data about rates of complications and identifying risk factors for harm and insight into the efficacy of various rescue techniques. Some examples are listed in Table 3.5.
Database or registry | Area of practice and year started | Data source | Notes |
---|---|---|---|
|
| Routine data from > 70% of cases nationally | Detailed data on > 600,000 cases. Useful incidences and risk factors of e.g. difficult mask ventilation, difficult intubation and eFONA |
|
| One-year registry of all UK hospitals | Captured complications of 2.8 million anaesthetics. Contemporaneous denominator survey enabled incidence and risk factor determination https://www.nationalauditprojects.org.uk/NAP4_home |
APRICOT (Anaesthesia PRactice In Children Observational Trial: European prospective multicentre observational study: Epidemiology of severe critical events) |
| One-off collection of routine data from > 30,000 paediatric anaesthetics in 250 hospitals |
|
|
| Rolling database of closed litigation cases from USA |
|
University of Arizona College of Medicine registry |
| Single centre database of all ED intubations | Run by Drs J Sakles and J Moiser. Well reported and comprising > 6000 cases. |
|
| Database of all intubations in ED in 40+ contributing units | http://www.thesharpend.org/airway-registry |
|
| Database of all intubations in ED in > 20 contributing units |
|
|
| 22 specialised paediatric hospitals |
|
|
| 13 children’s hospitals in the USA | North American registry of difficult intubation in specialised centres with data on > 1000 difficult intubations |
|
| UK initially, expanding to Europe |
|
The Airway App |
| Self-reported cases |
|
RCoA-DAS FONA database |
| Self-reported cases | Due 2020 |