Within hospital practice, emergency anaesthesia outside of the operating theatre commonly occurs in the emergency department (ED) and critical care areas. The conduct of RSI in these areas is associated with a higher rate of poor views at laryngoscopy and increased complication rates (Taryle et al. 1979). The fourth National Audit Project (NAP4) of the Royal College of Anaesthetists and the Difficult Airway Society established a 12 month registry of major complications of airway management across NHS hospitals in the UK (Cook et al. 2011). It found that at least one in four major airway complications reported, occurred in the Intensive Care Unit (ICU) or ED, and that the outcomes of these events were more likely to lead to permanent harm or death than events in operating theatres. Such findings may relate to the difficulties faced by staff working outside of the familiar theatre environment and with less familiar equipment. This problem is further compounded by the fact that the sub-set of patients requiring emergency airway interventions tend to be more physiologically compromised, have a higher requirement for manual in-line stabilisation (MILS), and are generally more complex to manage.

Historically, within the UK, anaesthetists have conducted almost all in-hospital emergency anaesthesia. As different specialties have developed, notably Emergency Medicine (EM) and Intensive Care Medicine (ICM), RSI is increasingly also undertaken by other clinicians. The curricula of the Faculty of Intensive Care Medicine (2014) and Royal College of Emergency Medicine (2010) both reflect this, with emergency anaesthesia and airway management being an essential part of training.

Whether non-anaesthetic staff should be performing RSI unsupervised by an anaesthetist as routine practice has been much debated, and significant variation in practices exists between different hospitals. Previous UK studies of ED RSI comparing EM staff to anaesthetists found that anaesthetists achieved significantly better views at laryngoscopy and had a greater success rate for first-pass intubations (Graham et al. 2003; Stevenson et al. 2007). A recent census of UK-wide ED airway management found that 80 % of ED RSIs were performed by anaesthetists, predominantly senior anaesthetic trainees of ST3 and above (Benger and Hopkinson 2011).

In contrast, a recent study published by Edinburgh Royal Infirmary, had EM physicians performing 78 % of intubations, and anaesthetists 22 %, utilising a joint RSI protocol, in which a senior anaesthetist provides immediate support whilst trained EM staff perform RSI (Kerslake et al. 2015). This study is one of the largest series reported, including 3738 ED intubations at one large Scottish teaching hospital. Over the 13 years reviewed (1999–2011) they found that EM physicians achieved similar views but lower initial intubation success rates for RSI when compared to anaesthetists. Overall intubation was successful in 99.6 % of patients, with first time successful tracheal intubation in 85 % (anaesthetists 92 %, EM consultants 94 % and EM trainees 83 %; p < 0.001). In subgroup analysis of intubations since 2007, the difference in initial success rate between ED and anaesthetic physicians is no longer apparent (87 % vs. 88 % respectively). One explanation could be that the intubation skills of EM physicians have improved due to increased training and skills maintenance in anaesthesia and emergency airway management. It may also demonstrate the advantages of collaboration between anaesthetists and ED staff in performing emergency airway management, with prediction of ‘difficult’ intubation and use of the most appropriately experienced physician for the initial intubation attempt. Complications occurred in 9.4 % of RSIs, with no significant difference in rate of complications between anaesthetists and EM physicians.

In a similar prospective study, RSIs performed by critical care staff, encompassing those taking place in the ED, ICU and wards, demonstrated that non-anaesthetists had a significantly higher incidence of unsuccessful intubation attempts and multiple attempts, compared with anaesthetists (10.8 % vs. 3.9 %) (Reid et al. 2004). However overall (albeit self-reported) complication rates were not significantly different between groups.

A large US study reinforces the importance of optimising the intubation conditions for RSI, by suggesting that the incidence of complications increases significantly when more than one attempt at intubation is required (Thomas 2004). This was replicated in the findings of Kerslake et al (2015), where complications occurred in 8 % of patients overall; however the incidence increased significantly with the number of intubation attempts (7 % with one attempt, 15 % with two attempts and 32 % with three attempts). Such complications include desaturation/hypoxia, hypotension, vomiting/regurgitation, arrhythmia, cardiac arrest, misplaced endotracheal tubes and airway trauma.

In their retrospective study of patients intubated within the first 2 h of arrival at a Level 1 trauma centre, Sise et al. (2009) found a similar significant correlation with the rate of complications increasing with the number of intubation attempts. They also demonstrated that complications were more common when resident physicians (postgraduate years 2–4) were involved in the intubation, than when more senior physicians took the first intubation attempt (14.2 % vs. 9.7 %), and that they were significantly less successful at first intubation attempt (78 % vs. 92 %). This unsurprising relationship between training level and intubation success, was also seen by Kerslake et al. (2015). EM consultants had intubation success comparable to anaesthetists; however EM trainees had significantly lower success rates.

In summary, intubation outside the theatre environment has less chance of first time success and more chance of complications. Current evidence demonstrates that, as one may expect by virtue of their training and continuing experience, anaesthetists are more likely to gain a better view at laryngoscopy and intubate successfully, with fewer complications. Physicians of other specialty backgrounds can achieve comparable intubation success, and it may be more appropriate for a senior EM physician to perform RSI than a junior anaesthetic trainee.

It should be remembered that familiarity with the side effects and appropriate doses of induction agents, along with the ability to minimise and manage adverse effects is as important as the ability to get the best view at laryngoscopy. As with many things, the key factors in a safe and successful outcome, are appropriate training and ongoing maintenance of skills.

The pre-hospital environment is one of the most challenging clinical settings in which to practice. It can be austere as well as emotionally and physically demanding. The conditions are at best suboptimal, characterised by difficult access to the patient, limited equipment, inclement weather conditions and poor lighting (too dark or too bright). All these factors and many other stressors can hinder the performance of an optimal pre-hospital RSI. The pre-hospital practitioner may also have less support in terms of additional medical assistance and specialist airway equipment in the event of difficulties. To compound matters, many patients will have sustained head or neck trauma, which may lead to a more challenging intubation, even in the controlled environment of a hospital (either directly e.g. airway blood/swelling or indirectly e.g. manual in-line stabilisation (MILS) of the neck. Despite these issues, the threshold to intubate is often lower. This can be influenced by the difficulties in managing a partially obstructed airway and/or an agitated patient during transfer to hospital, particularly if this is in a helicopter.

Anecdotal reports suggest that pre-hospital RSI (PRSI) has been practised since the 1970s. The first report in the literature of drug-assisted pre-hospital intubation described paramedics using succinylcholine to assist intubation (Hedges et al. 1988). Over subsequent years, the technique has been progressively adapted from hospital practice.

Intubation and ventilation in hospital is considered a cornerstone in the management of critically ill patients. The optimum timing for this intervention has never been studied but it is assumed that the earlier the intervention takes place the greater the benefit it will offer. It has not been proven that RSI in the ED (vs. theatre or ICU) improves outcome and reduces mortality, and probably never will be, as it is now accepted practice. This has become the case with PRSI, although it is clearly possible that the risks of carrying out this procedure in the pre-hospital setting may outweigh the benefits in certain situations. In other words, when RSI is undertaken in suboptimal conditions there is the potential for increased morbidity and mortality. In the absence of appropriate personnel with the relevant training and equipment, a conservative approach may be appropriate.

Lossius et al. published a systematic literature review of pre-hospital intubation in 2011. They highlighted the difficulty in establishing an evidence-base for pre-hospital intubation when the available studies demonstrate such heterogeneity in procedures, providers, patients, systems and stated outcomes. In addition, the vast majority of publications are retrospective studies.

There is currently no randomised controlled data showing clear benefit in mortality or morbidity following pre-hospital intubation. The first randomised controlled trial of pre-hospital intubation involved paramedics intubating children without drugs (Gausche et al. 2000). In addition to the lack of physicians and drugs (i.e., not pre-hospital RSI), the study had major flaws.

The long-awaited ‘Head Injury Retrieval Trial’ (HIRT), which aimed to be the first prospective randomised controlled trial (RCT) evaluating the addition of pre-hospital physician care compared with paramedic only care for patients with severe head injury, has recently been published (Garner et al. 2015). Within the duration of the trial (2005–2011), changes in the ambulance service policies made physician treatment of these patients the expected standard of care, leading to significant protocol violation, non-compliance with recruitment and randomisation, and considerable crossover between groups. Using an ‘as-treated’ analysis of data, rather than the standard ‘intention to treat’ analysis, they were able to report a potential 30-day mortality reduction in adult patients with blunt trauma and Glasgow Coma Score (GCS) <9 receiving additional physician care, such as drug assisted intubation, but were not able to demonstrate a significant improvement in overall mortality or 6-month Glasgow Outcome Scores. Due to such potential bias in the study the findings are not conclusive; but it does highlight the difficulties in carrying out high-quality prospective pre-hospital research. It is suggested that future studies need to take place in settings where physician staffed teams are not part of the established standard care.

There have been numerous retrospective studies, and overall these have not provided clear evidence to support PHA either. Some of the studies suggest survival advantage (Winchell and Hoyt 1997; Arbabi et al. 2004). Others showed no improvement in neurological outcome or mortality (Stockinger and McSwain 2004). Other studies appeared to show an adverse outcome from pre-hospital intubation (Murray 2000; Eckstein et al. 2000).

The majority of studies are from the US (several from the same trauma database) and involved paramedics. Many used either no drugs or simply sedation with midazolam and no neuromuscular blocking drug. Both failed intubation and other complications are reduced by the use of neuromuscular blocking drugs (Bulger et al. 2005). These trials, most of which suggest inferior outcomes, are therefore clearly not applicable to pre-hospital RSI as currently practised by physician/paramedic pre-hospital teams. The study by Davis et al. of paramedic RSI in traumatic brain injury compared to historical controls was associated with an increase in mortality (Davis et al. 2003). Periods of reduced SpO₂, and hyperventilation were noted and the relatively limited training and experience of paramedics were cited as potentially important factors. These critical factors are still relevant today.