To tube or not to tube? That is the question.

Author: Juan Pedro Carrasco-Alvarez, Govind Oliver / Editor: Charlotte Kennedy / Codes: / Published: 11/01/2019


The content you’re about to read or listen to is at least two years old, which means evidence and guidelines may have changed since it was originally published. This content item won’t be edited but there will be a newer version published if warranted. Check the new publications and curriculum map for updates

What is the AIRWAYS-2 Trial?

The AIRWAYS-2 trial compared the effectiveness of i-gel supraglottic airways (SGA) with endotracheal intubation (TI) in the initial airway management of adult patients in non-traumatic out of hospital cardiac arrest (OHCA). Despite being a study in the prehospital setting, orientated towards paramedics, it is interesting and important to consider how the results of this trial apply to the Emergency Physician waiting for such a patient to arrive at the Emergency Department (ED). Here we describe the trial, what it showed and how it might affect our practise in the future.

Full reference:

Benger JR, Kirby K, Black S, et al.(2018) Effect of a Strategy of a Supraglottic Airway Device vs Tracheal Intubation During Out-of-Hospital Cardiac Arrest on Functional Outcome The AIRWAYS-2 Randomized Clinical Trial. JAMA. 2018;320(8):779791. doi:10.1001/jama.2018.11597


Why was the study needed? What knowledge gap was it trying to address?

The College of Paramedics supports the use of TI as the initial advanced airway of choice in cardiac arrest (College of paramedics, 2018). However, other advanced airway management strategies such as SGAs have increased in popularity. Observational studies suggest there are potential advantages to either approach but large-scale randomized control trial evidence to ascertain which approach is optimal has hitherto not existed.

What question were the researchers trying to answer?

The AIRWAYS-2 trial aimed to answer whether SGA or TI is superior in achieving better functional outcome as the initial advanced airway management of OHCA.

How did they answer their research question?

This was a prospective un-blinded trial with participating paramedics from four large ambulance services in England, responsible for 21 million people (40% of England’s population), randomised to use either TI or SGAs. Non-traumatic OHCA patients were automatically enrolled when attended by participating crew and assessed retrospectively for inclusion criteria. It is important to note that paramedics retained the freedom to adapt the airway management at their clinical discretion, despite being randomised to one trial arm.

The primary outcome measure of the study was the level of disability at 30 days. This was assessed using the Modified Rankin Scale (MRS). The authors divided this dichotomously, considering scores of 0-3 as a good outcome (with 0 being no symptoms and 3 representing moderate disability) and scores of 4-6 a poor outcome (with 6 being death). They covered a range of secondary outcomes, including but not limited to: initial ventilation success, regurgitation and aspiration, and achievement of ROSC.

What were the main findings from the research?

The initial cardiac arrest rhythm, proportion of patients receiving by-stander CPR, paramedic time of arrival and number of patients achieving ROSC were similar across groups. The initial airway management before the arrival of a trial paramedic was also comparable between groups, with airway management in progress in 31.5% and 30.1% of cases in the SGA and TI arms respectively. A similar proportion of these involved SGA insertion or TI prior to study paramedic arrival, 18.9% vs. 13% and 0.2% vs. 0.2% respectively.

Primary Outcome:

A total of 9,296 patients (4,886 SGA and 4,410 TI) were enrolled in the trial. The primary outcome, measured by the MRS, was available for 9,289 (99.9%) patients. 311 (6.4%) patients had a good functional outcome (MRS 0 – 3) with SGA management, compared to 300 (6.8%) patients managed with TI; this gave an adjusted odds ratio of 0.92 [95% confidence interval (CI) 0.77-1.09]. Superiority in functional outcome at 30 days was not demonstrated by either approach.

It should be noted that an intention to treat analysis was performed. Of the 4,886 patients in the SGA group, 4,145 (84.8%) patients received advanced airway support, of which 3,994 (96.4%) received a SGA; this represents 81.7% of the SGA intervention arm overall. Of the 4,410 patients in the TI group, 3,413 (77.4%) received advanced airway support, of which 2,718 (79.6%) received TI; this represents 61.6% of the TI intervention arm overall.

Primary Outcome Subgroup analyses:

Looking only at those patients in either intervention arm who received advanced airway management, more patients who received a SGA had a good outcome compared to TI: 163 (3.9%) vs. 88 (2.6%) respectively, with an adjusted odds ratio of 1.57 [95% CI 1.18 – 2.07]. This effect was also observed when outcomes were analysed by the first type of airway management intervention the patient received irrespective of the trial arm. A good outcome was seen with SGA insertion in 193 patients (4.2%) compared to 58 (2.0%) who received TI, giving an adjusted odds ratio of 2.06 [95% CI 1.51 2.81].

Secondary Outcomes:

There were several key secondary outcome measures of note and of interest to the ED physician.

The odds of achieving successful ventilation (after up to 2 attempts) were greater with a SGA (87.4%) compared to TI (79.0%), with an adjusted odds ratio of 1.92 [95% CI 1.66 – 2.22]. However, there was a greater risk of unintentionally losing a previously established airway when using a SGA (10.6%) compared to TI (5.0%), with an adjusted odds ratio of 2.29 [95% CI 1.86 2.82].

Overall, rates of regurgitation and aspiration in both groups were similar. A trend was seen towards greater regurgitation and aspiration before intervention when TI was undertaken (21.1% and 13.5% respectively) compared to SGA use (17.4% and 11.0%), whereas SGA use had greater regurgitation and aspiration rates during or after intervention (18% and 9.8% with SGA vs. 12.5% and 7.0% with TI).
There were slightly higher odds of arriving at the ED with ROSC for those in the SGA arm (30.6%) compared to the TI arm (28.4%), with an odds ratio of 1.12 [95% CI 1.02-1.23].

Are there any limitations we should be aware of?

The study sites several limitations. Our aim is not to critically appraise the trial but we will cover these briefly to add context to the findings.
The key consideration to note is that the trial population included patients who both did and did not receive advanced airway management and more patients in the SGA arm received advanced airway intervention than in the TI arm. An intention to treat analysis was applied, meaning that patients assigned to the intervention arms were included and analysed according to their initial randomisation, irrespective of whether or not they received advanced airway support. This is a possible confounding factor and raises the potential of a type 2 error (false negative). An intention to treat analysis is methodologically the most robust but we must understand its impact when interpreting the primary outcome measure. Another important consideration, given that an intention to treat analysis was performed, is that paramedics delivered advanced airway support based on clinical grounds, despite being randomised to a particular intervention arm. This resulted in crossover of interventions between each arm and therefore the analysis based on randomisation is not entirely representative of the actual intervention received.

Another potential limitation, the significance of which is unknown, is that the time from initial confirmation of cardiac arrest (or call to the emergency services) to successful management with an advanced airway was not recorded. The time of arrival of the trial paramedic was part of the data set but this did not include the actual time at which the airway was successfully inserted. Pragmatically this may not have been reliably recordable. However, it does means it is difficult to compare the strategies in the context of a meaningful therapeutic window. It is also unclear if or when the airway strategy in the SGA group was changed on arrival to the ED.

Finally, the participating paramedics were volunteers, not randomized from the overall paramedic population, and therefore might not be representative of the whole paramedic workforce. In addition, the findings are not generalisable to areas where there are physician-led crews.

What is the take home message from this paper, will it change how we practice?

The results of this study indicate that SGAs are not superior to TI for the initial management of OHCA by paramedic crews. No favourable functional outcome was seen at 30 days with either intervention. There was evidence from the subgroup analysis that the odds of a good recovery were higher in the SGA group, however this effect was only found on secondary analysis.

This trial does not present evidence that we can use to directly influence our practice in the ED. However, as the trial findings are disseminated and applied amongst ambulance services and paramedic crews we may see more patients arriving with a SGA as their advanced airway intervention following OHCA. The implications of this are that we may find ourselves increasingly facing the decision of when or whether to change to an endotracheal tube during on-going resuscitation in the ED.


  1. College of paramedics, 2018. Consensus Statement a framework for safe and effective intubation in paramedics.
  2. Taylor, J. et al., 2016. Design and implementation of the AIRWAYS-2 trial: A multi-centre cluster randomised controlled trial of the clinical and cost effectiveness of the i-gel supraglottic airway device versus tracheal intubation in the initial airway management of out of hospital cardiac arrest. Resuscitation, , 109(), pp. 25-32.

Leave a Reply