JC: The last breath for apnoeic oxygenation?

The last gasp for passive oxygenation in RSI_(2)

Apnoeic oxygenation in emergency intubation in the ED is a concept that has emerged in the past few years and has been promoted by the FOAMed community quite extensively.

I have to admit it is something we have recently introduced as a standard of care in Virchester along with other revolutionary techniques (like RSI checklists – I promise to share this via a blog about this soon) and I am an avid fan of it too.

The rationale behind it is that it is supposed to provide additional oxygen during the critical phases of the intubation in order to prevent episodes of hypoxia/destauration.

It has been largely endorsed across the world of emergency medicine but an American article published last week in the American Journal of Respiratory and Critical Care Medicine seems to suggest a lack of evidence and therefore contradict the above benefit.

I suggest that this week, we have a look at this paper (with, as always, a critical eye) as part of our now regular Journal Club series…

As always, read the full paper if you can. Abstract below.

Screenshot 2015-10-10 09.45.27


A bit of background for those needing to catch up with FOAMed:

I might be stating the obvious but I would start by saying that hypoxaemia during emergency intubation is bad for the patient’s brain (and indeed other vital organs) but is also bad for the emergency physician’s nerves.

The idea is that provision of supplemental oxygen via nasal cannulae during laryngoscopy has been hypothesised (and demonstrated to work in small studies – see later) to reduce episodes of desaturation. This works even without the presence of ventilation secondary to diffusion due to alveolar pressure difference.

As mentioned above, this technique has been recently adapted, adopted and endorsed by many EDs and other critical settings using RSIs.

It has been used to prevent desaturation in patients undergoing brain death examination, endoscopy, bronchoscopy or elective endotracheal intubation.

“Not our cohort of patients!” I hear you say. There have been out-of-operation room intubations papers published around this topic, which have suggested benefits (1,2) I would reply.

What type of paper was this?

The FELLOW (Facilitating Endotracheal intubation by Laryngosocpy technique and apnoeic Oxygenation Within the intensive care unit) was a randomised, open-label, parallel-group, pragmatic trial.

As you will have guessed, the setting was a critical care unit and we could argue that this population group is probably close to ours in the ED – critically ill patients, you might say. I will come back to this a bit later.

The trial was also concomittantly comparing direct video laryngoscopy to standard laryngoscopy and the authors presented these results too. I am deliberately not discussing those results as I am always dubious at trials looking at several different outcomes (not necessarily related) at the same time. These papers read as if the authors set out to look at something, came up with an idea mid-way and jumped on the bandwagon. But I am slightly speculating here, so let us have a look at the results…

Who was included?

The study took place over a period of 12 months and 150 adult patients undergoing endotracheal intubation were recruited. Patients were similar at baseline. Benefit in specific subgroups could therefore not be excluded (a sub-group analysis was not undertaken by the authors).

Randomisation was undertaken by permuted blocks into intervention (apnoeaic oxygenation via nasal cannulae at 15L/min) and control (usual care). This seemed appropriate to me.

For obvious reasons , blinding was not feasible so clinicians and study personnel were aware of  the study groups they were intervening in.

What were the considered outcomes?

Primary outcomes were the lowest SpO2 between induction and two minutes after successful ETT placement. I wonder here how the authors measured that precise time of successful ETT placement (ETCO2? lung expansion?other method?) and this does undoubtedly have an impact on the SpO2 value (the longer the patient is paralysed without a definitive airway, the higher the likelihood of desaturation).

Secondary outcomes were incidence of hypoxaemia (SpO2 <90%), severe hypoxaemia (SpO2 <80%), desaturation (decrease in SpO2 greater than 3%) and change in saturation from baseline. I have to note the the definition of most of the secondary outcomes appear to be arbitrary to me but I guess the authors had to pragmatically quantify these terms.

I will not detail the tertiary outcomes of this article (you will need to read the full paper yourself – and you know very well that you always should!) and this is because I was interested in the findings around desaturation only, since this is the primary reason for utilising apnoeic oxygenation in the Emergency Department.

What were the findings?

Oxygen saturation at the time of induction was 99% [IQR 96-100%] with the intervention versus 98% [IQR 94-99%] in the usual care arm (p=0.03).

It is also worth noting that there was no difference in the performance of the procedure, time from induction to secured airway and in the rate of successful intubation (p values presented all non significant).

The authors main finding was no difference between the two arms with regard to the primary outcome: 92% [IQR 84-99%] versus 90% [IQR 80-96%] respectively (p=0.16).  Apnoeaic oxygenation did not impact the proportion of patients as defined in the secondary outcomes either.

To detail this more, no difference was found between the intervention and usual care groups in incidence of SpO2 <90% (44.7% versus 47.2%, p=0.87), SpO2 <80% (15.8% versus 25%, p=0.22) or decrease in SO2 >3% (53.9% versus 55.6%, p=0.87)

The authors therefore concluded that, for patients intubated in the medical ITU, routine use of additional oxygenation was safe but ineffective and the routine use during emergent endotracheal intubation could therefore not be recommended.

A bit more about the limitations

I have detailed some of my reservations above around the author’s arbitrary definition of desaturation(s) and other specific points.

Furthermore, the majority of patients recruited in the FELLOW study were intubated for respiratory failure.  If anything, the results of this trial should therefore re-assure us but I could argue about some issues here about external validity of the trial (and this despite my earlier statement that this patient group is probably closest to the ED population we see daily).

In the absence of prior data within the study population, the power calculation was based on a previous trial

So what? 

So, a trial comes suddenly along and seems to contradict a practice most of us had implemented in their RSI drills, checklists and practice.

Will this change your RSI procedures today based on this single paper or will you look at this with a critical eye and draw your own conclusions?

Remember that the wise thirst for questions, not answers…





Further reading/listening

Check out this podcast with one of the FELLOW trial authors at EMCrit

The FELLOW Trial at EMergucate

Ryan Radecki takes on the FELLOW Trial at EMLitOfNote

Read about the PREOXYFLOW trial at the clinical bottom line

Before you go please don’t forget to…


  1. Stuart Duffin

    I’m going to continue using it for now. Although it’s probably better called apneic peri-ox rather than pre-ox! The FELLOW study seems to have had a lot of patients with what might be better described as ‘non-apneic oxygenation’, which isnt really the same thing, but is a viable alternative. Feel strongly that the lack of harm arguement figures heavily in me continueing to use it. Anyone who’s done brain-death testing knows that true apneic oxygenation, with 1-2l/min and a definitive airway really does work, even in patients with not-totally-healthy lungs, i’ve seen it dont on someone with only one lung without desat!

    Really good post, thanks


  2. Pingback: The FELLOW Trial: An End to Apneic Oxygenation? - R.E.B.E.L. EM - Emergency Medicine Blog

  3. Rob Dawes

    I’m someone who doesn’t use it. What I do though is always use a Water’s circuit to preoxygenate – whether in hospital or prehospital (50% of my time). My view has always been the nasal specs just make up the extra % of oxygen you lose when using a BVM (even the best designed BVM’s don’t come close to delivering an FIO2 of 1.0) whereas the Water’s circuit does, and you can add enough PEEP in those patients with a high ET-PACO2 grad to splint the alveoli. To me it just adds more faff to a sometimes difficult scene.


  4. Pingback: Apnoeic Oxygenation | Emergency Medicine Podcasts

  5. Pingback: Episode 38 – The Nose | FOAMcast

  6. Pingback: Research & Reviews in the Fastlane 106 | LITFL: Life in the Fast Lane Medical Blog

  7. Pingback: Research and Reviews in the Fastlane 106

  8. Pingback: LITFL Review 201 | LITFL: Life in the Fast Lane Medical Blog

Thanks so much for following. Viva la #FOAMed