Category Archives: intubation

Old School/New School – Updating Classic RSI

Respect for the classics doesn’t mean being stuck with them. Here’s a refresher on why you might not want to do RSI like they used to by Dr Andrew Weatherall. This one is a cross post picked up from the paeds anaesthesia site he chips in on, 

Everything in medicine needs the occasional reboot. I mean not as often as Hollywood thinks we need to reinvogorate a superhero franchise but at least every now and then. Sometime that’s because we learn new things (cross reference here). Sometimes it’s because our perception of what is the biggest risk changes (more on that in a second). And sometimes we suddenly realise that the original reason something became fixed practice might not have been a thing in the first place.

Which brings us to RSI, a classic so many of us have grown up with.

What is this thing?

The story of RSI starts with excellent intentions (and for this version of events I’m leaning heavily on this review by the excellent Thomas Engelhardt). In this case the idea was to come up with a safer way to get the snorkel in the all important windpipe as quickly as possible to try and minimise the risk of things that should stay nestled in the gastrointestinal tract might find their way to the lungs.

And you can understand why. Serious aspiration can, sometimes, be deadly. The first piece of the puzzle was written up by Morton and Wylie way back in 1951 who described where with the patient sitting up the anaesthetist would give intravenous barbiturate then muscle relaxant and rapidly intubate them. A rapid sequence of induction and intubation. So really it’s RSII.

8 years later a description emerged of a thiopental/relaxant/40-degree head-up tilt foot-down tilt. It wasn’t for another 2 years that cricoid pressure popped up (thanks Sellick) although interestingly it included not just a bit of pre-oxygenation but also some bag-mask ventilation prior to putting the tube in.

It was another 2 years before the other classic bit of RSII became popular, with an exhortation to avoid bag-masking because of the perceived risk for gastric insufflation and hence regurgitation.

A classic technique derived from a series of “what abouts” and “I reckons”. I mean, you wouldn’t read about it. Except you just did.

That’s not to say that medicine doesn’t have space for a bit of logical derivation of good ways forward. It might just suggest that the whole approach is open to a refresh.

Re-evaluating the Likely

If the technique was designed to prevent aspiration, maybe we should start with looking at how likely this event is in a setting a bit more modern than 1951. In 1999 the epic writing team of Warner, Warner, Warner, Warner and Warner looked at 56138 patients under 18 having procedures (elective or emergency) over 12 years to see just how big this problem was. This covered 63180 procedures.

The time frame for defining aspiration was entry into the operating room until 2 hours post-anaesthetic. To score the label there had to be direct identification of bilious secretions or particulate matter in the tracheobronchial tree or new X-ray findings after an episode of regurgitation.  A total of 24 patients met the criteria.

11 of those were emergency cases so the rate in that group was 1 in 373 compared to 1 in 4544 in the elective cases. 21 of the 24 were around induction. 15 of the 24 had no symptoms develop despite the aspiration. 5 of the other 9 did need respiratory support of some kind and 3 of them needed ventilation for more than 48 hours. Well the paper says that but actually describes ventilation for 18 days, 14 days and 33 days in those cases.

And there’s the rub. It’s really very impressively rare. But then when it goes bad, the downside can be very, very down.

So fine, let’s prevent the bad thing. We’d better get on with the classic old RSII, right?

Remembering the Even More Likely

The problem with being so rigorously focussed on avoiding pulmonary aspiration that you do things like not help the patient breathe, is there are other basic functions that don’t get looked after so well. Like oxygenating.

Gencorelli et al looked at episodes of desaturation during RSI while describing the classic drugs/cricoid/no ventilation technique. Across 1070 children included they reported a 3.6% rate of desaturation to 89% or below (1.7% of the patients being in the under 80% group). Not surprisingly the under 2s were more likely to have a desaturation.

These rates are low of course and certainly lower than in some other areas of practice. Reports from emergency departments have indicated desaturation rates anywhere from 14% to 33% (with the latter reporting rates of desaturation of up to 59% in the under 2s).

So amongst the various things we’re trying to do to prevent the 1 in 400+ event are we at risk of failing on another key thing. You know? The oxygen provision thing.

What’s the alternative?

Neuhaus and team subsequently described very well their approach to RSII, which they badged as cRSII (where the “c” is for “controlled” not some other “c” word like “cheese” which wouldn’t make sense anyway but would be a good reminder that cheese is great).

They key features for them (putting to the side “lots of preparation”):

  • 20 degrees of head up (though they say only for the over 2s)
  • Suction any NG in situ.
  • Give the drugs.
  • Avoid cricoid pressure (with a few exceptions).
  • Provide gentle facemark ventilation with peak pressures of 12cmH2O.
  • Neuromuscular monitoring to ensure the muscle relaxant has really, really worked.

This last point makes a heap of sense as active regurgitation is a problem created by airway instrumentation when you don’t have adequate anaesthesia and paralysis.

It’s a big list.

Talk is cheap though, what were their results?

They report on 1001 patients They had a moderate hypoxaemia (89-80%) rate of 0.5% and a severe hypoxaemia (< 80%) rate of 0.3% and the 8 patients this represents had a median age of 0.8 years. They had 1 patient with regurgitation but no evidence of aspiration.

That’s pretty impressive.

Putting it Together

So if we accept that we should really try and optimise oxygenation, and that the risk of this is higher than the risk of aspiration then we have to accept that modifications to that original technique are reasonable. What are a few steps for practically putting it together?

1. Assess that risk of a full stomach

It might well be that we’re going to avoid cricoid most times, but there are still a few situations where that risk of aspiration is probably higher. In the Neuhaus paper they suggested achalasia, Zenker diverticulum or post-colonic interposition patients (done for oesophageal replacement) always need cricoid.

It certainly seems worth having heightened concerns in the patient with significant increases in intra-abdominal pressure.

2. Everyone sits up

Why wouldn’t you have a bit of head up? It makes sense if you’re avoiding passive regurgitation and is a good position for pre-oxygenation, facemark ventilation and intubation. I’m not quite sure why some authors have suggested the under 2s shouldn’t be head up. This is a routine option.

3. Have that suction handy

Goes without saying maybe, but I’m saying it.

4. Pre-oxygenation, but not with distress

Yes you want to pre-oxygenate. And most times you can talk kids through that and get a full 3 minutes in. Some kids will only get more distressed with oxygenation though, and insisting on pre-oxygenation only guarantees distress. Given that you’re going to apply gentle face-mask ventilation, it’s rare you need to go to the wall on this one.

And while I’m there what about apnoeic oxygenation? Well, as discussed in this post, the evidence that’s available in kids isn’t so persuasive as to suggest it should be routine. The stuff that has been done showing extended apnoeic time actually followed effective pre-oxygenation with face-mask ventilation. So as we’re going to put that tube in quickly after the same sort of effective face-mask ventilation, extending apnoeic time for minutes seems not that clinically relevant.

5. Cricoid yes or cricoid no?

Again this is a judgment call. I know plenty of anaesthetists who still prefer to start with it but with a low threshold to remove it. I’m more likely to mostly err on the side of not using it, except for those high risk of aspiration patients.

If you are going to use it, it is worth noting that, particularly in infants, the trachea is quite often more prone to distortion by cricoid pressure than you realise. Doing flexible bronchoscopy work you’re sometimes asked to manipulate the airway and I’ve seen the whole trachea get substantially compressed and distorted by seemingly innocuous manipulation. Distort it enough and you can increase the resistance to air going in and out enough to make it easier to get down to that stomach.

In addition, as covered very nicely in this review, cricoid relies on the alignment of trachea and oesophagus and the evidence is that in kids < 8 years old 45% had displacement of the oesophagus so you’d be unlikely to get compression of the oesophagus even with perfectly delivered cricoid (at least on the CT scanning mentioned).

So for the very high risk ones I’d tend to start with it (well start with it once I’m sure the kids won’t react to it going on), but that leaves almost everyone where I would’t be too concerned. And if it is on, I’d be quick to take it off if it was impeding either view or tube passage.

Maybe I included this picture of an echidna because they have a reputation for being good at waiting and not because it’s a prickly situation.

6. Wait

We’re going to take our time with face-mask ventilation and maintain oxygenation. So where’s the extreme rush getting the tube in? Being too obsessed with that step, even though you’re achieving oxygenation, is a way to end up instrumenting the airway while the patient is only lightly anaesthetised or inadequately provided with paralysis. What was that thing we’re preventing again? The regurgitation thing that’s worse if we get going while the kid is lightly anaesthetised? Oh, right. Slow down.

The description suggests using a nerve monitor. I can’t say this is routine myself, but once the muscle relaxant is onboard I do publicly note for the team I’m working with how long we’ll be waiting on the clock before we start trying to intubate. (“The clock says 09:30 now. Once it ticks over to 09:32, we’ll start with the intubation.”)

I then remind everyone that this will take an unnervingly boring period of time and they might want to come up with a good joke to fill the time.

7. Ventilating

Yes, this is a thing that’s necessary because kids desaturate quickly. Particularly the younger ones. Achieving gentle face-mask ventilation relies on really good technique with the bag in hand. Plus it’s very therapeutic to gently squeeze that bag.

7. What about parents?

This one also needs an assessment of what might help and what won’t. For lower risk kids, as a paediatric anaesthetist doing it regularly, I’d be comfortable having them along. But if it was the sort of case that was likely to be difficult, or if I was back at the training junior doctor stage, there’d be no dilemma for me. I’d tell the parents that they wouldn’t be coming in. Having them alone to help their child relax (not always a guaranteed result of having parents in) has some advantages. But the prime job is safe management of the peri-induction period. And that might mean less people around.


So those are the simple things that have shifted over the course of my time in the big wide medical world. It’s a realignment of the priorities in a way that makes the ‘R’ in ‘RSII’ look smaller and smaller so that the oxygenation is placed at the top of the tree.

Put together though it’s a reboot worth endorsing. I mean the 60s just weren’t that great, surely?



How many bits that are really important aren’t covered here? There must be some. So leave a comment. We’ll all learn.

And if you like the post and other things around the joint, maybe throw your email in the relevant spot so you’ll get an email each time a new post pops up.

This post is a cross-post from another site that this Weatherall bloke works on called Songs or Stories. It’s about paediatric anaesthesia.

That echidna pic came from flickr’s Creative Commons area and is unchanged from Duncan McCaskills’s post.

Now to the literature, because going to the direct papers is always rewarding.

That review by Engelhardt where he makes it clear what he thinks is this one:

Engelhardt T. Rapid sequence induction has no use in pediatric anesthesia. Pediatr Anesth. 2015;25:5-8. 

The paper by the anaesthetic equivalent of the Brady Bunch or something I assume is this one:

Warner MA, Warner ME, Warner DO, Warner LO, Warner JE. Perioperative Pulmonary Aspiration in Infants and Children. Anesthesiol. 1999;90:66-71. 

The benchmarking study is this one:

Gencorelli FJ, Fields RG, Litman RS. Complications during rapid sequence induction of general anesthesia in children: a benchmark study. Pediatr Anesth. 2010;20:421-4. 

The emergency department studies mentioned in passing for their demonstration of high rates of desaturation are these ones:

Long E, Sabato S, Baby FE. Endotracheal intubation in the pediatric emergency department. Pediatr Anesth. 2014;24:1204-11.

Rinderknecht AS, Mittiga MR, Meinzen-Derr J, Geis GL. Kerrey BT. Factors Associated with Oxyhemoglobin Desaturation During Rapid Sequence Intubation in a Pediatric Emergency Department: Findings from Multivariable Analyses of Video Review Data. Academic Emergency Medicine. 2014;22:431-440. 

That paper looking at controlled techniques in kids is this one:

Neuhaus D, Schmitz A, Gerber A, Weiss M. Controlled rapid sequence induction and intubation – an analysis of 1001 children. Pediatr Anesth. 2013;23:734-740.

And that other review is this one:

Newton R, Hack H. Place of rapid sequence induction in paediatric anaesthesia. BJA Educ. 2016;16:120-3.