The decision to intubate – the follow up

[Image Wikipedia Commons]

A few days ago I posted a few cases to try and elicit the reasoning behind why we intubate people. Check out the comments to get a feel for what people were saying. Here’s some follow up. Though first a bit of preamble on context.

For the US readers, it takes a lot to get into an ICU over here. Most of our non-invasive ventilation and DKA patients and lots of other folk that you might send to the ‘ICU’ go to a ward here. From what I read/listen to, there are a lot more patients going to ICUs in the US, and a lot less sick patients in the ICU.

For starters, intubation is of course just one part of the care. It’s all well and good intubating an obtunded leaking AAA but it kind of misses the point. Intubating someone, while a vital part of management, is rarely going to fix the problem in isolation. Intubation, however, does tend to set in motion a whole series of clinical procedures and (if they remain intubated) dictates disposition to an ICU setting.

Secondly, there are some genuine and major issues surrounding bed availability in most of the UK/Ireland. Critical care beds are at a premium and it’s very common to have intubated patients in the ED for prolonged periods (by which I mean overnight, rather than the 2 or 3 days that I heard can happen in the US) and often having to be transferred to other hospitals where a bed is available. Just as a full ED is bad for patients, a full ICU is going to be bad for patients and those patients trying to get in.

Just like in an overcrowded ICU, there are rarely the nursing staff or medical staff to look after the ongoing needs of a ventilated patient in the ED for prolonged periods and this is of course, by no means ideal.

Decisions about intubation (which implies that we think they need ongoing critical care – remember that might change when we find out they’ve had a massive, non-survivable ICH) are influenced by all kinds of ‘non-clinical’ needs like these. Registrar to registrar referrals at 4am are different from 10am referrals when all the bosses are around. In the rather dysfunctional, ‘sickest looked after by the thickest’  model (UK/Ireland) where trainees continue to provide the vast majority of care there are a lot of pragmatic decisions made that perhaps would be made differently if the only people involved were consultants at 10am following a nice latte and a ward round on a half-empty ICU. You may disagree with that, but I think that’s the honest view of things.

In my experience decisions regarding intubation and ICU admission (and indeed most major decision in medicine) are frequently different when there are consultant level staff involved, on both sides – ED and ICU. Let me put it this way, the outcome (eg intubation or ICU admission) might be the same but the process to reach that decision will be somewhat different (usually quicker and more convincing) when the people deciding it are on a higher pay grade than me.

A remarkably good job is done in this non-ideal situation and it is rare that anyone actually comes to harm from it. But it is frankly naive to think that we all (as trainees or consultants) make wonderful, unbiased purely clinical decisions all of the time.

So… in this resource limited, pretty screwed up situation that I find myself, the decision to intubate is a fairly major one. In my current dept, we seem to have a good relationship with ICU and they seem more supportive than usual of the ED performing RSIs. Having said that, my threshold for intubating someone is still pretty high.

I don’t know what you all did with the 3 cases I presented but I have managed 3 very similar cases, all without intubation.

Case 1 

This person has what Weingart calls SCAPE. (creepily Dave Menzies used the exact same phrasing in his comment…) This is acute pulmonary oedema. With NIV and massive doses of nitrates it is rare for me to need to intubate someone like this. They look much sicker than lots of other patients that you might intubate and the nurses will stare at you in shock when you say you don’t want to intubate them.

Neither GCS, pH, resp rate, even hypoxia tip your hand to intubate these people. In another condition then of course you would intubate them if they were this sick but the big difference is that this is easily (well relatively easily) reversible, and rapidly reversible without intubation.

Most of the ones I’ve seen present at 5am (why?) and you can frequently have them drinking tea by 9am if you play your cards right.

Case 2

this person has fairly significant cocaine toxicity. If he seized then I’d intubate him but otherwise what this man needs is neither ventilation nor airway protection but massive doses of benzos. I’m yet to break 200mg of diazepam yet but I’ve got close. If he got too sedated, then of course, he may need an advanced airway but most of the time you can get away with it.

A lot of people mentioned that he’ll have renal failure but CK doesn’t seem to predict full blown AKI that well that I’m aware of. With teenage kidneys, good fluid resus and a catheter I’m not sure he’ll need dialysis.

In general I’ve found that these guys don’t get floored with sedation. I stand at the bedside cracking open amps of diazepam, doubling the dose each time until the diaphoresis stops. Hardly the most scientific of end point but a ‘conscious sedation’ level comes about the same time as the sweating stops, then I let them sleep it off. Treat till they’re dry?

Case 3

this is the tricky one I think. 8 times out of 10 this chap is just seizing from missed meds and too much booze. But of course he could well have a big bleed in there causing it. The decision to CT him is easy. The decision to intubate him isn’t as clear cut. The key thing that isn’t mentioned in the vignette is treatment. This man may be in status but he has not yet had any treatment. If he seizes after reasonable doses of benzos and phenytoin then yes he gets a tube. If there are genuine airway concerns (as opposed to simply “GCS less than 8 concerns”) then he gets a tube too.

Some final thoughts

It is, of course, perfectly reasonable to intubate all 3 of these patients. Lots of you would and it would be hard to disagree. And, of course, just cause you might ‘get away with it’ by not intubating them, doesn’t mean that you’ve done a good job – I don’t think we need a specificity of 100% for intubation, it’s likely more important that our sensitivity for intubation is on the high end. Airway protection is often cited as reason to intubate but whether or not the airway is safe or not can only be determined in retrospect – did they obstruct or aspirate – if they didn’t then you could argue they never needed the airway protected in the first place.

Perhaps the only thing I’m trying to illustrate here is that the decision to intubate is a tricky one. Bottom line, when you’re with the patient in resus it’s often your call (depending on whether you need to outsource your airway management/ICU admission to someone else.)

Someone may disagree with you, sometimes for good reasons, sometimes for shitty ones. The techniques of making things happen that Cliff teaches so well, are quite simply vital to the care of your patient when you’re coming up against some less than pristine clinical decision making on the other end of the phone. You won’t always need it though – there are plenty of patients where my initial inclination has been to intubate but with the presence and discussion with ICU team, there’s been a collaborative decision to manage without intubation.

This is just scratching the surface of a really deep and complex question. To think that it all boils down to a decision to put plastic in the trachea is a little bit simplistic so help me out in the comments with what else you think makes the decision to intubate.

PS The general trend in your responses for intubation was:

1) No

2) Yes though some wouldn’t

3) Yes


The decision to intubate

[Image Wikipedia Commons]

[This has been floating about as a draft post here for ages, but Cliff has finally inspired me to put it out there and have the discussion following his recent post on The tongue-in-cheek non-intubation check list]

This is one of the trickiest decisions in EM in my opinion. Now for lots of patients it’s really easy: the severe sepsis with white out pneumonia or the severe head injury.

However, lots of people fall somewhere in between. In the past 6 months I’ve treated a number of patients where the decision could have gone either way. For example, here’s some theoretical cases. I’ll let you work out the (hopefully) obvious diagnoses.

Case 1

72 male presents via ambulance at 0500 having woken from sleep with shortness of breath.

He is clearly unwell with a systolic BP of 180, HR of 180 a horribly broad LBBB ECG with a rate of 110. He is diaphoretic, agitated but responsive to voice with sats of 70 on oxygen. His RR is 30 and the chest is wet like an Irish summer. GCS is single figures.

pH 7.0 pCO2 8 pO2 6 on arrival (they’re in kPa if you’re wondering)

Do they get the blue cigar?

Case 2

A young male presents via ambulance after being found agitated and delirious trying to catch a bus.

He is massively diaphoretic, agitated, tachy at 170 with a normal QRS but a dominant R in aVR. He is moderately hypertensive at 150 systolic. He tells you he wants to leave so he can catch his bus.

His lactate is 20 and he is producing dark coloured urine and has a CK greater than 100000.

Do they get the blue cigar?

Case 3

A middle aged male, with a preponderance to drink strong beverage and a tendency to have seizures, is found on the floor of a hostel seizing with a wound and grossly swollen right peri-orbital area. He has two further seizures on route to hospital and one in the department. It is now 30 mins from the first seizure and he has not returned to baseline.

Do they get the blue cigar?

Let me know in the google form and give some reasons in the comments. You’ll all want more information but you’re not getting any. I’ll give some follow up next week.

Stop putting IV cannulae in the 2nd ICS for tension PTX

Though I suspect this is old news by now, I’ve just read a couple of papers on it and thought I’d share my thoughts. They’re both by the very smart and talented Kenji Inaba.


  • where should we put a cannula if we want to decompress a tension pneumothorax?

Inaba, Kenji, Bernardino C Branco, Marc Eckstein, David V Shatz, Matthew J Martin, Donald J Green, Thomas T Noguchi, and Demetrios Demetriades. “Optimal Positioning for Emergent Needle Thoracostomy: a Cadaver-Based Study.” The Journal of Trauma: Injury, Infection, and Critical Care 71, no. 5: 1099–1103. PMID 22071914


  • 20 fresh frozen cadavers
  • 14G 5cm cannulae placed in 2nd ICS and 5th ICS
  • throacotomy to assess pleural puncture (considered a +ve outcome)
  • chest wall thickness measured in each cadaver at each puncture position


  • total of 80 punctures
  • all succeeded in the 5th ICS; 57% succeeded in the 2nd ICS


Inaba, Kenji, Crystal Ives, Kelsey McClure, Bernardino C Branco, Marc Eckstein, David Shatz, Matthew J Martin, Sravanthi Reddy, and Demetrios Demetriades. “Radiologic Evaluation of Alternative Sites for Needle Decompression of Tension Pneumothorax..” Archives of Surgery (Chicago, Ill. : 1960) 147, no. 9: 813–818. PMID 22987168


  • chart and image review of all their trauma pts over a year who got a chest CT
  • split them into BMI quintiles
  • measured the chest wall thickness on CT  at the 2nd ICS and 5th ICS


  • 680 pts
  • 46mm in the 2nd ICS; 33mm in the 5th ICS
  • half of the pts had a chest thicker than 50mm at the 2nd ICS


  • if a standard 14 G cannula is 50 mm then we’re going to fail to drain a whole bunch in the 2nd ICS


I’ve drained a number of tension PTX in the past – it’s usually fun to get the hiss  – all have been in the ED and quickly followed by a chest drain. I’ve received a number from the pre-hospital environment that probably worked initially but now the cannula has kinked and a second cannula shows a second release of air.

All this is a little bit silly I think – as I’m now convinced that we just need to get on and do the bloody drain (or at least make a cut and stick a finger in). If the patient is crashing in front of you and the quickest option is to place a cannula then great – just don’t get lulled into thinking you’ve fixed the problem.


Haldun Akoglu left a comment about a paper that he just published (PMID 23116647). It was a review of 150 CT scans of people with PTX, mainly traumatic. They found that a 5cm cannula would also fail a lot of the time but failed to find a thinner chest wall at the 5th ICS when compared with the 2nd. I’m not sure why they they didn’t find a difference when Inaba did but either way 5cm cannulas are not the way to go.