Mounting evidence against intubating cardiac arrest patients...

[ChaseZ]

In our hospital resuscitations we have been progressively moving intubation further and further back on the priority list. The official policy is definitive airway at 10 mins. Anesthesia can usually get them tubed while compressions are going.

[Kiwiology]

I don't think there is really a role for ventilation in primary cardiac arrest at all regardless of which piece of plastic you are using.

The evidence clearly shows that intubating people makes outcomes worse, and it may be the case with supraglottic airways such as King or LMA.

I hypothesise that this is due to over-ventilation causing increased intra-thoracic pressure (and therefore causing venous return to fall) or the arteriolar constrictive effects of 100% oxygen causing myocardial perfusion to be lowered. Not giving oxygen to a normoxaemic patient with myocardial ischaemia / MI is all the rage now, so why suddenly when they are unconscious is it OK to cram 100% O2 down their gob with a bag mask?

And as for the Japanese Paramedics (called Emergency Life Saving Technician or ELST) if memory serves me right they require a Bachelors Degree (of 3-4 years duration I am unsure which it is in Japan) and that they are, or at least up until very recently were, highly restricted in the treatment they could provide; I heard it was limited to LMA, manual defibrillation, adrenaline for cardiac arrest and one or two other small things; not unlike the original Paramedics of the late 1960s in the USA.

Having said that, I did read somewhere that some were allowed to tube people (clearly that must now be the case if this study is out) and Japan already has Doctor staffed MICU Ambulances anyway.

[systemet]

This is an interesting study. It's huge (n = 650,000) -- compare that to something like ROC-PRIMED (n=10,000), or the San Diego RSI Trial (n = 200 trial patients, 627 controls). So it's well-powered to detect even a small difference.

The majority of the advanced airway patients were SGAs (~ 85%) vs. ETT. Both advanced airway subgroups showed harm.

There are potential confounders there -- it could be that there's a survival bias for patients in the BVM group, if they regained circulation prior to any airway attempts. Nowadays a lot of services are deferring advanced airway until 5-10 minutes after initiation, so if the patient regains circulation prior to that, they probably end up in the BVM group, and bias that cohort towards survival / better neurological outcome. But the magnitude of the difference is pretty big, The 95% confidence interval OR for intubation is 0.37-0.41, and for SGA 0.36-0.40, suggesting patients in the BVM group were 2.5X more likely to survive with good neurological outcome.

Kiwi suggests a couple of plausible mechanisms. As the authors suggest, there needs to be an RCT for advanced airway in cardiac arrest now.

One potential confounder that the researchers didn't identify was the use of epinephrine. 10.6% of the patients in the advanced airway group received epinephrine, versus 2.9% in the BVM group. Now this makes sense, as you'd assume that less of the providers using BVM would be authorised to give epinephrine -- but what if this is the variable at play, and not the method of ventilation?

As they say, it's time for an RCT. Intubation vs. no-intubation in cardiac arrest. Can you imagine walking in to the patient's house, directing a couple of firefighters to move the living room table out of the way, putting the patient on the floor and going to open an envelope with an 'A' or a 'B' in it? Interesting times ahead, potentially. I wonder what the ethics of a waiver of informed consent are for this situation?

-------------------------------------------------------------------------------------------------------------------------------------------

For people less familiar with research trials:

* An n-value is an overly complicated way of saying how many patients were in a study. The higher the n-value, that is, the greater the number of people in the study, the more it's able to detect small differences. If we see a 3% difference in a comparison between two groups of 10 patients each, it's unlikely to be statistically significant -- it could just be chance. When we have 650,000 people, a 3% difference is often more meaningful. This is sometimes referred to as the "power" of a study.

* Significance testing is a statistical tool that looks at how much data varies, and lets you estimate how much of a difference between two groups might be due to chance. The standard in most biomedical trials is that there should be a less than 1 in 20 chance that any difference observed is due to random variation (sometimes referred to as p = 0.05).

* Confounders are factors that may influence the outcome of a study. If I try to investigate whether coffee causes cancer, I need to make sure that I take account of how many people in each group smoke. If I simply compare coffee-drinkers to non-drinkers, I'll find an increased cancer rate in the coffee drinking group as coffee drinkers are more likely to be smokers than people who don't drink coffee. This may seem obvious, but is frequently missed.

* A confidence interval is an estimate of the precision of a number. It tells us how confident we are in a given number, so a 95% confidence interval gives us a number that is likely to be right 19 times out of 20. Sometimes you read a study that shows a 20% decrease in mortality, but then you see that the confidence interval is from -10% to 50%, i.e. there's a greater than 1 in 20 chance that the intervention may actually save the lives of 10% of the patients enrolled, there's also a more than 1 in 20 chance that the mortality effect is far worse than 20%, i.e. 50%.

* An Odds ratio shows an association between two factors. If the odds ratio is 1, then there's no difference. In this study group the odds ratio for intubation was 0.40, that is if you were able to take a time machine, take 10 of the BVM patients and intubate them instead, you'd expect four of the patients that would have otherwise survived to have died.

* An RCT is a randomised controlled trial, where you roll a dice, flip a coin, open an envelope or pull a number off a computer to assign patients to a control or intervention group. What the authors of this study suggest is that we take a group of patients, and randomly assign them to either have intubation / SGA insertion or to BVM-only.\

* WOIC is "waiver of informed consent" -- when you do a clinical trial you have to explain what you're doing and the risks and benefits to the patients, if your patients are unconscious (or, in this case, dead, or some version of it) there's a process whereby you have to inform the community, and apply to an ethical committee to be allowed to study people without first obtaining verbal consent, which is impossible in this setting.

I wonder why the outcome was worse in pts with an advanced airway. Is this related to interruption of compressions to secure an airway?

This is possible. They didn't capture that data, so we won't know from this study. One would assume that that would be more of an issue for the ETT group than the SGA group, but they don't describe how many intubation attempts there were, if any, before SGA insertion (I think there was a simiilar problem with ROC-PRIMED).

When you look at the group compositions, the BVM group had a much higher rate of VF/VT as the presenting rhythm (20% v. 6-8%), bystander witnessed arrest (5% v. 2%), and EMS witnessed arrest (9% v. 2-3%), and a much shorter time to ROSC (6 min v.14 *although the CIs overlap here, so it's not significant). However, they've adjusted for this, so it's probably not a big deal.

[triemal04]

I don't get it really. Why do we need to stop compressions to intubate? I mean, I used it stop them years ago but then I thought..."Hey, why am I stopping when I can still see all my landmarks?" so I decided to give it a go. Since then I've only stopped compressions a couple of times, and only for about 3 - 5 seconds to pass the tube. I still intubate farther down the algorythm than the old days, but stopping CPR to do it just isn't always a requirement.

You shouldn't need to in almost every case. Part of the problem may be that it used to be/probably still is taught that compressions are held during intubation, or it's never mentioned at all. I know that many areas are advocating not stopping, but it will still take awhile before that becomes standard across the board, and as ETI is moved away from as a whole it will likely get mentioned less and less.

Also factor in that if someone is only intubating a few times a year (which I'd say is all to common in the US) they may not be thinking about that and/or have the confidence to try it.

More needs to be known about this (intubation during cardiac arrest) before a definative answer can be given. The skill level of the providers taking part in the study...the number of required attempts...how long if at all compressions were stopped for...when it was done...what was the presenting rhythm...initial downtime...CPR before arrival or not...cause of the arrest...all that and more needs to be known before a real answer is available. And to be really accurate, the variables need to be the same for each group participating.

This is an interesting study. It's huge (n = 650,000) -- compare that to something like ROC-PRIMED (n=10,000), or the San Diego RSI Trial (n = 200 trial patients, 627 controls). So it's well-powered to detect even a small difference.

* WOIC is "waiver of informed consent" -- when you do a clinical trial you have to explain what you're doing and the risks and benefits to the patients, if your patients are unconscious (or, in this case, dead, or some version of it) there's a process whereby you have to inform the community, and apply to an ethical committee to be allowed to study people without first obtaining verbal consent, which is impossible in this setting.

However, they've adjusted for this, so it's probably not a big deal.

I agree it's time for a RCT. I think it can be done, and done in a way that many of the potential variables are accounted for and removed. As for as the ethics, I doubt it'll be a problem. ROC has been doing studies that sometimes push the boundaries for awhile; the way they've got around the ethics commission is to announce in the news media in all their locations that such and such a study will be starting, and give people the ability to opt out by requesting a bracelet to identify themselves.

And given that they got approval to study the efficacy of lidocaine vs amiodarone vs saline in cardiac arrest...I think that getting approval to look into intubation wouldn't be to hard. And appropriate.

[systemet]

This is a great thread. Thanks for participating everyone!

[Kiwiology]

I agree this is a great topic and a very relevant issue

In New Zed it has been decreed that regardless of problem, if an LMA is in place and working well that the patient should not be intubated (unless they will benefit from RSI and control of carboxaemia e.g. TBI) and specifically in cardiac arrest, ventilation in primary arrest is at or one above the bottom of the list, above perhaps only IV drugs from the bottom of priorities.

I cannot help but wonder if a more appropriate study would be ventilation vs. passive oxygenation?

I know there were some studies from Arizona that looked very promising where the patient was not actively ventilated?

[BAYAMedic]

Thanks to every one who informed me of the educational requirements of the Japanese system, I had no idea. This is by far the best post I have seen on here in awhile. Good work all.

BAYAMedic

[mikeymedic1984]

1. First and foremost, I still think the stats for out of hospital arrest are around a 95-99% death rate, so it really does not matter what airway device you choose to use or not use. The odds are against you despite your best efforts and technology. I could probably produce a study that shows rubbing "deer antler spray" over their heart works better than intubation or epinephrine. Just saying.

I think we spend too much time valuing our jobs and existence on these patients, lets spend more time on patients that we can have a positive outcome on.

[Kiwiology]

1. First and foremost, I still think the stats for out of hospital arrest are around a 95-99% death rate, so it really does not matter what airway device you choose to use or not use. The odds are against you despite your best efforts and technology. I could probably produce a study that shows rubbing "deer antler spray" over their heart works better than intubation or epinephrine. Just saying.

I think we spend too much time valuing our jobs and existence on these patients, lets spend more time on patients that we can have a positive outcome on.

Whilst I completely agree with what you are saying I think it is still a valid question; if what is currently practiced is somehow harming patients somehow regardless of how dead the overwhelming majority of them are going to remain then it deserves attention.

I still say that ventilation is not a priority and should probably be removed from primary cardiac arrest until ROSC is achieved; I've been saying forever and a day that adrenaline is probably harmful and should be abandoned.

[mikeymedic1984]

I am "old school" and disagree. I dont think the "device" used matters, I still believe "proper ventilation does". I can not tell you how many times I have watched Para-Gods dig in someone's throat for minutes in an attempt to get the tube in, versus just ventilating with a BVM. I do not see how the deprevation of oxygen improves anyone's survival chances. I have been around long enough to see "new studies" that prove what we have been doing for years is now somehow wrong, but the success rates never improve. I say go back to two rounds of Epi, Bicarb, and D50 for any arrest; it worked just as good as what we are doing today.