Author: Tucker Matthews
Four more papers about airways!
Paper 1: Nauka PC, Moskowitz A, Fein DG. Appraising First-Pass Success: During Emergency Airway Management, What Does It Mean to Be Successful? Ann Am Thorac Soc. 2023 Jan;20(1):21-23. doi: 10.1513/AnnalsATS.202208-661VP. PMID: 36227712; PMCID: PMC9819272.
This paper is pretty much mandatory for anyone interested in airway literature.
For whatever reason we have seemingly decided that first pass success is the metric that we’re all going to use in airway literature. That being said, there isn’t even a clearcut definition of what this term means. The traditional use of the term equates to a single insertion of the blade into the oropharynx, but it actually doesn’t define anything in regards to endotracheal tube insertion attempts. And it also, of course, doesn’t specify anything in terms of time required for successful intubation. So a provider could theoretically keep the blade in the mouth and just keep repositioning without full removal in between attempts, and that would still count as a first attempt. Or a provider could stand there ramming the tube against the aretynoids or vocal cords for minutes, and still count it as a first pass success if they eventually get it. Even if the patient is profusely bleeding from all of the trauma.
More recently, people are trying to include a single ETT insertion as part of the definition, which probably makes more sense, but then leads to another issue with the term: does a first pass success actually matter?
There are a few general things that we all probably agree are bad. First, we don’t want to take forever and leave the patient without oxygenation/ventilation. And secondly, we don’t want to be ramming tubes and blades around and causing a bunch of direct trauma. And so generally, first pass success rates are attempting to give a quantifiable datapoint for avoiding those negative effects.
The problem though, is that it’s only one part of effective airway management. For relatively easy airways, first pass success is definitely a good goal. But for difficult airways, it may just be more important to eventually get the tube in, and to avoid bad outcomes like hypoxia, hemodynamic instability, or airway trauma. So, the authors provide a helpful example, that if a provider is taking a longer time to get a tube in on the first attempt, leading to prolonged hypoxia, that will still show up as a successful first pass, and thus will mislead the data.
The authors also call for the importance of determining other metrics that could be widely used in addition to first pass success rates. I love this idea. They suggest metrics representing the possible adverse outcomes (things like hypoxia, vasopressor use, etc.). They also discuss a recent study that used a composite outcome, though that obviously comes with the secondary complication that we’d have to figure out a good composite outcome to use. Finally, they discuss the all-important topic of using long-term patient-centric outcomes like mortality. These data points are often most useful in larger studies, but are definitely another area that studies should include.
So all in all, I’m psyched that people are talking about the limits of this metric, and hopeful that we can start having a little more nuance to our data collection.
Paper 2: Garcia SI, Sandefur BJ, Campbell RL, Driver BE, April MD, Carlson JN, Walls RM, Brown CA 3rd. First-Attempt Intubation Success Among Emergency Medicine Trainees by Laryngoscopic Device and Training Year: A National Emergency Airway Registry Study. Ann Emerg Med. 2023 Jan 18:S0196-0644(22)01202-1. doi: 10.1016/j.annemergmed.2022.10.019. Epub ahead of print. PMID: 36669924.
Fittingly covering this paper after talking about the limits of first pass success rates. Here, an American group wanted to look at first pass success rates in residents at different PGY levels of training. It’s a prospective observational trial using the NEAR database, which included 25 academic medical centers.
So the NEAR database is awesome, and there’s tons of great information that has been published from it. It’s just also important for folks to remember the limitations of the database, in particular, the fact that all of this is observational, and subject to loads of bias. And like almost all airway studies, NEAR studies also have the issue of being based on self-reported metrics. And as someone who does the airway QI at my institution, I unfortunately am aware that many individuals don’t report things as accurately as they maybe should (that’s me trying to be diplomatic). I just want to put that out there before talking more with this study.
In total, they used 15,204 intubations, most of which were for medical indications, and which were actually spread pretty evenly among DL, hyperangulated VL, and standard geometric VL. There were actually far more performed by senior residents, which I found pretty surprising, given that at my program, we always prioritize junior residents in getting procedures.
About two-thirds of the patients didn’t have Mallampati scores assessed, so it’s pretty limited as a data point, but interestingly, Mallampati 4 airways seem to have more frequent use of hyperangulated blades, though only slightly.
Also really interesting is that there’s a trend for standard geometric VL to be used in residents PGY-3 and above.
Their primary outcome of first pass success rates showed decreased rates among DL intubations over all PGY training levels compared with both types of VL. Those differences were pretty impressive too, at just under 10% for each level. Not surprisingly, the first pass success rates improved with PGY level among all three methods. Also, adverse events tend to decrease in likelihood with PGY level, though there was some variability here, and this is as a composite outcome of all adverse events.
This is all really fascinating. Again, the limitations from before still apply, but it’s a great collection of data. One other problem though is the different in airway training quality/style at different residencies. Some places are almost definitely teaching one modality far better than others, and so there will be pretty large biases with which patients those residents are choosing to use their backup modalities on. For example, a resident who feels most comfortable with hyperangulated VL, is probably going to use that on all of their difficult patients, and reserve their DL and SG-VL for anticipated easy airways. So that will skew the results.
The authors talk a bit about how this data helps push the suggestion that residents should probably be trained on standard geometric VL setups, since they seem to have great first pass metrics, but that they also will help trainees to be prepared for DL, should they need to. I generally agree with that statement. I’m not sure what the point of training on hyperangulated would be, since it would then take a complete separate effort to learn standard geometric skills. Trainees should all learn hyperangulated approaches as well, but probably not until relative mastery of standard geometric blades.
At the end of the day, we have a lot of room to investigate the best training practices for intubation, and hopefully, we’ll start to see more studies investigating the topic. This paper is definitely another great start though for determining how best to train our residents.
Paper 3: Khorsand S, Chin J, Rice J, Bughrara N, Myatra SN, Karamchandani K. Role of Point-of-Care Ultrasound in Emergency Airway Management Outside the Operating Room. Anesth Analg. 2023 Jan 24. doi: 10.1213/ANE.0000000000006371. Epub ahead of print. PMID: 36693019.
This paper is a review by some anesthesiologists that discusses the use of point of care ultrasound for emergent intubation.
One area they discuss is with fluid-responsiveness. As the PREPARE II trial has ignited a bunch of conversations about whether or not to use fluids peri-intubation, these authors make the valid point that maybe ultrasound assessments of fluid responsiveness could be the next step forward. This way, patients who demonstrate responsiveness would get the fluids to potentially mitigate peri-intubation hypotension, while those not responsive would avoid the over-resuscitation that has shown poorer outcomes as well.
They also talk a bunch about the use of POCUS for general assessment of hypotension with a RUSH exam, which is nice, but for emergency department patients, we should really already have figured out whether they have a hemodynamically important pericardial effusion or hemothorax, etc. So this doesn’t feel as important of a role of POCUS, at least not in the minutes just before intubation. I would hope that any undifferentiated hypotensive patient that is getting intubated has already had the equivalent of a RUSH exam.
The authors also discuss the use of lung ultrasound, but they pretty much just discuss it irrespective of intubation. And I’m just not sure in an emergent or urgent intubation, if ultrasounding the lungs is really your best bet, unless you have some concern for a pneumothorax or pleural effusion (or maybe pulmonary edema). So maybe this is really only helpful for the undifferentiated patient who you don’t have much time to do any workup before intubation.
In the next section they discuss measuring the gastric antral cross-sectional area to evaluate aspiration risk. I’m just not going to go into this, but there is no way that I’m doing this. They also talk about using it for determining likelihood of difficult anatomy, but it’s unconvincing as well.
There’s a brief segment on use of ultrasound for ETT placement confirmation, with the ability to visualize the tube in the trachea and also to see bilateral lung-sliding. I like this possible use personally, since the only other reliable method is waveform capnography, which isn’t always available quickly.
This paper was okay. It was more of an outline of all of the proposed uses of ultrasound revolving around intubation, but it really didn’t discuss what’s most important, and that’s how ultrasound is an improvement over other modes of evaluation. I find ultrasound to be fairly easy to use, but I only do it when it actually changes management and outcomes. And I’ve found it frustrating to learn all about the uses of ultrasound that wind up just taking extra time helping us with diagnoses that don’t impact care. Here, they just spat out every possible use, but it didn’t feel like they had as much of a concept of practicability. So personally, I still think I’ll use it on some cases pre-intubation if I have a specific clinical question, and in some cases post-intubation for confirmation if capnography isn’t readily available. Otherwise, I don’t think it will be a major part of my peri-intubation care, outside of using it for hemodynamic assessments if the patient is unstable (which I would be doing even if not intubating).
Paper 4: Steffen R, Hischier S, Roten FM, Huber M, Knapp J. Airway management during ongoing chest compressions-direct vs. video laryngoscopy. A randomised manikin study. PLoS One. 2023 Feb 9;18(2):e0281186. doi: 10.1371/journal.pone.0281186. PMID: 36757942; PMCID: PMC9910718.
This is a study out of Switzerland, comparing VL versus DL for intubation with ongoing chest compressions using a manikin model.
Importantly, this was using anesthesiologists and anesthesia nurses, which isn’t a perfect comparison for ED providers. They also don’t say conclusively, but it seems like the VL attempts were all with standard geometric blades.
It’s important to minimize chest compression interruptions during CPR, and it unfortunately seems in my experience that providers often forego that recommendation when intubating. Instead, we should probably be completely comfortable intubating while the chest compressions are actively occurring, and not just assuming that we can quickly slip in a tube during a pulse check.
Here, they used a manikin that was getting chest compressions by a mechanical device, and I’m just going to assume it accurately modeled a real-life code.
I appreciate that they used both first-pass success rates along with time until first breath. That helps avoid some of the issues with solely using FPS as the outcome, and in a manikin study like this, there is no reason not to have additional outcomes included.
They only have 28 data points for each group (DL vs VL), which really isn’t a lot, but the data that they show here is pretty clear-cut. VL had a 100% FPS rate, while DL was only at 67.9%. That only led to a 2.5 second difference in time for intubation though, which may further the point that first pass success means different things in DL and VL setups. That being said, this definitely suggests pretty effectively that VL works better in this model, though, I don’t think we can really extend this to real patients until a study conclusively shows that. The authors are well-aware of that and point it out in their discussion.
For myself, I am fully on the VL train, and I think this is a scenario where it’s hard to argue against it. If we’re really trying to minimize chest compression interruptions, whoever is intubating can even have the VL blade inserted into the mouth before the pulse check, beginning to optimize their view, so that during the actual check, they are able to use those ten seconds to insert the actual tube. And if it takes more than those 10 seconds to get the tube in, there’s probably something else that needs to be optimized, and the patient should be bagged in the meantime.