Author: Tucker Matthews, MD
Paper 1: Singh N, Sawyer T, Johnston LC, et al. Impact of multiple intubation attempts on adverse tracheal intubation associated events in neonates: a report from the NEAR4NEOS. Journal of Perinatology : Official Journal of the California Perinatal Association. 2022 Sep;42(9):1221-1227. DOI: 10.1038/s41372-022-01484-5. PMID: 35982243.
Methods: Using a prospective airway registry database (like the adult NEAR database) of delivery room and NICU intubations, the authors retrospectively analyzed for associations between intubation attempt number and a variety of variables.
-52% success rate on 1st attempt, with 78% success after 2 attempts
-Gestational age less than 30 weeks was associated with increased attempts
-The comorbidity more associated with increased attempts was acute respiratory failure (though this feels like it is more of an indication than a comorbidity, and so could be just indicative of intubations that were less expected and had less time for preparation)
-The indication most related to increased attempts was surfactant administration, while unplanned extubation was the indication that seemed to be more likely to have success within 1 or 2 attempts
-When residents attempted the first trial, there was a strong link with increased eventual attempts until successful intubation
-Video laryngoscopy and use of a paralytic were both associated with first pass success
-42% of intubations had a severe oxygen desaturation (and this unsurprisingly increased with the number of attempts)
-Also not surprisingly, more attempts were associated with pretty much every adverse event type
The NEAR databases (and any airway registry) are incredible, but have obvious limitations. None of this is randomized, and so the decisions for the specifics behind the technique used for each intubation was almost certainly effected by the particular patient case. Unfortunately the authors don’t really talk about that in the limitations, and instead sort of suggest that we should all start to use VL and a paralytic on each neonate intubation. This just isn’t what the data say. Maybe it’s true, but in all likelihood, the decision for VL vs DL was made in many instances based on some thought of patient anatomy or physiology. Same with the decision for paralytic use. For example, you wouldn’t paralyze a patient in cardiac arrest, and that patient will also almost certainly have some pretty severe adverse events, no matter how successful you are with the actual intubation.
Overall though, these database studies are always super interesting. I’ve never tubed an infant, and so I have really no concept for how hard it would be, but I’m pretty amazed at how low those first pass success rates are, and how many of these kids had significant desaturations. I would also be so curious as to how good EM providers would be here, since we are generally pretty good at intubating, even if we don’t do it on babies a lot.
I’m guessing that surfactant administration is a risky procedure, and there was some adverse event during the procedure that necessitated intubation. And maybe those that needed surfactant had particularly tough anatomy, but I don’t really know, and it doesn’t seem particularly EM relevant.
Neonates have the dangerous combination of unique anatomy in addition to vulnerability to any sort of insult. As such, it’s important to try and optimize these airways. This paper provides some data about these neonatal intubations, while also generating some hypotheses about potential strategies to improve first pass success.
Paper 2: Vahdatpour, C.A., Ryan, J.J., Zimmerman, J.M. et al. Advanced airway management and respiratory care in decompensated pulmonary hypertension. Heart Fail Rev 27, 1807–1817 (2022). https://doi.org/10.1007/s10741-021-10168-9
Methods: This is a review/expert opinion.
What they say:
Background: Pulmonary hypertension occurs when there is increased right ventricular afterload, leading to myocardial hypertrophy and subsequent RV failure. In decompensated PH, there is decreased cardiac output and/or increased CVP secondary to this increased RV afterload. The authors propose a definition using three criteria: hypotension, end organ dysfunction, cardiac cath or echo abnormalities. Intubating these patient’s is extremely dangerous, and there is super high rates of complication. In particular, the combination of sedatives and positive pressure ventilation can be devastating to already tenuous hemodynamic statuses.
-Avoiding supine positioning can minimize preload, so consider maintaining incline
-Have a goals of care discussion
-There’s a really scary tradeoff between wanting to avoid atelectasis (which increases RV afterload through hypoxic vasoconstriction and increased pulmonary pressures) and wanting to avoid the elevated intrathoracic pressure from NIPPV, so pre-oxygenate carefully with NIPPV. It’s probably better to just use HFNC (and continue it for apneic oxygenation)
-Have an arterial line and central line
-Have a pulmonary artery catheter (maybe not in the ED…)
-MAP goals of >65mmHg throughout peri-intubation
-Consider delayed sequence or awake intubation
-Once ventilated, use low PEEP and low volumes, and avoid hypercarbia
-Don’t use propofol
-Both etomidate and ketamine are reasonable choices, though ketamine does have some myocardial depressant properties
-Have inhaled pulmonary vasodilators ready, and if you’re already giving them pre-intubation, continue them through the HFNC/NC used for pre-oxygenation and apneic oxygenation
-Have vasopressors ready, ideally vasopressin or maybe norepinephrine
-Have inotropes ready: dobutamine or milrinone
My thoughts: These patients scare the hell out of all of us, so it’s good to have a laid out plan of how to approach them. I think if this happens to me in the ED, I’m calling for help immediately from a cardiologist or intensivist, who can help me with all the considerations behind using some of these medications.
Paper 3: Lee, G.T., Park, J.E., Woo, Sy. et al. Defining the learning curve for endotracheal intubation in the emergency department. Sci Rep 12, 14903 (2022). https://doi.org/10.1038/s41598-022-19337-8
Methods: This was a single-center (South Korea) retrospective analysis from their institutional airway registry. They took every intubation done by an EM resident (who started residency during their range of data collection so that they could also be sure to monitor this resident’s progress through their training), and had observers record information about each intubation (no self-reporting).
-22 total trainees, of whom only 1 had more than 5 intubations prior to starting in the program
-Overall first attempt success (FAS) of 78.6% (given teaching center, maybe this is expected)
-25.5% were anticipated difficult airways (FAS rate in these was 66.2%, as opposed to 82.9% in the remainder)
-37.2% were “crash approach”, meaning they didn’t need medications
-52.6% RSI, 6.3% sedative only
-Not surprisingly, a pretty big jump in FAS rate from year 1 to 2 of training (68.7% to 79.1%), but then a more gradual improvement.
-Based on their model from their data, 119 cases were needed to achieve a FAS probability of 85% or greater
-68 intubations were needed for 80%, and 88 were needed for a 90% success within two attempts
-Only 35 intubations are required by ACGME, though multiple studies in the anesthesiology field have shown a minimum of 50 are needed for some proficiency
-I like the choice of 85% for proficiency here, because other studies have indicated that ED FAS rates are somewhere in the ballpark nationwide
-They show their model to calculate the relationship between cumulative number of intubations and the predicted probability of FAS, and it just keeps trending up towards 100% FAS, which is flawed, because we know there is a reasonable upper limit, since so few providers are needing just one attempt for those really difficult intubations. And since none of their trainees were getting 200 tubes, or any amount in that area where we may see a plateau, it’s a pretty unsupported assumption they make about FAS rates past a certain value. Unfortunately, they don’t show the data (or I can’t find it) showing how many tubes each trainee got, other than saying the median number of cases was 82.5. In the supplementary information they do have a graph that shows a plateau with one of their models, and it seems to show a predicted plateau of first pass success rate at around 150 intubations.
-They also aren’t following their trainees going forward, we don’t really know what happens to these skills as they leave training. Maybe some residents will continue to improve with more cases, while others will plateau or even diminish in skill if they stop intubating much as attendings.
-Another limitation they talk about is the lack of information on attempt duration. For almost every intubation paper, I feel like that is an important missing data point, which probably tells us even more about the intubation than the number of blade insertions.
There’s obviously a huge learning curve for trainees to become more proficient in intubation skills. This paper indicates that it may take around 119 cases for trainees to reach FAS rates at the national average. I’m not sure how actionable this information is (we can’t just create intubation opportunities when they aren’t there), but it’s really interesting at least, and can maybe even just inform us of how far we have to go to reach the levels that we want to reach.
Paper 4: Nancy Carney, Annette M. Totten, Tamara Cheney, Rebecca Jungbauer, Matthew R. Neth, Chandler Weeks, Cynthia Davis-O’Reilly, Rongwei Fu, Yun Yu, Roger Chou & Mohamud Daya (2022) Prehospital Airway Management: A Systematic Review, Prehospital Emergency Care, 26:5, 716-727, DOI: 10.1080/10903127.2021.1940400
Methods: Systematic review of 99 studies comparing BVM, supraglottic airway (SGA), and endotracheal intubation (ETI) in patients needing pre-hospital ventilatory support.
-There was a higher frequency of ROSC in adults w/ SGA vs ETI, however, ETI was associated with better neurologic outcomes on one type of scale
-No clear benefit to advanced airway insertion vs BVM.
-No survival benefits found in any comparison.
-There doesn’t seem to be any difference in adverse outcomes in any comparison
-There is such a wide range of types of patients requiring prehospital ventilatory support, so lumping them all together might not show much. But it’s still important to know that there’s no golden ticket here that all paramedics should be following.
-We don’t know if studies were using 1 or 2-person bagging techniques.
-They discuss the possible bias where some of the BVM patients likely weren’t seen as needing an advanced airway either because the providers thought ROSC was imminent, or the transport was so short (which would be associated with better outcomes).
-I think this all looks pretty equivocal. There aren’t a lot of RCTs, and generally the strength of evidence here is pretty weak. But nothing is jumping out as an obvious best practice, so I think SGA is probably best just because of ease of use. That being said, it’s really hard to set up a quality RCT for this sort of question, so this might be the best we can get.
Summary: While we often think of ETI as the gold standard, the prehospital setting is unpredictable, and has been shown to have variable success rates, along with higher complication rates with ETI. This paper attempted to investigate the available literature on the topic of optimal prehospital ventilatory support, and found that generally, there is no evidence of one technique performing better than any other.