Research Roundup (June 2024) – First10EM

Back again with another month’s worth of interesting, inane, or sometimes important emergency medicine literature. As always, podcast version on BroomeDocs or YouTube.

Obviously, GCS 8 doesn’t mean intubate

Freund Y, Viglino D, Cachanado M, Cassard C, Montassier E, Douay B, Guenezan J, Le Borgne P, Yordanov Y, Severin A, Roussel M, Daniel M, Marteau A, Peschanski N, Teissandier D, Macrez R, Morere J, Chouihed T, Roux D, Adnet F, Bloom B, Chauvin A, Simon T. Effect of Noninvasive Airway Management of Comatose Patients With Acute Poisoning: A Randomized Clinical Trial. JAMA. 2023 Dec 19;330(23):2267-2274. doi: 10.1001/jama.2023.24391. PMID: 38019968

I skipped right over this paper when it first popped up in my feed, because the premise seemed a little ridiculous, but after seeing the online discussion, it appears that there are a large number of people who practice this way: but obviously, GCS 8 does not mean intubate. This is an unblinded multicenter trial from 20 emergency departments (and 1 lonely ICU) in France, randomizing adult patients with a clinical suspicion of acute poisoning and a GCS less than 9 to have intubation withheld unless ‘emergency intubation criteria’ were met, or just standard of care. The very unsurprising headline news is that outcomes were better when you just leave these patients alone, but the details of this paper are sort of wild. First of all, almost 2/3rds of these patients were primarily intoxicated with ethanol. In my entire career, I have never even considered intubating a drunk patient. In fact, for the majority of my career, these patients weren’t even a health care problem. They were brought to a police drunk tank to sleep it off, and clearly don’t need intubating. More importantly, this is an unblinded trial with no deaths in either group, and so we are really just comparing ICU use. But ICU use is an asinine outcome to choose in an unblinded trial of intubation, because of course intubated patients spend more time in the ICU. That isn’t a question that needed to be asked, but it is the only question this paper answers. Honestly, the approach I was taught, in which these drunk patients are never intubated, could be wrong. It is possible that aspirations are occurring at a higher rate than I am realizing. That is the outcome that really needed to be addressed, and this paper really doesn’t do that at all. There are many other issues here, such as their focus on oxygenation without any mention of basic airway maneuvers, and their odd decision to stop the trial at 4 hours and allow doctors to do whatever they want at that point. You can read all the details in the main post. 

Bottom line: This unblinded trial shows that more liberal intubation results in more ICU usage. It does not show any difference in patient oriented outcomes, and has numerous sources of significant bias. That being said, I think the conclusion is clearly correct, because of course you shouldn’t be intubating drunk patients.

The “Syringe Hickey”

Issa EC, Ware PJ, Bitange P, Cooper GJ, Galea T, Bengiamin DI, Young TP. The “Syringe Hickey”: An Alternative Skin Marking Method for Lumbar Puncture. J Emerg Med. 2023 Mar;64(3):400-404. doi: 10.1016/j.jemermed.2023.01.013. PMID: 37019501

These authors discuss a technique for marking skin for lumbar puncture that does not involve ink and therefore will not wipe off with chlorhexidine or ultrasound gel. Honestly, when I downloaded this, I expected to be excited about it, and I even tried this a few times, but I have settled on this being a little ridiculous. To do this, rather than just pushing into the skin with a syringe (or any object) like most of us do, they used a 10 mL syringe, pushed into the skin, pulled back to the 5 mL mark, and left it there for 1 minute. They provide some nice pictures to demonstrate that it works, but don’t provide the comparison to simple skin indentation, so I am not sure that pulling back on the syringe adds anything at all. (The ‘hickeys’ they show really just look like indentations, not hickeys or petechiae.) Honestly, although this works, spending 1 minute to make a mark when simple skin indentation or a proper marker will work just as well seems like overkill. I have tried it a few times, and I get antsy well before the hickey is made, and proceed with no problems with just the skin indentation left in the first 10 seconds. I am also not sure the value of showing that this mark holds out to 30 minutes. If you are still trying a lumbar puncture at the same level using the same mark 30 minutes after you started, you have some major problems. I need a mark that lasts about 5 minutes, if I mark before starting, to allow for adequate time to anesthetize and clean. However, pre-marking for a lumbar puncture can be a significant pitfall, because you are assuming that the patient does not move at all. I will occasionally mark at 2-3 levels so I can easily see midline, and can tell if the patient has moved, but I alway re-palpate for anatomy before inserting my needle, making marking sort of moot. You will probably hear about this technique from multiple sources, and so it is worth knowing what people are talking about, but I think this is silly. 

Bottom line: If you have the patience, this is a fine trick, but if you can wait this long for a skin mark, you probably aren’t working in emergency medicine. 

The invisible gorilla strikes again

Drew T, Võ ML, Wolfe JM. The invisible gorilla strikes again: sustained inattentional blindness in expert observers. Psychol Sci. 2013 Sep;24(9):1848-53. doi: 10.1177/0956797613479386. Epub 2013 Jul 17. PMID: 23863753

I assume everyone has seen, or at least heard of, the invisible gorilla test. I like that they repeated it in radiologists. 24 radiologists were given 3 minutes to review 5 chest CTs, with the specific goal of looking for nodules. They were not told about the gorilla in the last CT, positioned near a lung nodule. Overall, the task was challenging, even for experts, and only 55% of the nodules were found. Twenty of the 24 radiologists failed to note the gorilla, despite most looking straight at it based on eye tracking software. They repeated the study with non radiologists who were trained to look for nodules. These participants were worse at finding nodules (12%), and none of them spotted the gorilla. So, despite the poor numbers, radiologists were better than a random Joe off the street. The gorilla test is a classic example of “attentional blindness”, where you are so focused on one task that you miss other pertinent details. (This has many implications in medicine, such as when you are so focused on intubating you fail to notice that the patient’s blood pressure has plummeted.) In radiology, we often refer to this as “search satisfying”, meaning that once you identify the thing you are looking for, you are likely to miss other abnormalities. I think emergency doctors are particularly prone to search satisfaction, because we know the clinical details, and have short enough attention spans that we really only want to look at the affected area, not the entire image. Although radiologists often disagree, this is an argument for leaving the radiologist without a lot of clinical detail, to ensure the full image is reviewed, but obviously the real lesson is that we need to force ourselves look at the entire image, rather than just focusing on the gnarly comminuted fracture that catches our eye.

We have a false positive problem in medicine

Kotani Y, Turi S, Ortalda A, Baiardo Redaelli M, Marchetti C, Landoni G, Bellomo R. Positive single-center randomized trials and subsequent multicenter randomized trials in critically ill patients: a systematic review. Crit Care. 2023 Nov 28;27(1):465. doi: 10.1186/s13054-023-04755-5. PMID: 38017475

I usually avoid the ultra-nerdy papers that focus specifically on evidence-based medicine topics in this clinical summary, but once in a while I think it is important to include some samples to help people understand why I am often very skeptical of new research. In this systematic review, they looked for any single center RCT published in New England Journal of Medicine, JAMA, or Lancet that demonstrated a mortality benefit in critically ill patients, and asked the simple question: what did follow-up multicenter trials show? Of the 19 single center RCTs that demonstrated a mortality benefit, only 16 had a follow-up multicenter RCT, and only one of those follow-up RCTs confirmed the original findings, with another follow-up multicenter RCT actually showing the exact opposite, with an increase in mortality. We have a false positive problem in medicine. We use ridiculously lax p-values, and many sources of conflict of interest, and don’t adequately adjust for the many sources of bias in research. This paper really demonstrates why replication is the core of science. You should always consider initially positive RCTs to be hypothesis generating, even if they demonstrate changes in important outcomes like mortality, and always take guidelines based on these trials with a massive grain of salt. (This also really should make you question research, such as the trials of PPIs for GI bleed, that have never been able to demonstrate important changes in patient centered clinical outcomes.)

Bottom line: When encountering new research, have a very low pretest probability that it will help patients. That isn’t cynical; its science. 

Charters B, Foster K, Lawton B, Lee L, Byrnes J, Mihala G, Cassidy C, Schults J, Kleidon TM, McCaffery R, Van K, Funk V, Ullman A. Novel Peripheral Intravenous Catheter Securement for Children and Catheter Failure Reduction: A Randomized Clinical Trial. JAMA Pediatr. 2024 Apr 1:e240167. doi: 10.1001/jamapediatrics.2024.0167. PMID: 38558161

This is a practical and potentially practice changing article that would be easy to overlook because physicians often aren’t involved in managing IVs (at least in North America). In a multicenter RCTinvolving 383 children who were being admitted from the ED and who were thought to require an IV for at least 24 hours, they compared 3 strategies to secure the IV. Standard care was a Tegaderm. The second group had a ‘integrated securement dressing’ (ISD), brand name “Sorbaview Shield” (pictured below). The third group combined the ISD with a tissue adhesive (TA). 

For their primary outcome of PIVC failure, the combination of ISD and TA was the best at 12%, while standard care was the worst at 34% (an absolute improvement of 21%, 11% of which was from preventing dislodgement). Secondary outcomes were not statistically different, but seems to either favour the ISD/TA combo or simply had too few numbers to say anything. Satisfaction, both of caregivers and staff, was identical across the board (which is weird given the massive absolute difference in IV failure and therefore presumably need for re-poking children). This is an unblinded trial, and although the outcomes seem somewhat objective, bias is highly likely. Nurses spend variable amounts of time trying to save IVs that appear to have stopped working, and fancy new dressing might convince them to spend more time than standard care. Personally, I wouldn’t have expected this intervention to change the rate of IV occlusion, so given that there was also a big change in occlusion rates alongside the change in unintentional dislodgement, I wonder if that might be telling us the data is biased. Perhaps the biggest problem with this paper is it doesn’t compare to what I think is standard care in Canada, which I might describe as the “just wrap the hell out of it, with many layers of cling, and a splint of some sort” technique. (I don’t know how successful that technique is.) This is yet another trial that was stopped early, not because of any specific endpoint, but because COVID derailed research. Although you feel bad for researchers when that happens, these early stoppages add a lot of uncertainty.

Bottom line: Despite some significant limitations, this is the best data I know of, and we know that standard of care is imperfect, so adding tissue adhesive and better designed dressings routinely to pediatric IVs might make sense as we wait for better data. 

Real emerg docs should be able to intubate in outer space

Thierry S, Jaulin F, Starck C, Ariès P, Schmitz J, Kerkhoff S, Bernard CI, Komorowski M, Warnecke T, Hinkelbein J. Evaluation of free-floating tracheal intubation in weightlessness via ice-pick position with a direct laryngoscopy and classic approach with indirect videolaryngoscopy. NPJ Microgravity. 2023 Sep 8;9(1):73. doi: 10.1038/s41526-023-00314-y. PMID: 37684267

I can’t remember, but intubating in outer space seems like the kind of topic I might have covered before. Apparently that would be easy to do, because there is a lot of prior research that shows people are bad at direct laryngoscopy in outer space. However, like everyone in emergency medicine, astronauts are also fans of video laryngoscopy. (Sorry, Casey. I know we said no more video versus direct laryngoscopy papers, but this is airway in outer space, so it doesn’t count.) If you weren’t convinced by all the prior studies, this might be the one that puts you over the edge: in outer space, VL beats DL. Specifically, direct laryngoscopy with an ice pick approach had a success rate of 33% and video using a standard approach had a success rate of 58%. Or, put plainly, both sucked and it is very clear that people should not be intubating in outer space. Why we would want to at this stage is well beyond me. Astronauts may be ‘cowboys,’ just like emergency doctors, but it seems like they also need to learn the lesson that plastic in the trachea is almost never the most important intervention in critically ill patients. A modern LMA might be reasonable in a space emergency pack, but having novice astronauts intubate each other seems like poor judgment. (That being said, I have no idea how well an iGel would seal in microgravity, nor what the aspiration risks are when fluid can travel in any direction.)

Bottom line: I really only included this paper so I could copy and paste some pictures of astronauts trying to intubate upside down. Sorry paramedics, it seems like intubating the 550 pound patient behind their toilet wasn’t difficult enough, and the spacemen had to show you up. (Gendered language seems appropriate when you consider the intent of this paper and look at the pictures.)

Heresy, witchcraft, Jean Gerson, scepticism and the use of placebo controls

Kirakosian R, Möllenbrink L, Zamore G, Kaptchuk TJ, Jensen K. Heresy, witchcraft, Jean Gerson, scepticism and the use of placebo controls. J R Soc Med. 2024 Jan;117(1):36-41. doi: 10.1177/01410768231207260. Epub 2023 Nov 22. PMID: 37991475

An interesting essay looking into the deep history of placebo. In the early 1600s in France, in an attempt to settle religious disputes based around demonic possessions and exorcisms, the idea arose of exposing an individual to sham religious items that looked identical to items that were actually consecrated to see if there was a difference in response. Going further back, Jean Gerson was concerned with using “trick trials”, or what we would now call placebo controlled trials, to assess the veracity of divine revelations. Unfortunately, his methodology was adopted by the church before science, giving rise to a very brutal history of witch hunts, based on pseudo-scientific ideas. It took another 200 years for these ideas to be repurposed away from witches and towards charlatans, when Antoine Lavoiser definitively proved that Anton Mesmer and his claims of animal magnetism were fraudulent, based on one of the earliest examples of a placebo controlled trial. A frivolous paper, but one I think many would enjoy reading for themselves. 

Cheesy Jokes of the Month

I am a big fan of the metric system, but I have to admit that if America changed from pounds to kilograms overnight, there would be mass confusion.