Articles, articles, articles. So much wonderful science to improve patient care. So little time. Do you just skip to the bottom line?
These articles can be digested in podcast version, if you prefer, either through the BroomeDocs site, or on YouTube.
High dose nitro for SCAPE
Siddiqua N, Mathew R, Sahu AK, Jamshed N, Bhaskararayuni J, Aggarwal P, Kumar A, Khan MA. High-dose versus low-dose intravenous nitroglycerine for sympathetic crashing acute pulmonary edema: a randomised controlled trial. Emerg Med J. 2024 Jan 22;41(2):96-102. doi: 10.1136/emermed-2023-213285. PMID: 38050078
Despite being a small, single center, unblinded trial, I think this RCT should be practice changing for many people. What dose of nitroglycerin do you use for patients with flash pulmonary edema, or sympathetic crashing acute pulmonary edema? This trial randomized patients to high or low dose nitroglycerin. High dose was a 600-1000 mcg IV bolus followed by an infusion started at 100 mcg/min. The low dose got no bolus and had their infusion started at 20-40 mcg/min. All patients were started on BiPAP. The outcomes were dramatically better with high dose nitroglycerin. At 6 hours, 65% of the high dose group had resolution of their symptoms as compared to only 12% of the low dose group (p<0.001). The difference was still large at 12 hours, with 89% resolution as compared to 19% (p<0.001). (These numbers are all surprisingly low to me, as I get close to 100% resolution of SCAPE within the first hour or two with high dose GTN and BiPAP.) Basically all outcomes were worse in the low dose group. Intubation was 19% vs 4% (although not statistically significant). (Again, I have not intubated one of these patients in my career, so this group seems way sicker, or different in some other way, than what I am seeing in Canada.) Admission rates, hospital length of stay, and MACE were all higher with low dose nitroglycerin. In terms of adverse events, no patients developed hypotension in either group. 11 patients in the low dose group and 3 in the high dose group developed a headache. Obviously, there are tremendous limitations in a small, single center, unblinded RCT, and I am usually very cautious in recommending practice change based on such data. However, there is a lot of existing data on the safety of this practice, and with such a dramatic benefit, I think this is probably practice changing (at least until better RCTs are run).
Bottom line: High dose nitroglycerin seems to result in better outcomes in SCAPE, at least according to this small, single center, unblinded trial.
Is our PPE evidence-based?
Diekema DJ, Nori P, Stevens MP, Smith MW, Coffey KC, Morgan DJ. Are contact precautions “essential” for the prevention of healthcare-associated methicillin-resistant Staphylococcus aureus? Clin Infect Dis. 2023 Sep 21:ciad571. doi: 10.1093/cid/ciad571. PMID: 37738565
Science is about nuance. Conversely, IPAC practitioners seem only to care about draconian rules. Thus, it is never surprising to me to find out that IPAC’s advice (or, as they prefer, rules) are ascientific, or just wrong. (See their ongoing misunderstanding about the mechanism of spread for respiratory viruses, or the general misrepresentation of the evidence around N95s. In fact, even the handwashing dogma has been very well challenged by Scott Weingart, if you have a EM:RAP subscription.) Overall, this controversy may not be that important to emergency physicians, because we usually don’t know whether our patients are colonized with MRSA, but these authors challenge the idea that contact precautions are essential for MRSA colonized or infected patients. There are no direct RCTs of contact precautions for MRSA. However, there are a number of cluster RCTs from the ICU setting, and they show no benefit. (There appears to be a benefit from routine decolonization with chlorhexidine and mupirocin, but not from contact precautions.) Observational studies are used to support contact precautions, but the results are actually mixed, and they need to be interpreted in their broader context. For example, although MRSA rates decrease in some studies, so do rates of all infections across the hospital, indicating a broader trend unrelated to contact precautions (perhaps a simultaneous handwashing campaign). Perhaps the more interesting conversation is about the harms of universal contact precautions. (They are often painted without any nuance as pure benefit only). These authors discuss evidence that contact precautions lead to decreased physical exams, less contact with healthcare professionals, longer hospital lengths of stay, higher hospital costs, and higher rates of readmission. The added time needed for donning and doffing PPE, when it is for many patients, limits time for clinicians to actually do clinical work. There is also a sort of ‘alarm fatigue’, such that if PPE is required for every patient, we are less likely to be careful about appropriate PPE when it really matters. Finally, there are broader societal issues to consider. The healthcare system is estimated to produce almost 10% of greenhouse gasses in North America, and the gowns we use for contact precautions are single-use petroleum-derivative plastics. (Think about how many plastic straws I could use for my milkshakes – a beverage ill-designed for paper straws – if I just stopped using stupid yellow gowns.) These authors aren’t advocating abandoning contact precautions altogether, but rather for a more evidence-guided approach that uses them only when there is an outbreak, or rising MRSA rates. Personally, I find the yellow gowns to be silly, especially as they often hang away from your body and potentially increase the risk of contamination of the environment. However, I don’t touch patients with bare hands, and so I guess I use partial contact precautions on everyone?
Bottom line: Evidence based medicine is important, even in infection control.
Great evidence for button battery ingestions (if you are treating a dead pig)
Chiew AL, Lin CS, Nguyen DT, Sinclair FAW, Chan BS, Solinas A. Home Therapies to Neutralize Button Battery Injury in a Porcine Esophageal Model. Ann Emerg Med. 2023 Sep 19:S0196-0644(23)00671-6. doi: 10.1016/j.annemergmed.2023.08.018. PMID: 37725021
Everyone knows that button battery ingestions are dangerous, and that seems to lead to a lot of discussion about potential medical therapies we can use in the emergency department. I have always sort of assumed these therapies were like medical management for esophageal food boluses: general wastes of time suggested by GI as delaying tactics because no one seems to want to scope (after hours at least). This study looked at the effect of many home products on a cadaveric porcine esophagus. They used the standard recommended honey, but also jam, orange juice, Coca-Cola, milk, and yogurt. Both honey and jam seemed effective at lowering the pH in the esophagus (lower being better, as button battery injuries are alkali injuries), as well as reducing the total surface area of the burn. None of the other substances seemed effective. In a second experiment, they just compared honey and jam to saline, and claim honey looks better based on histopathology, but I find these results very unconvincing, as the numbers seem very similar. Furthermore, for some reason they didn’t blind the histopathologist to the exposure. Blinding would have been simple and cost nothing, and would have dramatically decreased the bias in this study. Of course, the main question is whether a cadaveric model is appropriate. Living tissue is not like dead tissue in many important ways. (I was going to just leave it at that, but I do think it is important to consider the impact of peristalsis. In a dead esophagus, is the honey more likely to pool around the battery, whereas a living esophagus will propel it into the stomach? The cadaveric model may explain higher viscosity liquids seeming to help, but that might not translate to real life. Really, I think the most important point of data in this trial is that there was almost no mucosal damage within 90 minutes with either honey or jam, so the key is still urgent endoscopy.
Bottom line: There are a lot of issues with trying to extrapolate from cadaveric animal studies, but I see limited risk (unless you use this therapy as an excuse to delay endoscopy). Until we see more science, it seems very reasonable to provide honey, jam, or sucralfate (which wasn’t studied here), every 10 minutes while arranging for that emergent scope.
The current European guideline is 10 mL every 10 minutes, with a maximum of 6 doses of honey or 3 doses of sucralfate. Don’t do this if there are signs of perforation, mediastinitis, or sepsis, and avoid honey in patients less than 1 year (for botulism risk).
Myths and misconceptions about the treatment of acute hyperkalemia
Gupta AA, Self M, Mueller M, Wardi G, Tainter C. Dispelling myths and misconceptions about the treatment of acute hyperkalemia. Am J Emerg Med. 2022 Feb;52:85-91. doi: 10.1016/j.ajem.2021.11.030. PMID: 34890894
I like myths. I like managing hyperkalemia. Thus, this paper caught my eye. I don’t think we need to go into any depth, but if any of these comments surprise you, it might be a good idea to read the full paper.
Myth 1: Kayexalate is safe and useful. Really, the drug was introduced after only a case series, and there have only ever been 4 RCTs conducted, 3 of which were negative. It may have no effect at all, based on the data we have, but at most it seems to change potassium levels by about 1 mEq/L 3-7 days later, which isn’t very useful in the acute setting. However, there have been a number of deaths attributed to colonic necrosis from kayexalate, so you are probably better to just avoid it altogether.
Myth 2: Lactated Ringer’s is contraindicated in hyperkalemia. It is easy to see where this myth started, as LR has potassium in it, but even a brief pause to think should have dispelled this myth. The concentration of potassium in LR is only 4-5 mEq/L, meaning even in a closed system, LR would dilute down the potassium levels. However, we know that the vast majority of potassium is intracellular, so this tiny amount of potassium is meaningless. That being said, no one has ever shown LR to be any better than normal saline, so you can probably just use normal saline, like I still do.
Myth 3: The ECG changes from hyperkalemia are predictable and reliable. They aren’t.
Myth 4: All patients with hyperkalemia should be treated with calcium. There are some (very small) risks of intravenous calcium, and it almost certainly isn’t needed unless there are ECG changes.
Nebulized ketamine: the solution to a non-existent problem?
Kampan S, Thong-On K, Sri-On J. A non-inferiority randomized controlled trial comparing nebulized ketamine to intravenous morphine for older adults in the emergency department with acute musculoskeletal pain. Age Ageing. 2024 Jan 2;53(1):afad255. doi: 10.1093/ageing/afad255. PMID: 38251742
People are obsessed with ketamine for analgesia, and alternative methods of providing medications, so this trial of nebulized ketamine caught a few peoples’ eyes. This is a well designed trial of elderly patients presented with MSK based pain of at least 5 on a 10 point scale. They compared 0.1 mg/kg of IV morphine to 0.75 mg/kg of nebulized ketamine, and looked at the reduction of pain at 30 minutes. They started with a reasonable dose of morphine, but it is still a dose that we know will only be adequate in 50% of patients. Despite that, for some reason they switch to fentanyl rather than just sticking with a single opioid, which seems odd to me. It was a properly blinded trial, so you got either nebulized or intravenous saline as a placebo depending on your group. They enrolled 261 patients, and pain scores dropped by 2/10 in both groups at 30 minutes (from about 7 to about 5). This allowed them to conclude that ketamine was ‘noninferior’ to morphine, based on their threshold of 1.3. (In other words, they designed the trial so that ketamine could be as much as 1.3 worse on the 10 point scale and still be considered non-inferior, which seems like a mistake, when the majority of our analgesia trials show benefits of less than 1.3) As I have recently discussed, noninferiority trials are a bit of a mess, and I am not sure that we care to prove that nebulized ketamine is non-inferior to IV morphine. Instead, if we are going to adopt this therapy, I think that we want to see that it is better in some way. More importantly, we need to see long term outcomes for ketamine, because we are usually treating painful conditions expected to last more than a few hours. What are you doing with the ketamine group after you send them home? Was there really any benefit from one dose of ketamine, or are you still sending the patient with a prescription for morphine anyway? (And wouldn’t it be better to give the patient the same medication in the ED as you are going to prescribe at home, so you are aware of potential side effects?)
Bottom line: Although it is always useful to have more tools in your belt, I can’t see this study having much clinical relevance at all.
Small BVMs: When physiologic theory fails
Snyder BD, Van Dyke MR, Walker RG, Latimer AJ, Grabman BC, Maynard C, Rea TD, Johnson NJ, Sayre MR, Counts CR. Association of small adult ventilation bags with return of spontaneous circulation in out of hospital cardiac arrest. Resuscitation. 2023 Dec;193:109991. doi: 10.1016/j.resuscitation.2023.109991. PMID: 37805062
Association is not causation. However, we have identified over-bagging as a significant issue since I was a resident. Despite a decade of teaching on the topic, I still routinely see large volume breaths being given at 40 breaths a minute when I walk into a code. Changing the system or the equipment so that clinicians don’t have to rely on memory in high stress situations makes a lot of sense to me. The Seattle Fire Department (EMS) agreed, and in 2017 changed the BVM they stocked from the standard 1500 mL bag to a small BVM, measuring 1000 mL, and delivering a breath between 450 and 700 mL. They expected the change would increase return of spontaneous circulation (ROSC), but they were wrong. This is a retrospective look at prospectively collected data (historically, this is an excellent database). They looked at 1994 cardiac arrest patients in a before and after design, and it turns out that the change to a smaller BVM was associated with lower rates of ROSC (33% vs 40%, OR 0.74, 95%CI 0.61-0.9). They only included patients with an advanced airway (I think so they could display pretty data), but that seems to be a mistake, as the change to a smaller bag would also affect patients being ventilated with only a BVM, so we really need to know their outcomes as well. Because of the advanced airway and use of quantitative end tidal CO2, they have excellent data, and both groups had a consistent respiratory rate of about 12, which is somewhat higher than I teach in cardiac arrest, but definitely reasonable. Of course, this is just before and after observational data, so we don’t know what else might have changed over time, but it doesn’t seem to support a strong argument that tidal volumes have a big impact on survival in cardiac arrest.
Bottom line: In this before and after cohort, the change to a smaller BVM was associated with lower rates of ROSC.
Peripheral pressors: not only safe, but beneficial?
Yerke JR, Mireles-Cabodevila E, Chen AY, Bass SN, Reddy AJ, Bauer SR, Kokoczka L, Dugar S, Moghekar A. Peripheral Administration of Norepinephrine: A Prospective Observational Study. Chest. 2023 Aug 21:S0012-3692(23)05350-3. doi: 10.1016/j.chest.2023.08.019. Epub ahead of print. PMID: 37611862
At this point, I sort of assume that everyone knows it is OK to run vasopressor peripherally. First10EM readers should, given that we have covered the topic previously, with another post with more evidence here. Therefore, we can probably leave this summary very short. It is prospective observational data from 603 consecutive patients from a single medical ICU who received vasopressors through a peripheral IV. In terms of safety, extravasation occurred in 35 patients (3.9%), and the longer the IV was in place, the higher the risk (obviously). However, essentially all of these events caused no or minimal tissue damage, with no patients requiring surgery. (There was one patient who was graded as having a high grade extravasation by the bedside nurse, but who was never evaluated by the study team as the patient was made comfort care before that could occur.) Thus, there is a small risk of extravasation, with an incredibly low risk of real tissue damage. However, what this study adds as compared to prior studies is a clear measure of the benefit. About half of these patients never required a central line at all, thus preventing exposure to an invasive procedure that seems to have a higher complication rate than peripheral pressors. Overall, the use of peripheral pressors prevents an average of 1 day of central line use per patient.
Bottom line: Although there is a small risk of extravasation, the benefits of running vasopressors will outweigh that risk in many patients. Although I miss performing the procedure, these days I almost always leave the central line until the patient is stabilized and settled into their ICU bed.
Is pregnancy a disease?
Smajdor A, Räsänen J. Is pregnancy a disease? A normative approach. J Med Ethics. 2024 Jan 29:jme-2023-109651. doi: 10.1136/jme-2023-109651. PMID: 38286592
This is a paper for quiet reflection by a roaring fireplace, not rapid summary by an emergency doctor with undiagnosed ADHD. Some will be tempted to dismiss it. Some will probably be upset by it. But the question “is pregnancy a disease?” has important implications for many of our patients (not just those who are or who may become pregnant). To start, the authors compare pregnancy with measles. Both are self-limiting conditions with a predictable trajectory that almost always results in full recovery. Both involve a variety of symptoms and physiologic changes that can impair one’s normal functional ability. In terms of risk, the chance of dying from pregnancy is actually higher than the chance of dying from measles. So, the authors ask, what is the difference? They explore the many modifiers often used to define disease, such as whether it is subjectively good or bad, wanted or unwanted, but essentially use pregnancy to illustrate the significant limitations of such subjective criteria. It feels like doctors should be able to clearly and easily define “disease”, but if you spend some time with this paper, I think you might find the term slips from your fingers. Disability advocates have been screaming this for ages, and would be happy if more of us internalized the many problems with trying to define “disease”. It is unlikely to change your day to day practice in the emergency department, but if you want to spend a couple hours thinking about how your profession fits into the structure of the world, this paper might be a good jumping off point.
A randomized deception trial?
Iyer R, Park D, Kim J, Newman C, Young A, Sumarsono A. Effect of chair placement on physicians’ behavior and patients’ satisfaction: randomized deception trial. BMJ. 2023 Dec 15;383:e076309. doi: 10.1136/bmj-2023-076309. PMID: 38101923
The benefits of sitting when talking to patients have been discussed since I was in medical school, but there was just no way I was going to skip discussing a “randomized deception trial”. This is a “single center, double blind, randomized controlled deception trial” that compares placing a chair within 3 feet of the bed facing the patient to the usual position of the chair (which for some inexplicable reason in this hospital was in a cupboard). Neither the physicians being observed nor the patients were aware of the objective of the study. They observed 51 hospitalist physicians over a total of 125 patient encounters. When a chair was placed in the room, physicians were much more likely to sit (odds ratio 20.7, 95% CI 7.2 to 59.4, p<0.001), although I was surprised that only 2/3rds of the physicians sat when a chair was readily available. There was a statistically significant difference in patient satisfaction scores, but the absolute difference seems very small (88% vs 84% on a TAISCH score), and I am not sure it would be considered clinically meaningful. Having a chair didn’t impact the length of the patient encounter, and also (in contrast with prior research) did not impact patients’ perceptions of time. (That might be because these hospitalists were spending 10.6 minutes in the room, as compared to the average 45 seconds an emergency physician might spend.) Despite the low overall difference in scores, when I look through the components of the score, there are some pretty big differences in what seem like important questions, such as “how much confidence do you have in your physician’s plan for your care?’
They talk about this as a “nudge”, as if physicians need to be pushed towards sitting, but that seems like a crazy way to frame this issue. Of course physicians want to sit while talking to patients. Why wouldn’t we want to be comfortable while talking? This is a design issue, with most hospitals being so horrendously designed that there just aren’t available sitting options. (I wonder whether I have been unknowingly participating in this trial for years? I always try to sit when talking to patients, but consistently struggle to find a chair in most of the patient rooms, and this is especially hard when you are forced to see patients in hallways or waiting rooms.) I think this study emphasizes my design complaints beautifully, as apparently the “usual” location for the chair in these hospital rooms was inside a closed cupboard.
Bottom line: This study reinforces prior research that suggests sitting while speaking with patients might improve patient satisfcation. They want to spin this as a study of ‘nudging’ physicians towards sitting, but I think it is a study that emphasizes the need for much better design in healthcare spaces.
Cheesy Joke of the Month
Why should you never brush your teeth with your left hand?
Because a toothbrush works better
