I posted this paper and it’s revisited update on room9er about 3 years ago. July is always a good time for it!

Here is the original paper.

Here is the update.

The fundamentals do not change much!

Maryland pearls post about a recent paper in JEM about patients coming in with uncertain head trauma.

1. Subscribe to the Maryland pearls if you don’t already
2. Do not automatically go scan every geriatric patient who might have hit their head. But consider it on a patient to patient basis.

While the unknown patients had a lower % of positive head CT, it was not negligible. See the description below:

In this prospective study looking at geriatric patients with unknown head injury vs. known head injury, the unknown head injury group had an ICH 1.5%, neurosurgical intervention 0.3% and delayed ICH 0.1% when compared to known head injury (10.5%,  1.2% and 0.7% respectively).  The authors concluded that the risk of ICH was high enough in uncertain head injury patients to warrant scanning.

Turchiaro ML Jr, Solano JJ, Clayton LM, Hughes PG, Shih RD, Alter SM. Computed Tomography Imaging of Geriatric Patients with Uncertain Head Trauma. J Emerg Med. 2023 Dec;65(6):e511-e516. doi: 10.1016/j.jemermed.2023.07.009. Epub 2023 Jul 26. PMID: 37838489.

Brief but informative post from the Canadian Medical Protective Association (CMPA). They apparently do have lawsuit risk in Canada with as many as 24% of EM physicians named in a case in 5 years.

Check out our UL DEM 2018 article* that appeared in ABEM’s LLSA list. Many of the same diagnoses remain high risk today: Fractures, Lacs/wounds, Stroke, ACS, Appendicitis (And other GI), and less commonly seen in other reviews, respiratory system infections.

*Brian Ferguson, Justin Geralds, Jessica Petrey, Martin Huecker. Malpractice in Emergency Medicine-A Review of Risk and Mitigation Practices for the Emergency Medicine Provider. J Emerg Med. 2018 Nov;55(5):659-665.

Risk reduction reminders from the CMPA article:

The following risk management considerations have been identified for physicians providing care in the emergency department:

• Perform an objective and thorough assessment of patients and when appropriate, incorporate clinical practice guidelines and clinical decision rules for investigating common conditions encountered in the ED.
• Take time to pause and reflect on the differential diagnosis, being careful to consider any relevant risk factors, including comorbidities and surgical or family history. Obtain a second opinion if you are unsure of your diagnosis.
• Provide patients with appropriate follow-up and clear instructions (verbal or written), including symptoms and signs that should alert them to seek further medical attention and how urgently to do so. Confirm patients’ understanding of the information being provided, and answer questions honestly and openly.
• Communicate clear instructions during formal written handovers of care that include relevant patient history, pertinent findings on physical examination, differential diagnosis, diagnostic investigations performed, outstanding results, and the next steps in the patient treatment plan.
• For patients with continued or worsening symptoms or those who repeatedly return with unresolved complaints, re-evaluate the diagnostic assumption, repeat the physical examination, and consider alternative diagnoses, ruling out possibilities that may be life-threatening.
• Document differential diagnoses, pertinent positives and negative findings, reassessments, and discharge discussions

I always say that it takes more skill to NOT intubate a patient. That is especially the case with severe CHF, where BiPAP, nitro and a couple of hours can turn them around completely. Intubation is a dangerous procedure, and I think we have come a long way in EM from the days when we had a low threshold to intubate intoxicated patients.

The TL;DR is that in patients suspected of “poisoning” (which to them means alcohol, drugs or medication), an expectant approach of “restricted intubation” led to improved outcomes of shorter ICU stay, shorter hospital stay, less pneumonia, and of course less adverse events from intubation.

Read this article and the commentary (both linked below). I thought this was too important to wait for Journal Club, but we could still cover it at JC in the near future. 20 different EDs, RCT, 225 total patients, excluded some poisonings that had easy reversal or needed antidotes. No patients died.

Here is the original research article. Here is the accompanying editorial.

Check out this recent lecture from Dr Stephen Smith of the famous ECG Blog. The link is to a google doc but it is still live after months. He describes some sophisticated ST segment and T wave changes that ER doctors must know to pick up subtle ischemia. You have to just love how Smith is so candid about his opinions. He does not mess around with anyone who argues with the Occlusion MI (OMI) paradigm or even individual tracings. He says that sadly some people just can’t see the subtle findings, but I maintain hope that every studious ER doc can master the image pattern recognition he teaches. And if they can’t, maybe AI can. Smith advises multiple companies that use AI to detect these subtle ECG findings, to determine when patients are having OMIs. His software appears to be quite effective.

Something to keep in mind while watching, he is going over (tons of) cases that all have a relatively high pretest probability of ischemia. He has selected them out. We are working on evaluating and treating more ‘cardiac patients’ at ULH, but his patient population (at Hennepin) would be more like Jewish or maybe even other centers in town with more active cath labs.

A few of the rules that came up repeatedly in the video:

Smith of course talks about the weakness of a STEMI/NSTEMI paradigm, arguing instead for the occlusion MI paradigm.

Proportionality, proportionality, proportionality – T wave size in proportion to the QRS. A medium sized broad T wave after a tiny QRS is concerning!

Similarly, the morphology of the Hyperacute T is usualy broad based, and not tall and peaked like hyperK+. Thus, thinking about the area under the curve of the T wave makes more sense.

Biphasic T waves (down up meaning recip change) – we usually talk about biphasic reperfusion T waves in the leads involved in ischemia, but here mostly shows biphasic T waves reciprocal to infarction pattern in other leads.

If you have the urge to order Mag for a patient, follow your intuition. But Mag might be even more important when replacing potassium.

When treating hypoK+, if the patient has an IV or will have one, I just order Mag 2g IV rapid infusion (never need to do it slowly, 20 minutes is great) along with 60meq oral K. If they are below 2.0 K+, I either give 2 K runs or if they need/can tolerate volume, a liter of D51/2NS with 20meqK (K runs which are painful and seem to take forever to get to the bedside). Of note, people with CHF very often Mag depleted from diuretics and other etiology.

If you give the Mag IV and K po a the same time, Mag is hitting them first. Tough ones are when you don’t have an IV, because Mag oxide is trash. Sometimes I still order it. But people with no IV are likely not very sick and probably have K above 3.

Something might happen when K gets below 3. So if below 3, I usually give some IV, some po.

Dr Harmon asked me about this before she graduated, and I went looking for a few papers on it. I did a little lit search and cant find a true RCT of K repletion VS K repletion WITH or AFTER Mag repletion. It would be expensive to do and lots of confounders and no one wants to spend $ on a Mag study because it’s an old, cheap medication. **(Although some press coverage now on a Magnesium L-Threonate [which I take] study that showed cognitive benefit in Alzheimers patients).

In the few relevant papers I found, authors just generally recommend Mag with K. But we should always be careful when doing something that is logical but isn’t proven empirically. Sometimes things that make sense in theory don’t pan out in studies (vitamin C in sepsis).

One study in ICU patients compared those getting lots of Mag vs those getting none, and looked at their K balance. They found an obvious benefit to Mag repletion for K balance. They cite lots of basic science research on the K-Mg interplay.

At the end of the day, most people are Mg deficient and people with low K maybe even more likely Mg deficient. Mag is awesome, no downside, patient might flush or get sleepy. Just watch out in those with renal failure.

As mentioned in conference recently, we have years of various studies on the use of Magnesium Sulfate in COPD and asthma. See below a Cochrane review on asthma.

But right after conference, I checked my email to find hot off the press in Annals of EM, a brief review on Mag in COPD.

Take-Home Message

Among patients with an acute exacerbation of chronic obstructive pulmonary disease, intravenous magnesium sulfate may be associated with fewer hospital admissions, reduced hospital length of stay, and improved dyspnea scores.

Here is the Cochrane review on Mag in Asthma. Authors’ conclusions: This review provides evidence that a single infusion of 1.2 g or 2 g IV MgSO4 over 15 to 30 minutes reduces hospital admissions and improves lung function in adults with acute asthma who have not responded sufficiently to oxygen, nebulised short-acting beta2-agonists and IV corticosteroids. Differences in the ways the trials were conducted made it difficult for the review authors to assess whether severity of the exacerbation or additional co-medications altered the treatment effect of IV MgSO4. Limited evidence was found for other measures of benefit and safety.Studies conducted in these populations should clearly define baseline severity parameters and systematically record adverse events. Studies recruiting participants with exacerbations of varying severity should consider subgrouping results on the basis of accepted severity classifications.

Strangulation injuries are a tough chief complaint. We have many considerations in evaluating and managing these patients. Top priority is ABCs, and then ruling out other serious injuries in the patient.

We may tend to have too low a threshold for CTA in these patients. But they often end up in court proceedings and one could argue for the more aggressive imaging strategy for this reason. Of note, strangulation in the setting of domestic violence represents a VERY high risk mechanism to predict subsequent fatal injury in intimate partner violence.

I am not offering a clear cut answer on when to CTA and when not to. This should be a decision you make with the patient and your attending, considering patient age, injury severity, etc. But the two resources below can at least provide some context on evidence for imaging.

1. Check out this algorithm, authored by Dr. Bill Smock who was UL EM faculty for years and worked with the LMPD as the police surgeon for years. He writes the forensic medicine chapters in a few textbooks as well.

2. Also check out the paper below, authored by UL Emergency Radiology physicians including Dr. Jonathan Joshi.

https://pubmed.ncbi.nlm.nih.gov/31055673/

Check out this new essay in Annals EM. Well written. Our job is pretty amazing.

Check out this great article about nurse intuition on acuity level of patients. Link posted by Sam Ghali, MD who you should follow on Twitter. TL;DR Listen to the nurses!

The study asked nurses in 2 medical and 2 surgical units in Rochester, MN to score patients based on a 5 point “Worry Factor.” Basically deciding sick or not sick. 31,000 shifts in 3551 hospital admissions. The Worry Factor was highly accurate, with a LR of ICU transfer of 17.8 for WF>2 and LR 40.4 for WF>3. Accuracy was higher for RNs with more experience. AUROC was 0.92 for ICU transfer in 24 hours. The article specifies that they couldn’t assert whether RNs used intuition or analytical skills (something our Gut Instinct study DID try to determine).

This paper reminds me of an article I wrote a few years ago about a teaching tool for the ED, asking EM residents to decide admit vs discharge (or try to guess diagnosis, etc,) the moment they see a patient.

The references for this article are fantastic as well. Multiple primary sources and reviews on the various scores MEWS, NEWS, EWS, etc that try to identify who will decompensate in the hospital. I like to think ER nurses and doctors are especially skilled here, although we should be better about following up on patients we admit. You called the ICU and they deflected to PCU: check the chart the next couple of days, were you right or wrong? That feedback is necessary to modify your mental models and learn. At least 5 of the references cover Nurse Worry, including one systematic literature review and one prospective trial in Denmark. The references also go into intuition, expertise, they even cite the book Thinking Fast and Slow, our inspiration for the Gut Instinct abstract that was presented by Carter and Giddings.

I have had this meme in my head for a while but don’t think I ever made it or saw it on the Internet. Maybe it will go nerd viral.

As I have lectured in didactics, a paradigm shift is taking place toward the OMI vs nonOMI, and perhaps moving away from STEMI vs nSTEMI.

The ACC may be getting on board with this change that began with ER docs, chiefly Stephen Smith at Hennepin. *Unstable Angina still exists.

Check out his tweet linking to the paper:

1/2 For first time, the Am Coll of Cardiology recognizes Occlusion MI in clinical guidelines (and references our first of many OMI/NOMI studies: Meyers HP, … Smith SW. Comparison of STEMI vs. NSTEMI & OMI vs. NOMI Paradigms of AMI. J Emerg Med 2020) https://jacc.org/doi/epdf/10.1016/j.jacc.2022.08.750

2/2 And also, for the first time in any Guideline (as far as I know), they recommend EKG criteria that were developed by an Emergency Physician (Smith Modified Sgarbossa Criteria). Page 7 of the pdf, references 10, 11, 21. https://jacc.org/doi/epdf/10.1016/j.jacc.2022.08.750

Interesting paper here, retrospective data pulled from the NEAR registry, a big high quality airway registry.

The take home point was NO difference in desaturations during induction in the ramp vs supine position. They reported DL and VL cohorts separately. Now I like to ramp patients, especially those with low GCS or obesity. So I am going to find the issues with the paper, and try to hold strong in my beliefs.

But seriously, there are some problems. It is retrospective. The patients who were ramped were probably sicker and more obese! It turns out they are:

However, obesity and subjective impression of difficult airway were more common in the ramped cohorts (Table 1) and independently associated with postinduction hypoxemia (Tables 3 and 4).

The paper is in AEM and therefore very well done. They perform adjusted analyses to try to tease out any real effects. But you cannot infer causation with this study. They do cite one paper on ICU intubations that found no benefit and possible adverse effects of ramping. But this study did not control for … wait for it … apneic oxygenation! Remember the post this week on airway success, apneic oxygenation is awesome, do it. But at least in this paper all patients had apneic O2.

Something else left out is how long they were ramped before intubation (i wouldn’t expect 30 seconds of ramping to help), they excluded trauma patients, they didn’t talk about how ramping can prevent vomiting / aspiration, they did mention that perioperative data suggests benefit to ramping : ).

Overall this paper is worth reading and the stats get pretty thick. Maybe a journal club in the future. But we have to be very careful making practice changes or any strong assertions based on a confounded (direct quote: “we are unable to control for unmeasured confounders”) retrospective paper.

Check out this paper in the upcoming Annals of EM. Data from the Videography in Pediatric Resuscitation (VIPER) Collaborative. Not a huge number of patients (494), but a solid N for a pediatric airway paper.

Research Pearl: Never just read the abstract, at least also look at the tables and figures! Then of course when you write a paper (or even an abstract), spend tons of time on your tables and figures, they are often the most efficient way to convey your findings to the reader.

Much of their findings are of course applicable to adult airway. Some interesting stats:

– The first-attempt success rate was 67%

– Median laryngoscopy duration 35 seconds (interquartile range 25 to 40)

– Hypoxemia occurred in 15% of the patients.

– Videolaryngoscopy was used for at least a part of the procedure in 48% of the attempts, and it had no association with success or the incidence of hypoxemia.

– Intubation attempts longer than 45 seconds had a greater incidence of hypoxemia (29% versus 6%). Furthermore, apneic oxygenation was used in 8% of the first attempts.

***What is happening? Why do we not set up apneic oxygenation on everyone? They had two of the 18 people (11%) with apneic O2 desat, but 18% of those with no apneic O2. Of course 18 is a tiny number and we can’t draw any conclusions, but there is no reason not to throw a nasal cannula (>15L) on every patient you intubate.

Take a look at Table 1 (pasted below), impressive intubation success for EM residents.

Ok that’s probably enough for one post, check out the paper.

Here are four papers on ovarian torsion. If you suspect torsion clinically, do NOT be reassured by normal flow on USN. Only the last paper (12 years old) showed a high sensitivity of ultrasound doppler flow for torsion. The other findings matter!

Diagnostic Efficacy of Sonography for Diagnosis of Ovarian Torsion (2014)

323 subjects. The ultrasound correctly diagnosed 72.1% of ovarian torsion and missed 27.9% of them (false negatives)

Ovarian torsion: Case-control study comparing the sensitivity and specificity of ultrasonography and computed tomography for diagnosis in the emergency department (2014)

20 cases, 20 controls. Pelvic US for ovarian torsion was 80.0% sensitive (95% CI, 58.4-91.9%) and 95.0% specific (95% CI, 76.4-99.1%) for reader 1, while 80.0% sensitive (95% CI, 58.4-91.9%) and 85.0% specific (95% CI, 64.0-95.0%) for reader 2.

Diagnosis of Ovarian Torsion: Is It Time to Forget About Doppler? (2018)

55 cases of surgically proven torsion, 48 controls. Sixty-one percent of right ovarian torsion case and 27% of left ovarian torsion cases had normal Doppler flow. Presence of ovarian cysts was significantly associated with torsion. Sensitivity of ultrasound was 70% and specificity was 87%.

Doppler studies of the ovarian venous blood flow in the diagnosis of adnexal torsion (2009)

One hundred and ninety-nine patients presented with adnexal mass and intermittent lower abdominal pain. Sensitivity and specificity of tissue edema, absence of intra-ovarian vascularity, absence of arterial flow, and absence or abnormal venous flow in the diagnosis of adnexal torsion were: 21% and 100%, 52% and 91%, 76% and 99%, and 100% and 97%, respectively. All patients with adnexal torsion had absent flow or abnormal flow pattern in the ovarian vein. In 13 patients, the only abnormality was absent or abnormal ovarian venous flow with normal gray-scale US appearance and normal arterial blood flow. Of these 13 patients, 8 (62%) had adnexal torsion or subtorsion.

TL;DR

1. Ovarian Torsion is a clinical diagnosis. Ultrasound is NOT 100% sensitive.

2. Read the USN report, Just like a cardiac cath*, normal must really mean normal. If you can’t visualize one ovary, or have normal ovarian flow but a large cyst, or have edema, etc, that is NOT a normal pelvic ultrasound.

* A cath report that has 50% blockage in 2 vessels is not “normal” or “clean”! Caths with absence of a lesion that requires PCI (stent) can still have abnormalities that are very important. Remember, the 50% coronary plaques are the most likely to be unstable and rupture.