Articles of the Month (January 2017)

Another month and another batch of articles to keep your practice informed. As always, I have no specific selection criteria. These are all just papers that I found interesting. I am always happy to receive suggestions if you encounter a paper that makes you think. And, of course, make sure to have a listen to me and Casey Parker making fools of ourselves as we try to come up with intelligent things to say about these papers on the BroomeDocs podcast.


Gotta shock em all?!

Liu WC et al. Prognostic impact of restored sinus rhythm in patients with sepsis and new-onset atrial fibrillation. Crit Care. 2016 Nov 18;20(1):373. PMID: 27855722 [free full text]

Liu 2016 Figure 2.PNGI have previously covered a paper that illustrated that actively managing atrial fibrillation in patients who have underlying medical conditions does not help and is associated with more adverse events. This paper has been circulated around with essentially the opposite claim. It concludes that in patients with sepsis cardioversion out of atrial fibrillation is associated with a lower mortality. So maybe we should be cardioverting these patients? But after reading the paper, I am pretty sure that conclusion is wrong. This is a retrospective chart review where they identified all patients admitted to the ICU with sepsis over 3 years at a single centre. The first major problem with this paper is their exclusion criteria: patients with an ICU stay of less than 3 days are excluded. This is a common mistake in retrospective research. It is easy, when looking retrospectively, to determine which patients were in the ICU less than 3 days, but I cannot predict that at the outset, so I have no idea which patients to apply this data to. Ultimately, they include 503 patients and break them into 3 groups: no atrial fibrillation, atrial fibrillation that was converted to sinus, and atrial fibrillation that remained in atrial fibrillation. They report in-hospital morality of 17.5%, 26.1%, and 61.3% (p<0.01 for all comparisons) respectively, which is what leads to the conclusion that we might want to cardiovert these patients. However, the fatal flaw is in how they defined these groups. It wasn’t patients they attempted to cardiovert versus those who they didn’t; it was patients who they were successful in cardioverting (or who converted spontaneously) versus those who they were unsuccessful or just didn’t try. In other words, your group was determined not by the medical management you received, but just by whether or not you happened to have any more atrial fibrillation over the following week. What that means is that we are just comparing a sicker group of patients to a healthier group, and not surprisingly, the sicker group has a higher mortality. This is confirmed by looking through the comparisons between the two groups: the group that remained in atrial fibrillation had a higher APACHE II score, higher SOFA score, more pressor use, and even more attempts at cardioversion than the group that returned to sinus rhythm.

Bottom line: The only conclusion here is that sicker patients die more often. This data does not support cardioversion in sepsis.


Make sure your syncope patients have excellent car insurance

Numé AK, Gislason G, Christiansen CB. Syncope and Motor Vehicle Crash Risk: A Danish Nationwide Study. JAMA internal medicine. 176(4):503-10. 2016. PMID: 26927689

The results of this paper surprised me and will probably change my practice, although I’m not sure how. This is a large registry study done in Denmark, where every patient encounter is entered into a national registry. They looked at all patients with a first time diagnosis of syncope. This included 41,039 patients over a 4 year period. (They can only identify these patient based on their ICD10 codes, and based on other research, they think this coding will catch about 66% of syncope patients.) In this group of patients, they then looked for subsequent MVCs that were either fatal or severe enough to require emergency department presentation. The big number here: 4.4%. That is how many patient had a significant MVC in the median 2 year follow up period. 0.3% of these MVCs were fatal and 78.1% resulted in injury. The crude rate of MVC was 20 per 1000 person years in this syncope population, as compared to 12 per 100 person years in the general population. One thing I alway worry about when reading European studies is selection bias. In many European countries, almost all patients are seen by their primary care physicians before being referred the the emergency room, which means the emergency department patients are a much sicker cohort than we are used to seeing in North America. I don’t know how the Danish system works. Maybe a reader from Denmark will be able to comment on this? However, if this is the case, this could be an overestimate. There is a hint that this population is sicker than what we see in Canada: 62% of the syncope patients were admitted. That number is an order of magnitude higher than my admit rate for syncope. This all leads to some important questions: Is this just a sign that people who faint are older and sicker? Can we prevent these MVCs? Is this rate high enough that we should be taking away people’s licenses? Will self driving cars solve all our problems? (I only have an answer to that last question, and it is clearly: yes.)

Bottom line: 1 in 20 syncope patient here were involved in an MVC within the next 2 years. That clearly needs to be discussed with our patients and their families.

I have to point out that I probably wouldn’t have found this paper if it wasn’t for the excellent EBM summaries done by the BEEM group. This is one of the papers I will be reviewing at sunBEEM in the Domincan Republic February 16-18. It’s going to be a lot of fun, and educational too, and the best news is: there is still time to sign up. See you there!


Clavicles need surgery?

Smeeing DP, van der Ven DJ, Hietbrink F. Surgical Versus Nonsurgical Treatment for Midshaft Clavicle Fractures in Patients Aged 16 Years and Older: A Systematic Review, Meta-analysis, and Comparison of Randomized Controlled Trials and Observational Studies. The American journal of sports medicine. 2016. PMID: 27864184

This is a systematic review and meta-analysis that includes 8 RCTs and 12 observational studies. I haven’t read the original research, but they are reported as reasonable quality, although no specific information is given in the paper about methodology such as blinding. There also doesn’t seem to be any significant publication bias, based on their funnel plot. Contrary to what we were all taught, this meta-analysis concludes that surgical management of mid-shaft clavicle fractures decreases non-unions (the primary outcome; odds ratio 0.18 95% CI 0.10-0.33). The absolute numbers here would appear to be clinically significant, with 10.5% of the non-surgical group experiencing a non-union, as compared to only 1.4% of the surgical group. That rate of nonunion seems high to me, but I don’t see these patients in follow up, so I am not sure. When looking at only the high quality studies, surgery also appeared to improve malunions and return to work. Unfortunately, they don’t discuss harms at all.

Bottom line: Although we don’t make the decision about surgery in the ED, we shouldn’t be surprised if more of these patients end up with surgical management, and we should counsel them appropriately.


It’s a good month to be an orthopod

Clark W, Bird P, Gonski P. Safety and efficacy of vertebroplasty for acute painful osteoporotic fractures (VAPOUR): a multicentre, randomised, double-blind, placebo-controlled trial. Lancet. 388(10052):1408-1416. 2016. PMID: 27544377

This is another interesting paper about a treatment we don’t actually do. This is a randomized, double-blinded trial comparing vertebroplasty to a sham procedure in 120 patients 60 years and older with osteoporotic compression fractures less than 6 weeks old. 3 patients in each group (5%) died at 6 months, which gives you a sense of the age and comorbidities of these patients. The primary outcomes was the proportion of patients with pain less than 4/10 at 14 days, and it was higher (better) in the vertebroplasty group (39% vs 20%; NNT = 6 95% CI 3 – 43). These differences appear to hold out to 6 months in terms of absolute numbers, but were not statistically significant. There were harms in both groups. In the treatment group, one patient had a respiratory arrest during sedation (but was successfully resuscitated) and another suffered a broken humerus while they were transferring her to the procedure table. Two patients in the control group had retropulsion of the fracture fragments, one had surgery and the other, not being a surgical candidate, was left paraplegic. These are pretty significant adverse events in both groups. Almost every orthopedics paper I read is about a surgery we are currently doing that doesn’t work (knee scopes, etc), so I was excited to be able to include 2 papers that indicated surgery might actually help our patients. That being said, we really need to see a much bigger trial of vertebroplasty to confirm long term benefit and further clarify the possible harms.

Clark 2016 Figure 2.PNG

Bottom line: There may be some role for vertebroplasty in compression fractures, but we really need to see this study repeated in a larger population.


I don’t even like thinking about this: zippers and foreskin

Oquist M, Buck L, Michel K, Ouellette L, Emery M, Bush C. Comparative analysis of five methods of emergency zipper release by experienced versus novice clinicians. Am J Emerg Med. 2016 PMID: 27836312

This is a prospective, randomized trial of different zipper release techniques using a simulated model (chicken skin stuck in the zipper instead of foreskin). A group of medical students and emergency medicine faculty all tried 5 techniques for releasing the zipper in random order: cutting the median bar, using a screwdriver to separate the face-plates, using mineral oil as lubricant, lateral compression of the zip fastener using pliers, and removal of teeth of the zip mechanism using trauma scissors. The most successful technique was simple manipulation after application of mineral oil, and it was also the quickest. Of course, I don’t care as much about the time it takes, unless the child is freaking out. I care most about pain, which this study can’t tell us about. The technique that resulted in the least damage to the skin was cutting the closed end of the zipper with trauma scissors and allowing it to unzip backwards. Lateral compression and rotating the screw driver were both unsuccessful and more damaging. Clearly this is a tiny study using chicken skin, but I think it is interesting, and with finicky procedures like this, it always good to have a few backup techniques.

Bottom line: Scissors and mineral oil will be my first options when this comes up.


Anxiety as a cause of chest pain and shortness of breath

Musey PI Jr, Kline JA. Emergency Department Cardiopulmonary Evaluation of Low-Risk Chest Pain Patients with Self-Reported Stress and Anxiety. J Emerg Med. 2016 PMID: 27998631

Patients have anxiety. Patients are allowed to have anxiety. I sometimes think emergency medicine culture disagrees. At conference after conference I am told “diagnose anxiety at your own peril”. The idea is that if we make a misdiagnosis, telling a patient that they have anxiety is interpreted as “this is all in your head” and that comment looks bad in front of a jury. This is an unplanned secondary analysis of a prospective cohort of 851 patients who presented to the emergency department with chest pain and shortness of breath. When asked “what do you think caused your chest pain?” 67 patients (8%) responded that it was caused by stress or anxiety. Obviously, there may have been other patients who were unwilling to admit to being anxious or unaware that their symptoms could be caused by anxiety. Of these 67, none were diagnosed with either MI or PE. Zero. That is low risk, but the overall rule in rate was only 3% for ACS and 2% for PE, so this was an incredibly low risk cohort overall. Also this is a secondary analysis of a small number of patients. But these patients did get a couple thousand dollars worth of tests and a couple millisieverts of radiation. I don’t doubt that we do too many tests in patients with anxiety. That’s what we are taught to do. That’s not why I include this paper. I include the paper because of these 67 patients, all of whom told their doctor exactly what was bothering them, not a single one received either a diagnosis of anxiety or any treatment for anxiety. Sure, run your tests to rule out ACS if you think you have to, but at least treat the patients for the condition they are telling you they have.

Bottom line: Anxiety is one of the causes of chest pain and shortness of breath. It scares us to talk about it, but our patients need our diagnosis and treatment.


You got what it takes to diagnose acute heart failure?

Martindale JL, Wakai A, Collins SP. Diagnosing Acute Heart Failure in the Emergency Department: A Systematic Review and Meta-analysis. Academic emergency medicine. 23(3):223-42. 2016. PMID: 26910112

This is a systematic review and meta-analysis that included 52 studies covering 17,893 patients looking at the diagnosis of acute heart failure. One major problem is that there is no gold standard for CHF, so all these studies had to use a physician’s judgement, which incorporates the history and physical, which makes conclusions about accuracy a little circular. It is worth looking through the numbers. On their own, no aspect of the history or physical exam is good enough to rule in or rule out CHF. JVD only has a sensitivity of 37%; orthopnea 52%; rales 60%. A finding of a third heart sound is the best for ruling in CHF, but still only has a positive likelihood ratio of 4.0. (Ultrasound nerds will probably want to point out that lung ultrasound was definitely the best test available, with a positive likelihood ratio of 7.4 and a negative likelihood ratio of 0.16). I include this paper for a few reasons. First, I do think it is a good paper to read through as a reminder of the accuracy of our various tests for CHF. Second, I think it is important to point out that, although we often treat it as such, CHF is not a single unified diagnosis. CHF is actually used to describe a number of very different pathophysiologies. Cardiogenic shock, acute valve failure, arrhythmia, gradual fluid overload, and hypertensive flash pulmonary edema from sympathetic overdrive are all very different. Treatments are different and presentations are different, so grouping them together is problematic. Third, although no single aspect of the history or physical is good enough to rule in or rule out CHF, many of the likelihood ratios are in a moderate range, and a combination of multiple factors (also know as our usual clinical assessment) could end up being pretty accurate. Finally, and most importantly, almost all of these diagnostic studies look at our clinical assessments as if they are performed at a single point in time, but all of our tests can be repeated over time, before and after interventions, and that will almost certainly increase their accuracy dramatically. My diagnosis of CHF 20 minutes into an emergency department visit might not be that accurate. Add multiple reassessments and a few days, and it probably will be. Diagnosis is not – and should not be taught as or studied as – a one time event. Diagnosis is iterative, additive, and always a little uncertain. When you combine this all together, I think that we are a lot better at diagnosing CHF than the numbers here would lead us to believe.

Martindale 2016 Table 1.PNG

Bottom line: No single factor taken in isolation can either rule in or rule out acute heart failure, a surprise to nobody.


Physio for ankle sprains? Think again

Brison RJ, Day AG, Pelland L, Pickett W, Johnson AP, Aiken A, Pichora DR, Brouwer B. Effect of early supervised physiotherapy on recovery from acute ankle sprain: randomised controlled trial. BMJ. 2016 Nov 16;355:i5650. PMID: 27852621 [free full text]

I actually enrolled a few patients to this trial back in residency. This is a randomized controlled trial in 503 adults with grade 1 or 2 ankle sprains comparing usual care (a 1 page handout describing analgesics, elevation, ice, graduated weight bearing, and expected recovery) to supervised physiotherapy sessions (up to 7 sessions). When this trial was enrolling, I was pretty convinced that physiotherapy would provide a huge benefit, so if I had seen any friends in the ED with ankle sprains, I would have probably tried hard to get this into the “good” group. Luckily this trial had excellent randomization and allocation concealment procedures, so I couldn’t screw it up. And boy was I wrong. There weren’t differences in anything, and they measured a lot. There are probably a lot of different way to do physiotherapy, so an expert might be able to tell me why the specific plan they used here didn’t work, and it doesn’t tell us what to about more severe sprains.

Brison 2016 figure 4.jpg

Bottom line: I won’t be referring ankle sprains to physiotherapy any more.


I’m sorry to tell you, but if you are over 50 you are “older”. (Also, the treatment for anaphylaxis is still epinephrine)

Kawano T, Scheuermeyer FX, Stenstrom R, Rowe BH, Grafstein E, Grunau B. Epinephrine use in older patients with anaphylaxis: Clinical outcomes and cardiovascular complications. Resuscitation. 2017 PMID: 28069483.

This is a retrospective chart review that looked at 492 adults meeting criteria for anaphylaxis at one of two emergency departments over a 5 year period. They looked to see how often older patients were treated with epinephrine and if there were complications. Older was defined as over 50 years old – sorry if that offends you. These older patients were less likely to get epinephrine (36% vs 60%), but for some reason the older population was much more likely to be given an excessive dose of epinephrine (more than 0.5 mg IM or 100 mcg IV; 16% vs 1%). Either way, those are really low rates of epinephrine use for anaphylaxis, and I agree with the definition of anaphylaxis that they used. There were 3 “cardiac complications” in patients who received IV epinephrine, all of whom received an excessive dose (300 mcg IV push). None of these were real complications: 1 asymptomatic and completely resolving ST depression, 1 self resolving 3 minute run of stable ventricular tachycardia with no further events, and 1 run of atrial fibrillation so short they couldn’t catch it on an ECG, followed by some transient ST depression and a negative stress test. There were 2 cardiac complications in the IM epi group, both at normal doses, and both in the elderly. Again, they don’t sound too bad. One was transient 3/10 chest pain and ST depression that resolved without therapy and didn’t require intervention. Another patient developed ST depression with sinus tachycardia and did have a slight troponin elevation, but that was followed by a completely normal cardiac catheterization. With retrospective data, we have to remember that there might have been a reason that epi was withheld in so many of the older patients.

Bottom line: This data reinforces that epinephrine is safe to give and should be given in all age groups in the presence of anaphylaxis, but please know the dose of IV epi


Cheesy Jokes of the Month

I really hate it when people ask me what I will be doing in 3 years.

I mean, come on! I don’t have 2020 vision!

Author: Justin Morgenstern

Community emerg doc, FOAM enthusiast, evidence junkie “One special advantage of the skeptical attitude of mind is that a man is never vexed to find that after all he has been in the wrong.” - William Osler

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s