Here is another collection of the articles I have found interesting in my reading from the last couple months. For this edition of the Research Roundup we have TB screening, exercise for concussion, conflict of interest, gestalt for ACS, and so much more. Podcast version over on BroomeDocs.
Time to stop yearly TB screening
Sosa LE, Njie GJ, Lobato MN, et al. Tuberculosis Screening, Testing, and Treatment of U.S. Health Care Personnel: Recommendations from the National Tuberculosis Controllers Association and CDC, 2019. MMWR. Morbidity and mortality weekly report. 2019; 68(19):439-443. PMID: 31099768 [free full text]
I had just accepted yearly TB tests as a fact of life. Then I moved to New Zealand and realized that silly exercise in jumping through hoops was not mandatory in other countries. As a result, I had added TB screening to my list of potential myths to look into one day, but the CDC beat me to it. In this new guideline, they recommend against yearly tuberculosis screening for healthcare workers. They still recommend an initial baseline screen when starting your career, and then screening after known exposures when adequate personal protective equipment wasn’t used. They also note that although universal screening isn’t recommended, individual groups at higher risk of exposure (pulmonologists and respiratory therapists) may still warrant yearly screening. Also, individual hospitals and departments might consider their local TB prevalence when deciding what to do. Unfortunately, although these recommendations are based on a systematic review and meta-analysis, this is my most hated type of guideline, in which the evidence is never referenced. Thus, I have no idea what the actual numbers are. What is the benefit? What is the risk? What is the cost? Guidelines should routinely explain the logic for their recommendations, and this one does not.
Bottom line: It might be time to eliminate the mandatory yearly TB screening most of us are subjected to.
Ectopic without a uterus?!?
Fylstra DL. Ectopic pregnancy after hysterectomy may not be so uncommon: A case report and review of the literature. Case reports in women’s health. 2015; 7:8-11. PMID: 29629309
This is an interesting case report of a 32 year old female who 2 years prior had a hysterectomy after a complicated c-section. She presented with acute onset severe abdominal pain, vomiting, and vaginal bleeding. Despite the hysterectomy, her BHCG was 2279. There was an ectopic in the right fallopian tube and about 2 litres of blood in her peritoneal cavity. Although this is rare, there are 72 case reports in the literature describing ectopic pregnancy after hysterectomy! 30 of these cases occurred very early after hysterectomy, and it is assumed that a very early pregnancy was present at the time of surgery (a fertilized egg still in transit in the fallopian tube) or that sperm was present in the fallopian tube at the time of surgery. The other 42 cases occurred well after surgery, so sperm must have found their way into the peritoneal cavity through a vaginal to peritoneal fistula. Changes in surgical technique might actually make this more common with time. The authors suggest that, despite the rarity of the condition, a pregnancy test should be performed in all women presenting with abdominal pain (even if they have had a hysterectomy).
Bottom line: Emergency medicine is scary. Even seemingly impossible things will happen, with potentially deadly consequences. Stay on your toes out there.
Emergency doctors’ psychological well being
Howard L, Wibberley C, Crowe L, Body R. How events in emergency medicine impact doctors’ psychological well-being. Emergency medicine journal : EMJ. 2018; 35(10):595-599. PMID: 30131355
This is a qualitative, interview based study that addresses the important question of how emergency work impacts emergency providers. The authors interview 17 emergency doctors at all levels of training in an open ended interview that started with the prompt, “could you tell me about a time when an event at work has continued to play on your mind after the shift in which it occurred was over?”. This paper discusses some of the themes that arose from those interviews. They are not overly surprising. Doctors were more likely to be psychologically impacted by clinical scenarios involving young patients, trauma, events that can be related to the physician’s personal life, or when the physician felt responsibility. These scenarios resulted in psychological symptoms, physical symptoms, and sleep disturbances. The fact that those results are predictable might be because these themes were all determined a priori, or were expected before the interviews were started. In my mind, the real value of qualitative research is the insight that emerges inductively from the data collected. Luckily, there were a number of “emergent themes” from this research, but the authors saved them for a second paper, so we will have to wait. This research doesn’t provide us with what we really want – solutions to prevent psychological distress from a job we all recognize as taxing – but it is an important step towards codifying and normalizing the impacts of this work.
Bottom line: You are not alone. We all have cases that make us feel this way.
Exercise is good for concussion
Leddy JJ, Haider MN, Ellis MJ, et al. Early Subthreshold Aerobic Exercise for Sport-Related Concussion: A Randomized Clinical Trial. JAMA pediatrics. 2019; PMID: 30715132
It’s not the first time I’ve covered this topic, but many people are still recommending pretty strict rest for patients with concussions (and in Canada I was constantly forced to sign silly school forms that had concussion plans that conflicted with good medical practice). This is an RCT that looked at 102 adolescent athletes (aged 13-18) with concussions. They were randomized to either aerobic exercise or stretching as a “non-exercise” control group starting on day 3 after the injury. All patients had an exercise tolerance test on day #1 and the aerobic exercise was specifically targeted to a heart rate 80% of that which caused symptoms during that exercise tolerance test. Time to total recovery was faster in the aerobic exercise group (13 vs 17 days, p=0.009). Obviously, we won’t be able to directly implement this protocol of heart rate guided exercise in the emergency room. However, there are now multiple studies that all say the same thing: exercise is good after concussion. I routinely tell my patients to start exercising again (in low risk environments where they won’t get another head injury) as soon as they feel capable. Rather than targeting a specific heart rate, I just tell patients to stop and scale back if they develop any symptoms.
Bottom line: If you are still prescribing rest in concussion, it is time to stop.
COACT: Should we rush to the cath lab after ROSC?
Lemkes JS, Janssens GN, van der Hoeven NW, et al. Coronary Angiography after Cardiac Arrest without ST-Segment Elevation. The New England journal of medicine. 2019; 380(15):1397-1407. PMID: 30883057
COACT was a big study that has been discussed at length elsewhere (EMCrit, REBEL EM, The Bottom Line). I thought I had covered it already, but it seems not, so I will include a quick summary for those who haven’t heard these results. This is a multi-center, open label, randomized trial looking at adult patients with out of hospital cardiac arrests with shockable rhythms who remained unconscious after ROSC and did not have STEMI on their ECG. These patients were randomized to either immediate angiography (initiated within 2 hours of randomization) or angiography delayed until after neurologic recovery. The primary outcome of interest was 90 day mortality, and the quick summary is that in the 538 patients included, there was no statistical difference (65% survival with early angiography versus 67% with delayed). There was also no difference in survival with good neurologic outcomes. My biggest caveat when reading this paper is that by studying all comers you may miss a benefit in the smaller minority that actually required an intervention.
Bottom line: The majority of our cardiac arrest patients do not need to be rushed directly to the cath lab after ROSC, although I would still discuss and advocate for patients with a high probability of ACS as the cause of their arrest (significant preceding ischemic symptoms). Also, remember that this doesn’t apply to patients with STEMI – they still need the cath lab.
Gestalt is pretty good in ACS
Oliver G, Reynard C, Morris N, Body R. Can emergency physician gestalt “rule in” or “rule out” acute coronary syndrome: validation in a multi-center prospective diagnostic cohort study. Academic emergency medicine. 2019; PMID: 31338902 [article]
Honestly, this study isn’t that interesting. It won’t change my practice in any way. But I felt the need to include it because I think it could be misinterpreted. The quick conclusion that you will see floating around the internet is “clinician gestalt is not good enough to rule out ACS”. I think it is important to know that this paper doesn’t show that at all. Why? Because this study only included patients with suspected cardiac chest pain. If you suspect chest pain is cardiac, clearly you can’t then use gestalt to rule it out. That is nonsense.
Bottom line: Despite the headlines, it is still OK to rule out ACS without any tests using only your clinical judgement.
You can find the full write up here.
A couple papers on conflict of interest
Niforatos JD, Lin L, Narang J, et al. Financial Conflicts of Interest Among Emergency Medicine Contributors on Free Open Access Medical Education (FOAMed). Academic emergency medicine. 2019; 26(7):814-817. PMID: 31204793
This study looks at possible financial conflict of interest in FOAMed blogs. Specifically, they were looking at 31 blogs recommended by SAEM and EMRA, focusing on posts that were doing critical appraisal of the literature. They used the United States open payments database to determine if authors had financial conflicts of interest, and if so, how often they were disclosed. (I don’t know if First10EM is recommended by anyone, but I have no financial conflicts of interest, as outlined in my disclosure page.) Of 391 total authors, 45 (15%) had a financial conflict of interest with a median payment of $191. The authors considered $5,000 in a year from a single company a “significant conflict”, and there were 12 authors at that level. None of those authors disclosed their conflicts in the blog posts. The open payments system only covers American doctors, so these numbers could be an underestimate. It isn’t clear whether the conflicts were related to the subject matter of the blogs. They used a select group of websites, presumably recommended for quality, so I doubt they are representative of all FOAMed content. Less respected sites may have more conflict of interest. It is possible that some “FOAMed” websites have been created with the sole purpose of advertising. On the other hand, these recommended sites might have more ties to academia, which brings drug dinners and research money into the mix. They also didn’t look at the total number of posts per author, but that could have an impact if people with conflicts are publishing more often and therefore have more influence.
Although this paper is not overly convincing in itself, the potential for hidden conflicts of interest concerns me. FOAMed started as a relatively small collection of doctors with little interest in financial gain. However, now that the format has proven influential, it is almost guaranteed that companies with lots of money will try to capitalize on this audience. Even more than websites, I worry about influential twitter users. You can dig into most websites and get a sense of who is making the recommendations and why. There are so many twitter accounts, potentially anonymous, or even fake, that make medical recommendations. The potential for intentionally biased information is high. Which brings us to our next paper…
Kaestner V, Brown A, Tao D, Prasad V. Conflicts of interest in Twitter. The Lancet. Haematology. 2017; 4(9):e408-e409. PMID: 28863800
This is a research letter looking at financial conflict on twitter among hematologists. In a prior publication, this group showed that 80% of hematologists on Twitter had a financial conflict of interest. (Which is interesting in comparison to the 15% reported in the FOAMed community.) Median payments here were $13,600 (as compared to $191 in the FOAMed crew.) Only 2 of the 156 physicians (1%) disclosed their conflicts of interest on Twitter. This study focused specifically on the content of tweets. They focused on tweets mentioning approved oncology drugs, and compared 100 tweets that had a financial conflict of interest to 100 tweets with no conflict. Tweets with a conflict of interest were more likely to be saying something positive about the drug (66% vs 50%, p=0.02). This is a convenience sample, so the data isn’t perfect, but it fits with what we know about financial conflicts of interest in medicine: lots of doctors have them, and the presence of a conflict of interest definitely makes you more likely to say positive things about the drugs that are sold by the company that is paying you. There is only one good solution to this problem: we need to stop doctors from taking money from the companies that are trying to sell us products. In the meantime, we need to stop letting conflicted doctors take prominent roles in the medical community, whether at conferences, on guideline committees, or on Twitter.
Bottom line: Financial conflict of interest is a problem in medicine. That conflict is sure to find its way into all modes of knowledge dissemination.
Quick update on NSAID ceiling doses
Motov S, Masoudi A, Drapkin J, et al. Comparison of Oral Ibuprofen at Three Single-Dose Regimens for Treating Acute Pain in the Emergency Department: A Randomized Controlled Trial. Annals of Emergency Medicine. 2019;
I’ve discussed the ceiling effect in NSAIDs before (for example: here and here). Most people have heard that 10mg of IV ketorolac is just as good as 30mg, but why are we using IV NSAIDs all the time? Most patients can swallow pills, and oral NSAIDs have been shown to be equally effective to IV or IM, so ibuprofen is probably a good choice for most patients. From the same group that gave us the famous ketorolac RCT, this is another double-blind RCT looking at 3 different doses of ibuprofen in emergency department patients: 400 mg, 600 mg, and 800 mg. The reduction in pain was the same in all 3 groups (a drop from about 6/10 to 4/10), so like all other NSAIDs, ibuprofen has a ceiling. There is no reason to use more than 400mg for analgesia. That being said, they only measured pain at 60 minutes in this trial, and it is possible that larger doses would offer longer pain control. (I am still unsure about the appropriate dosing for inflammatory conditions, like gout, but multiple rheumatologists have told me higher doses are required.) If you want more detail on this trail, check out the write up at REBEL EM.
Bottom line: The appropriate maximum dose of ibuprofen is 400mg.
I love winter in New Zealand
Shah S, Murray J, Mamdani M, Vaillancourt S. Characterizing the impact of snowfall on patient attendance at an urban emergency department in Toronto, Canada. The American journal of emergency medicine. 2019; 37(8):1544-1546. PMID: 31201115
I am now more than halfway through a winter without seeing any snow (except on the peak of a mountain many miles away). In that context, this paper from back home in Toronto seemed like fun. These authors retrospectively looked at the correlation between snowfall and daily attendance in the emergency department. They looked at 2542 days, and contrary to some peoples’ impression of my country as an icy wasteland, it only snowed on 11% of those days, and only 1.5% of days had greater than 5 cm of snowfall. There was a statistical decrease in the number of patients presenting to the hospital on days that it snowed. On days with more than 1 cm of snow, there was a 2.6% decrease in patient volumes, or about 5 fewer patients at this hospital. The problem, which anyone working in Canada knows, is that these patients aren’t evenly distributed over the hours of the day. In the middle of the snow storm, volumes dip as smart people stay inside drinking hot chocolate. However, as soon as the snow starts to lighten up, people head outside, which means a large bolus of patients is headed our way, whether it is the 65 year old with an MI from shovelling snow, or just the slew of broken wrists and hips from the icy sidewalks. I think these authors need to go back and get even more granular with their data, breaking it down hour by hour, if they want to get a real idea of the impact on snowfall on emergency department presentations.
Bottom line: Although snowfall might result in 5 fewer presentations to the ED, I would still prefer the warm weather of winter in New Zealand.
Cheesy Joke of the Month
I accidentally handed my wife the superglue instead of her lipstick. She still isn’t talking to me.
