A monthly collection of the most interesting emergency medical literature I have encountered
Here is this month’s summary of my favorite reads from the medical literature.
Troponin is king – why even send an CK?
Le RD et al. Clinical and financial impact of removing creatine kinase-MB from the routine testing menu in the emergency setting. Am J Emerg Med. 2015;33(1):72-5. PMID: 25455047
This is an observational study, looking at a period before and after CK-MB was removed from an automatic order set. Out of 6444 cases included in the study, there were only 17 cases with a positive CK-MB fraction and a negative troponin. All 17 were ultimately determined by the treating physicians to have non-ACS causes (ie, they were false positives). So, CK-MB was not clinically helpful. Removing it from the order set dropped ordering by 80% and saved the hospital about $47,000 a year.
Bottom line: We might want to keep this one in our back pocket for the next time the hospital demands cost savings – dropping the CK helps us and saves money
Speaking of troponin – high sensitivity and the 1 hour rule out
Reichlin T et al. Prospective validation of a 1-hour algorithm to rule-out and rule-in acute myocardial infarction using a high-sensitivity cardiac troponin T assay. CMAJ. 2015 (In Print). PMID: 25869867
This prospective observational study of 1320 chest pain patients attempted to validate a 1 hour rule out protocol. Using high sensitivity troponins, patients ruled out if they had trop of 12ng/L or less and a 1 hour delta of 3mg/L or less. They ruled in with a trop of 52ng/L or more or a 1 hour delta of 5ng/L or more. Everyone else was put in longer observation. It was a relatively high risk cohort, with 17% overall having an acute MI. 60% of patients were able to be ‘ruled out’ at 1 hour, and only one of those patients (0.1%) ultimately had an MI. It ruled in 16% of the patients at 1 hour, with 78% being true positives. The remaining 24% that couldn’t be ruled in or ruled out had an 18% chance of an MI – so the prolonged observation work up makes a lot of sense.
Bottom line: This could work (if we had the right assay), but I think our rule in rate for MI is way less than 17% – so this strategy could actually increase our testing and admissions without benefit to our patients
How often to you order pregnancy tests just for medication use?
Goyal MK et al. 2015. Underuse of pregnancy testing for women prescribed teratogenic medications in the emergency department. Academic Emergency Medicine (in print). PMID: 25639672
A retrospective study using the NHAMCS database (notoriously poor data) but still raises an interesting point. Looking at all women who were given or prescribed FDA pregnancy category D or X medications, only 22% had pregnancy testing done. (I will note that this is one area where I don’t trust NHAMCS at all – there was one study where 50% of patients diagnosed with ectopic pregnancies didn’t have a pregnancy test done – but then how did they get diagnosed with ectopic pregnancy?) This also doesn’t tell us how many of these women were actually pregnant, so it is difficult to tell how big an issue this really is.
Bottom line: Are you checking for pregnancy before giving Advil to ankle sprains in ambulatory care? Should we have quicker point of care testing to make this feasible? Does it matter?
Non-news of the month: there happen to be some bacteria in your blood post CPR
Coba V et al. The incidence and significance of bacteremia in out of hospital cardiac arrest. Resuscitation. 2014 Feb;85(2):196-202. PMID: 24128800
I ignored this one when it first came around a year ago, but I have heard it repeated so many times, with strange conclusions, that I guess it should be included. This is a prospective observational study of 250 adult out of hospital cardiac arrest patients who they drew blood cultures on in the ED, 38% of whom were found to be bacteremic. But come on, you get bacteremic after brushing your teeth. Are you surprised this happened with crash airways, CPR, and broken ribs? They note that mortality was higher in the bacteremic group, but again, in dead people as mucous membranes break down, I expect more bacteremia. This is a silly surrogate outcome, unless someone can show early antibiotics save lives.
Bottom line: Try to ignore this paper when it is mentioned over and over again in the coming years
Another one with strange conclusions
Schuch S et al. Effect of oximetry on hospitalization in bronchiolitis: a randomized clinical trial. JAMA. 2014;312(7):712-8. PMID: 25138332
This is a double blind RCT from Sick Kids, where they took 213 infants with bronchiolitis and randomized them to either have an accurate pulse ox reading, or one that displayed values that were 3 points higher than the actual value. When higher oxygen sats were shown, admissions went down from 41% to 25%. This is obvious – we admit hypoxic patients. I have heard lots of doctor bashing around this, but what this study didn’t show was that it was safe to discharge home babies with borderline sats. I admit a child with a sat of 89% because they are right at top of the steep part of the oxygen desaturation curve, and I am worried they might get worse. Telling me that the sat is 92% might change my mind – but how do we know those kids didn’t go on to have complications? This study certainly didn’t look for it. (I will admit we probably over-rely on the sat – but until someone proves 89% is safe with no treatment or monitoring, I will keep admitting.)
Bottom line: If you lie to doctors about important clinical parameters, their decisions change
Once again, forget about atypicals in the treatment of community acquired pneumonia
Postma DF et al. Antibiotic treatment strategies for community-acquired pneumonia in adults. NEJM. 2015;372(14):1312-1323. PMID: 25830421
Despite the theory of needing to cover for atypical organisms, this study is another in a long line of papers that all say the same thing. This is a large, multi-centre cluster-randomized trial of 2283 adult patients with community acquired pneumonia who did not require ICU care. They randomized months to to either use beta-lactam monotherapy, a beta-lactam plus a macrolide, or a fluroquiolone. The primary outcome was mortality at 90 days, and was statistically the same in all groups (but actually 1.9% higher in the macrolide group.) Secondary outcomes, like length of stay, were also the same. (The authors do note that during the time of the study, there was a low incidence of atypicals. However, multiple previous studies have show atypicals don’t matter, except maybe legionella.)
Bottom line: We already knew this, but are always taught differently: you don’t need to add a macrolide to beta-lactams to treat community acquired pneumonia. (Empiric evidence trumps petri dishes every day.)
Dental abscesses are like all abscesses – antibiotics don’t help
Tichter AM and Perry KJ. Are antibiotics beneficial for the treatment of symptomatic dental infections? Ann Emerg Med. 2015;65(3):332-3. PMID: 25477181
This systematic review was able to find 2 RCTs comparing antibiotics (both pen-VK) versus placebo for apical perdiodonitis or abscess. There was no difference in pain, swelling, or infection progression at 24, 48, or 72 hours. All patients were given oral analgesics and ultimately had the definitive management – surgical pulpectomy.
Bottom line: Dental infections are one more diagnosis where we give antibiotics but probably shouldn’t
Was this patient’s DVT caused by an unknown cancer?
Robertson L et al. Effect of testing for cancer on cancer- and venous thromboembolism (VTE)-related mortality and morbidity in patients with unprovoked VTE. Cochrane Database Syst Rev. 2015 [Epub ahead of print] PMID: 25749503
We know that cancer is a risk factor for VTE, so we frequently ask ourselves should we be searching for a potential cancer in people with an apparently unprovoked VTE? This is a Cochrane review, but they could only identify 2 studies with a total of 396 patients – so interpret with caution. Using a a specific suite of screening tests post VTE diagnosis, they did make more early diagnoses of cancer than in patients with usual care, but they were unable to find any cancer specific mortality benefit. (They didn’t even measure all cause mortality.)
Bottom line: This fits well with most screening data we have, in that we can always find more cancer if we look, but we are not good at changing mortality or quality of life (for the better)
More is not always better
Minotti V et al. A double-blind study comparing two single-dose regimens of ketorolac with diclofenac in paindue to cancer. Pharmacoptherapy. 1998;18(3):504-8. PMID: 9620101
With recent drug shortages, the topic of the appropriate ketorolac dose was raised a number of times around the department. This is a double blind RCT comparing ketorolac 10mg or 30mg or diclofenac 75mg (all IM) in adults with acute cancer pain. All three provided equal and reasonable relief over 6 hours. I just picked one, but this is consistent with multiple other studies showing 10 mg = 30 mg of ketorolac.
Bottom line: Toradol 10mg is probably identical to 30mg
We know we don’t talk to our patients – but apparently we can’t even talk to each other
Venkatesh AK et al. Communication of Vital Signs at Emergency Department Handoff: Opportunities for Improvement. Annals of Emergency Medicine. 2015 (in press). PMID: 25805116
This was a prospective observational study looking at ED handoffs. Out of 1163 total handoffs observed, 117 patients had episodes of hypotension, and they were not mentioned for 66 patients (42%). There were 156 patients with hypoxia, and 116 (74%) were not mentioned. (These numbers seem unbelievable, and if you look closer, attending docs rarely left this info out, it was primarily residents.)
Bottom line: Handoffs are important. Take a minute to review all the information. And we should probably be emphasizing this in resident education
Should H.pylori be an ED problem?
Meltzer AC et al. Treating Gastritis, Peptic Ulcer Disease, and Dyspepsia in the Emergency Department: The Feasibility and Patient-Reported Outcomes of Testing and Treating for Helicobacter pylori Infection. Annals of Emergency Medicine. 2015 (in press). PMID: 25805114
This is a prospective cohort study on a convenience sample of ultimately 212 patients. The attending doctor was asked if the patients’ symptoms could be attributed to gastritis, PUD, or dyspepsia, and if so they tested for H.pylori and treated if positive. 23% of the patients tested positive for H.pylori. With treatment, they were able to eradicate H. pylori in 41% of those patients. At 3 weeks, the pain scores seemed to have decreased about the same amount no matter what had happened to you. For me, this could go either way. I worry about the false positives and a potential anchoring bias where we say this pain couldn’t be ACS just because the patient is H.pylori positive. However, our patients may benefit from early treatment (though they didn’t in this study).
Bottom line: H. Pylori is probably the cause of a lot of the symptoms we see, but we currently don’t have any good strategy to address that
The “rocket launcher” hip reduction technique
Dan M et al. Rocket launcher: A novel reduction technique for posterior hip dislocations and review of current literature. Emergency Medicine Australasia. 2015 (in press). PMID: 25846901
This is a case report of 6 patients, so I wouldn’t pay any attention to the EBM side of things. They describe a technique for hip reduction I hadn’t heard of, and may be helpful for some, especially if you are to short to make the Captain Morgan easy. Essentially, you adjust the height of the bed so that you can put the patients knee over your shoulder. The foot faces forward, like you might picture someone holding a bazooka or ‘rocket launcher’. This allows you to use you shoulder as a fulcrum, and lift with your legs.
Bottom line: Captain Morgan is still my go to, but its nice to have this as a backup
Another reduction technique: syringe rolling for mandible reduction
Gorchynski J et al. The “syringe” technique: a hands-free approach for the reduction of acute nontraumatic temporomandibular dislocations in the emergency department. J Emerg Med. 2014;47(6):676-81. PMID: 25278137
This technique involves placing a syringe (5 or 10cc) between the posterior molars, and then turning the syringe in the direction that would push the mandible backwards (as if a wheel were rolling forward along the bottom teeth). In this prospective, convenience sample, they were successful in 30/31 attempts, with 24 of those attempts taking less than a minute. You can do this without sedation. In fact, patients can do this for themselves.
Bottom line: I haven’t tried it yet – let me know if you do
Angioedema of the bowel: I’ve probably seen it, but I’ve never diagnosed it
Bloom AS and Schranz C. Angiotensin-Converting Enzyme Inhibitor–Induced Angioedema of the Small Bowel—A Surgical Abdomen Mimic. Journal of Emergency Medicine. 2015 (In Press). PMID: 25886983
Just a case report, but I include it because we probably see this, but I had never really heard of it. We won’t necessarily rule it in, but in recurrent abdo pain, I might consider stopping an ace inhibitor as a trial. They note that CT findings, if you happen to get one, include ascites, small bowel thickening and straightening, and dilatation without obstruction.
Bottom line: Medication side effects should be part of the differential diagnosis for every chief complaint
Old people have high D-dimers – don’t send them if you can avoid it, but if you have to…
Righini M et al. Age-adjusted D-dimer cutoff levels to rule out pulmonary embolism: the ADJUST-PE study. JAMA 2014;311(11):1117-1124. PMID: 24643601
This is a prospective observational study of 3346 patients with suspected PE (the total rule in rate was 19%), of which a total of 331 had D-dimers greater than 500, but less than age x 10. Using the adjusted D-dimer level of age x 10, they would have missed 1 PE out of 331 patients (0.3%). Unfortunately, not everyone got the gold standard test (CTPA), so it is possible they missed a few more that we don’t know about. However, if the test threshold for PE generally is 2%, and the elderly are particularly prone to renal problems from CT contrast, avoiding 331 CTPAs at the cost of one missed diagnoses might be worth it. The other major problem is that D-dimers are not standardized and there are multiple different assays.
Bottom line: If the D-dimer is less than age x 10, the risk is probably low enough to stop further testing. I use this to (and this is crazy, I know) talk to my patients about whether or not to scan
Clowns cause pregnancy; AKA completely irrelevant paper of the month
Friedler S et al. The effect of medical clowning on pregnancy rates after in vitro fertilization and embryo transfer. Fertility and Sterility. 2011;95(6):2127-2130. PMID: 21211796
This is just too good not to include. Give women IVF, and then let them play with a clown and 36.4% become pregnant. Remove the clown: only 20.2%.
Bottom line: What exactly are they doing with that clown?
#FOAMed suggestion of the month
If you haven’t come across it yet, Scott Weingart and Steve Smith put together a list of all the reasons for cath lab activation, including the very subtle details. There are 2 podcasts summarizing, and one very handy pdf. Also, Steve Smith is just giving away his amazing ECG textbook. All can be found at:
Cheesy Joke of the Month
Why don’t you ever see Hippos hiding in trees?
Because they are really f***ing good at it.