Articles of the Month (December 2016)

Welcome to the year-end edition of the Articles of the Month (released well into the new year because of the craziness of emergency department holiday schedules). The podcast version of this post with Casey Parker is available through the BroomeDocs podcast.

I’ve always thought kids should come with warning labels

Seiger N, Maconochie I, Oostenbrink R, Moll HA. Validity of different pediatric early warning scores in the emergency department. Pediatrics. 132(4):e841-50. 2013. PMID: 24019413 [free full text]

Our department has just seen the implementation of a pediatric early warning score (PEWS). Is it any good? This is a large prospective cohort of 17,943 children in which they attempted to validate 10 different PEWS. It is a relatively sick cohort as compared to what I see, with 16% of the children admitted to hospital and 2% admitted to ICU. There are a ton of numbers in this paper, because there are so many scores being used. The scores aren’t great. The sensitivity for admission to ICU ranges from 61-94% and the specificities were 25-85%. The numbers for hospital admission were a little worse (sensitivities 25-85% and specificities 27-80%). The simple fact that there are so many scores is probably the best hint that none of the scores is any good. As an emergency physician, I can generally tell which kids need ICU level care, and a score with poor positive and negative likelihood ratios is not going to help me at all. Given that I don’t think we are missing very many children who ought to be admitted to hospital (and that for most conditions a hospital admission will not influence outcomes) the low specificities of these scores means that they will result in a lot of false positives and needless hospitalizations without providing any real value.

Bottom line: According to this cohort, none of the available PEWS are good enough for clinical practice.

---

Can’t break through the ketorolac ceiling

Motov S, Yasavolian M, Likourezos A. Comparison of Intravenous Ketorolac at Three Single-Dose Regimens for Treating Acute Pain in the Emergency Department: A Randomized Controlled Trial. Annals of emergency medicine. 2016. PMID: 27993418

We have always sort of known that 30 mg was too high a dose for ketorolac. The oral dose is 10 mg and yet we give 30 mg IV? This study that should put the issue to bed. It is a randomized, double-blind trial of 240 adult patients presenting to the emergency department with painful conditions in which the physician would use ketorolac. They compared 10 mg to 15 mg and 30 mg IV. (The medication was given as a slow IV push, which is nice to note, as it is probably a lot quicker than the slow drip mini bag I always see being used.) For the primary outcome of pain reduction at 30 minutes, there was no difference between the groups. The improvements, on a 10 point scale, were 2.5 with 10 mg, 2.4 with 15 mg, and 3.0 with 30 mg. There were no differences in any outcomes out to 120 minutes. Although there were no differences in adverse events across the groups, the trial is not powered to determine safety. For every medication we have, there is a dose response relationship, such that higher doses have higher rates of adverse events.

Bottom line: It is time to stop using any dose of ketorolac higher than 10mg.

---

Another NSAID myth: bad outcomes in fractures

DePeter KC, Blumberg SM, Dienstag Becker S, Meltzer JA. Does the Use of Ibuprofen in Children with Extremity Fractures Increase their Risk for Bone Healing Complications? The Journal of emergency medicine. 2016. PMID: 27751698

This is a retrospective cohort of children aged 6 months to 17 years with fractures of the tibia, femur, humerus, scaphoid, or fifth metatarsal (chosen because these are bones with higher rates of healing complications). They identified 1192 children with these injuries, but 298 did not have follow up with their orthopedic service, so were excluded. 808 patients were included in the final analysis, and 338 (42%) had exposure to ibuprofen. There were 27 (3%) bone healing complications: 8 non-union, 3 delayed union, and 16 re-displacements. The rate of complication was the same with exposure to ibuprofen (3%) as it was without (4%; odds ratio 0.8, 95% confidence interval 0.4–1.8; p = 0.61). Of course, accurately determining exposure to ibuprofen from a chart review, and accounting for home use, is not easy and could muddy the results.

Bottom line: This is another piece of evidence that NSAIDs don’t cause bone healing complications, and that we should just treat children’s pain

---

Anesthesiologists and their oral fixation

Beach ML et al. Major Adverse Events and Relationship to Nil per Os Status in Pediatric Sedation/Anesthesia Outside the Operating Room: A Report of the Pediatric Sedation Research Consortium. Anesthesiology. 124(1):80-8. 2016. PMID: 26551974

I’ve covered presedation NPO time before. I don’t think it’s relevant in emergency medicine. The ACEP clinical policy agrees. If there are any remaining hold outs, here is some more data for you. This is data from the Pediatric Sedation Research Consortium. It is prospectively collected data on a huge number of sedations (139,142) over 4 years at 42 hospitals. There were 25,401 patients who were not NPO prior to their sedation. Aspiration occurred in 8 patients who were NPO and 2 who weren’t, event rates of 0.8/10,000 and 1/10,000 respectively (odds ratio, 0.81; 95% CI, 0.08 to 4.08; P = 0.79). There also wasn’t any difference in major complications. Fitting with other data, sedation is incredibly safe, with major complications only occurring in 6/10,000 patients. One limitation is that NPO status was not available for 31,195 patients. Also, this is observational data, so it can only tell us about associations, rather than demonstrating cause and effect. But what this shows us is that the aspiration even rate is so low, even when you aren’t NPO, that there is no way that we will ever be able to conduct a RCT to detect a difference, nor would the difference be clinically significant, given these tiny numbers.

Bottom line: Emergency department patients don’t need to be strictly NPO before sedation.

Hear more: SGEM episode 165: I wanna be sedated – but do I need to be NPO?

---

I’m a little uncertain about this one

Simpkin AL, Schwartzstein RM. Tolerating Uncertainty – The Next Medical Revolution? The New England journal of medicine. 375(18):1713-1715. 2016. PMID: 27806221

This is a brilliant essay. I sort of just want to copy and paste it here. The authors point out that although we are all rationally aware of the uncertainty in medicine, the medical culture is deeply against acknowledging or embracing it. Our quixotic quest for the right answer conceals the fact that ideal clinical reasoning is iterative and evolutionary. They point out that two of the major problems that we face in modern medicine – the seemingly contradictory over-testing and premature closure – are two sides of the same coin. They both arise out of our discomfort with and inability to discuss uncertainty.

Bottom line: A quote: “Key elements for survival in the medical profession would seem, intuitively, to be a tolerance for uncertainty and a curiosity about the unknown.”

---

Antihistamines to prevent anaphylaxis? I don’t think so

Kawano T, Scheuermeyer FX, Gibo K. H1-antihistamines reduce progression to anaphylaxis among emergency department patients with allergic reactions. Academic emergency medicine. 2016. PMID: 27976492

The conclusions of this study state that “early H1a treatment in the ED or prehospital setting may decrease progression to anaphylaxis”. This would definitely be new information, but I don’t think it is actually what the data shows. This was a retrospective chart review (with excellent chart review methods) looking at 2376 patients coded as having allergic reactions at 2 Canadian emergency departments. They excluded patients with anaphylaxis. When comparing patients who received H1 antihistamines to those who did not, they found a higher progression to anaphylaxis in the patients who were not treated (3.4% versus 1.9%, NNT = 44.74, 95% CI 35.36 to 99.67). Of course, a NNT assumes a causal relationship, and I don’t think that is what is going on here. You have to wonder why some people were given antihistamines and other were not. Were the groups equal? They don’t seem to be, with differences in history of allergy, allergen, and mode of arrival. Most importantly, 7.8% of the antihistamine group (as compared to only 2.8% of the no treatment group) was treated with epinephrine “before the development of anaphylaxis”. I am not sure why these patients would be given epinephrine if the clinician didn’t think they had anaphylaxis, so I wonder whether the chart review methods just failed to identify the true condition of the patient. Either way, treating these patients with epinephrine presumably treats anaphylaxis symptoms, and the 5% difference in epinephrine use is larger than the subsequent 1.5% difference they base their conclusions on. If you assume the patients given epinephrine actually had anaphylaxis, the rate of anaphylaxis was 9.7% with antihistamines and 6.3% without, supporting the exact opposite conclusion to the authors’.

Bottom line: Antihistamines probably don’t prevent anaphylaxis, but this retrospective data can’t tell you either way

---

Late-night breastfeeding advice

Sutherland S. Late-night breastfeeding advice. Canadian family physician. 62(7):579. 2016. PMID: 27412213 [free full text]

I like this short essay. It describes the author’s struggle trying to breastfeed on her first night home with her first born child. The difficulties feeding lead to chaos in the house, anger, and feelings of inadequacy. After speaking to an on-call family doctor by phone, her husband draws her a warm bath, lights some candles, and leaves her a glass of sherry. The single formula feed that night made no difference to her long term desire or ability to breastfeed.

Bottom line: Not all medical advice needs to involve medicine. We treat people, not diseases. Sometimes, you just need a warm bath.

---

A better technique for nursemaid’s elbow?

Bexkens R, Washburn FJ, Eygendaal D, van den Bekerom MP, Oh LS. Effectiveness of reduction maneuvers in the treatment of nursemaid’s elbow: A systematic review and meta-analysis. The American journal of emergency medicine. 35(1):159-163. 2017. PMID: 27836316

This paper has already gather a lot of attention. Are you using the best technique to reduce pulled elbows? This is a systematic review and meta-analysis that includes 7 trials covering a total of 701 patients. None of the trials were high quality. According to their results, hyperpronation is better than the traditional technique of supination-flexion (risk ratio, 0.34; 95% CI, 0.23 to 0.49). The absolute risk difference between maneuvers was 26.4%, which results in a number needed to treat of 3.8. However, looking at those numbers, something seems really wrong with this data. To have an absolute difference of 26% you have to be failing AT LEAST 26% of the time using supination-flexion technique. I have been using this technique for years, and my success rate is very close to 100%. So this paper doesn’t pass the sniff test. I guess if you are just starting out or if you have had a number of failures, you could choose the hyperpronation technique, but if you are getting close to 100% success with supination-flexion like I am, this data should not make you change your practice. They do comment about pain, but honestly kids cry with both techniques and stop in seconds either way, so I’m not sure how they would accurately assess this.

Bottom line: Hyperpronation works for pulled elbows (but so does pronation-supination, or almost anything in my experience).

---

Tensor bandage only for 5th metatarsal avulsion fractures?

Akimau PI et al. Symptomatic treatment or cast immobilisation for avulsion fractures of the base of the fifth metatarsal: a prospective, randomised, single-blinded non-inferiority controlled trial. The bone & joint journal. 98-B(6):806-11. 2016. PMID: 27235524

This is a prospective, randomized, non-blinded trial comparing a below knee walking cast to tensor bandages (double layer!) for avulsion fractures of the fifth metatarsal (Lawrence and Botte type I). They enrolled 60 adult patients with injuries within 7 days. In both groups, patients were encouraged to bear weight as soon as they could tolerate. They used a previously validated functional outcome score, and found that that the tensor bandage was non-inferior to the cast. Unfortunately, there was a high rate of loss to follow up (40% at 3 months), which adds uncertainty to the results of this small trial.

Bottom line: Tensor bandage may be as good as walking cast for fifth metatarsal avulsion fractures, but this study is too small to promote wholesale practice changes.

---

Ummm.. OK

Osterberg EC, Gaither TW, Awad MA. Correlation between pubic hair grooming and STIs: results from a nationally representative probability sample. Sexually transmitted infections. 2016. PMID: 27920223

Just one of those bits of science that everyone needs to know: “extreme” pubic hair grooming is associated with increased risk of sexually transmitted infections. This is a survey of 7580 Americans asking about grooming habits, sexual behaviours, and STI history. The self-report nature of this data could produce some bias. “Extreme grooming” was defined as removal of all pubic hair more than 11 times per year, and was associated with higher rates of STI (OR 1.8 95% CI 1.4 – 2.2). Of course, (in contrast to the general media reports of this research) this is a classic example of association not equalling causation, as there are many potential confounders in this data. Also, although the difference is statistically significant, the differences may not be so clinically significant (14 vs 18% for any STI).

Bottom line: Keep this data in mind the next time a medical conference brings you to Las Vegas

---

Cheesy Joke of the Month

Yesterday I swallowed some food coloring. The doctor says I’m fine, but I feel like I’ve dyed a little inside