Articles of the Month (March 2017)

I decided to take a break last month, for the first time in a few years. I imagine some people were hoping I would take I much more extended vacation, but I am sorry to say, you are stuck with me. Here is another run down of medical research. I might be a little rusty, though, after a month off, so cut me some slack…

As always, the audio version is available on the BroomeDocs podcast (if you can tolerate very poor quality banter that Casey and I produce).

Most misunderstood and misinterpreted paper of the month

Barnett ML, Olenski AR, Jena AB. Opioid-Prescribing Patterns of Emergency Physicians and Risk of Long-Term Use. The New England journal of medicine. 376(7):663-673. 2017. PMID: 28199807 [free full text]

This paper has been discussed all over the place, including in the lay press, and I really don’t think that the conclusions being drawn are supported by the data here. This is a huge database study looking at 377,629 patients in the American Medicare system. The basic idea of the study was to compare patients seen in the emergency department by physicians with the highest and lowest opioid prescription rates, and then see if that influenced long term opioid use (180 days of use or more in the next year). According to their conclusions, this is exactly what they found and it is what has widely been reported. Even if this was what their data showed, I would be pretty skeptical of the relevance of this finding. The actual number of long term opioid users was 1.5% vs 1.1%, which is obviously statistically significant because of the massive numbers in the study, but is a very small difference that is at most marginally clinically significant. (Everything in this trial had a statistically significant difference, including the mean ages of 68.8 and 68.2 between the groups, which tells you something about relying on p values in a data dredge this large). Remember, that 1.5% number is just long term use. The abuse or harm rate will only be a fraction of that. (Opioid poisoning was seen in 7/10,000 and 10/10,000 in the two groups). However, if that was a real difference, it might be worth discussing, because we clearly have an opioid problem that needs to be addressed. What nobody seems to be talking about is that they didn’t actually compare high prescribing doctors to low prescribing doctors at all. They defined a doctor as high prescribing if in the week after a patient saw that doctor, the patient filled a prescription for an opioid. The prescription didn’t have to be from that specific doctor. There was no link at all between the doctor and the prescription in this study. Patients may have visited 10 other doctors looking for a prescription. We have no way of knowing. So what they actually studied was: are patients who are more likely to fill a prescription for an opioid in the first week after an emergency department visit also more likely to continue using opioids over the next year. When phrased that way, I think we all knew the answer before the research was started. Given the miniscule differences between the two groups, the massive confounders easily dwarf any reported differences between the two groups.

Bottom line: Opioids are a problem. Prescribe responsibly. Don’t use combination pills. Avoid euphorics like percocet and stick to plain oral morphine. Prescribe only short courses for acute pain. But ignore this paper – it doesn’t tell you anything.

---

Why we need resuscitation sequenced Intubation

Perbet S, De Jong A, Delmas J. Incidence of and risk factors for severe cardiovascular collapse after endotracheal intubation in the ICU: a multicenter observational study. Critical care. 19:257. 2015. PMID: 26084896 [free full text]

This is a secondary analysis of a prospective observational trial originally designed to look at predictors of difficult airways. In this analysis, they looked at the patients who experienced peri-intubation cardiovascular collapse, defined as a single systolic pressure less than 65 mmHg, or a systolic pressure of less than 90 for more than 30 minutes despite fluid resuscitation, or a requirement for vasopressors. They excluded anyone with pre-existing hypotension. Of the total of 1400 intubations, 885 occurred in patients without pre-existing hypotension, and cardiovascular collapse occurred in 264 (29.8%). Although not causal based on this observational data, mortality was significantly higher in the group of patients that experience peri-intubation hypotension (30% vs 20%, p=0.001). Not surprisingly, sicker, older patients were more likely to have problems. Patients being intubated for coma alone were less likely to experience collapse. Propofol was actually the agent least likely to be associated with collapse, but that is almost certainly the result of it being selected against in sicker patients by the treating physicians. Peri-intubation hypotension is a problem we all know about, but I think it gets less attention than the airway anatomy when we are discussing airway issues. An over-emphasis on the ABCs has convinced many that the airway must be secured early in most critically ill patients, but really, immediate intervention is rarely required. I include this paper mostly as a reminder of the importance of “resuscitation sequenced intubation”, or in other words, ensuring that your patients are hemodynamically suitable for intubation.

Bottom line: In critically ill patients, post-intubation cardiovascular collapse is common. We need to resuscitate these patients prior to intubation, and be prepare for deterioration.

---

Intranasal Ketorolac?

Rao AS et al. A Randomized Trial of Ketorolac vs. Sumatripan vs. Placebo Nasal Spray (KSPN) for Acute Migraine. Headache. 56(2):331-40. 2016. PMID: 26840902 [free full text]

I thought this was a neat little study, even if it won’t change my management in the ED. It is a RCT of migraine patients, randomized to either ketorolac, sumatriptan, or placebo, all given intranasally. Both medications looked reasonably good, with 70% of patients being free from pain at 2 hours in the treatment groups as compared to only 40% with the placebo group (p<0.001). The two medications were non-inferior to each other. A few problems with the trial: it was industry sponsored and takes place in a migraine clinic rather than an emergency department. However, my big complaint is that they didn’t compare the nasal route to an oral dose. Is the intranasal ketorolac here actually any better than an oral NSAID, or are they just trying to sell us a fancy, more expensive, recently patented version?

Bottom line: I don’t use any of these options in the emergency department for migraine right now. This paper won’t change my practice.

Avcu N, Doğan NÖ, Pekdemir M. Intranasal Lidocaine in Acute Treatment of Migraine: A Randomized Controlled Trial. Annals of emergency medicine. 2016. PMID: 27889366

Just to throw in another paper on intranasal medications in migraine, this is a double-blind RCT comparing intranasal lidocaine to placebo in 162 emergency department migraine patients. I guess it would be a neat trick to have if it worked, but it didn’t. There were no differences between the two groups, except that the lidocaine group had more nasal pain or local irritation after the spray.

Bottom line: Just in case you were wondering, intranasal lidocaine is unlikely to help your migraine patients.

---

Forget the sepsis bundles, these patients just need vitamins

Marik PE, Khangoora V, Rivera R, Hooper MH, Catravas J. Hydrocortisone, Vitamin C and Thiamine for the Treatment of Severe Sepsis and Septic Shock: A Retrospective Before-After Study. Chest. 2016. PMID: 27940189

I’ve seen this paper everywhere in the lay press. It’s a paper by Paul Marik, one of the biggest names in sepsis research, but it isn’t a paper you should spend much time on. Essentially, Marik had a few patients he was sure were going to die from septic shock, and decided to give them a combination of intravenous vitamin C, hydrocortisone, and thiamine. They seemed to have a miraculous recovery. Therefore, they started using this treatment routinely in their ICU. This paper is a retrospective before/after description of their experience. They looked at a total of 94 patients in the 7 months before and after the change. Before the change mortality was 40%. With the new treatment, the mortality was only 9%. They add that, in their opinion, none of the patients in the treatment group died of sepsis, but rather (in the unblinded eyes of the authors) died of alternative causes, such as dementia or heart failure. A 30% absolute decrease in mortality is an incredible change.
You might even call it unbelievable. I won’t list the multitude of problems that can result in significant bias in this kind of unblinded, retrospective, before and after trial. Clearly, this is not a treatment strategy anyone should be instituting clinically based on this data. However, the massive differences reported here are interesting and may warrant further study. (Marik has already registered a massive, multicenter randomized trial, as far as I know.)

Bottom line: Definitely not a game changer, but I think it is interesting, and you will hear about it, whether in a medical venue or on NPR.

---

A couple papers on rudeness

Riskin A, Erez A, Foulk TA. The Impact of Rudeness on Medical Team Performance: A Randomized Trial. Pediatrics. 136(3):487-95. 2015. PMID: 26260718 [free full text]

This is a randomized, double blind, simulation based trial. They studied 72 NICU teams (made up of one doctor and 2 nurses) in a neonatal resuscitation simulation. All the groups were told that they were going to be observed by a visiting international expert. For half of the groups, this expert (in a video) stated that he had already observed a number of teams and he was “not very impressed with the quality of medicine in Israel.” Halfway through the resuscitation, during a break, the expert stated that this team “wouldn’t last a week” in his department. The simulations were observed by experts who were blinded to this rudeness. The teams were rated in a variety of diagnostic and procedural categories. The teams exposed to rudeness performed less well, basically across the board. My biggest problem is that although the results are all statistically significant, it’s not clear what a clinically significant difference on this score is. (For example, for the overall diagnostic performance, the scores were 3.18 and 2.65 out of 5. Are those scores really different? I don’t know.) The results seem plausible, and I certainly believe we should avoid rudeness in our professional practice, but I am not sure this study shows an important difference that would translate into clinical care.

Riskin A, Erez A, Foulk TA. Rudeness and Medical Team Performance. Pediatrics. 139(2):. 2017. PMID: 28073958

This is a second study by the same group. The methods are essentially the same, except this time the rudeness is from the mother of the child, in the form of a statement that “I knew we should have gone to a better hospital where they don’t practice Third World medicine.” They also added in a computer based cognitive behavioural modification aimed at altering interpretations of threatening expressions from others and promoting a more benign interpretation of ambiguous environmental stimuli. The results are essentially the same, with the group exposed to rudeness performing less well. However, the cognitive behavioural modification technique appeared to neutralize the effect of rudeness.

Overall bottom line: Our performance is influenced by a large number of psychologic factors. It isn’t surprising that rudeness might impact our performance. Our performance is not directly influenced by outside stimuli – it is our interpretation of those stimuli that matters. That is good news, because you can train yourself to alter your interpretations and therefore maintain the excellent performance you are accustom to. (You can read more about this in my recent post Performance Under Pressure.)

---

The easy IJ

Moayedi S, Witting M, Pirotte M. Safety and Efficacy of the “Easy Internal Jugular (IJ)”: An Approach to Difficult Intravenous Access. The Journal of emergency medicine. 51(6):636-642. 2016. PMID: 27658558

This is a prospective look at the use of an “easy IJ” in 74 patients with difficult vascular
access in three emergency departments. The easy IJ is a technique of placing a peripheral IV (in this case they used 4.8 cm long 18 gauge catheters) into the internal jugular using ultrasound guidance and limited sterile technique (not a full drape). Overall, this was relatively successful, with 88% of patients getting a usable line. It was quick, with total procedure time clocking in under 5 minutes (although it takes that long for our ultrasound to boot up sometimes). It also looked pretty safe in this small cohort, with no cases of pneumothorax, line infection, or arterial puncture. The lines were used for up to 24 hours before being removed. There are a number of other vascular access techniques (IOs and ultrasound guided peripheral IVs) that I would go to first, but it is important to have multiple backup strategies. Although we would need a much larger trial to really assess the safety of this procedure, and I would love to see a controlled trial comparing this to other rescue vascular access options, this is clearly a reasonable option when urgent vascular access is required.

Bottom line: I will definitely keep this procedure in mind in patients with difficult vascular access

---

Treatment for first time seizures?

Leone MA et al. Immediate antiepileptic drug treatment, versus placebo, deferred, or no treatment for first unprovoked seizure. The Cochrane database of systematic reviews. 2016. PMID: 27150433

Should we be starting anticonvulsants for first time unprovoked seizures in the emergency department? This is a cochrane review looking at immediate versus delayed treatment for patients of any age after their first unprovoked seizure. They identified a total of 6 trials that included 1600 patients. Overall, starting anticonvulsants early was associated with a lower risk of early seizure recurrence. You were less likely to have a seizure in the first 5 years if you were treated, but there was no difference in the likelihood of obtaining any 5 year seizure free period. Obviously, adverse events were higher in the treatment group. At 5 years, 45% of the control group had a seizure, as compared to only 35% of the treatment group. However, this has to be weighed against the 30% of patients in the treatment group who had adverse effects (not very well described here) from these medications.

Bottom line: This is all about shared decision making to me. There isn’t a clear answer, so I will probably still leave this decision to the neurologists in follow up.

---

They might bend, but they won’t break

Jiang N, Cao ZH, Ma YF, Lin Z, Yu B. Management of Pediatric Forearm Torus Fractures: A Systematic Review and Meta-Analysis. Pediatric emergency care. 32(11):773-778. 2016. PMID: 26555307

This systematic review and meta-analysis looked at the management of pediatric forearm buckle fractures. They only included RCTs, of which there were 8, covering 781 patients. The results aren’t a huge surprise if you have been following this literature or reading the articles of the month. When compared to plaster, non-rigid removable splints or bandages had better functional recovery, a lower complication rate, and were preferred by patients.

Bottom line: There is no reason to be casting torus fractures. We should probably stock velcro splints in the ED instead.

---

Could facebook extend your life?

Hobbs WR, Burke M, Christakis NA, Fowler JH. Online social integration is associated with reduced mortality risk. Proceedings of the National Academy of Sciences of the United States of America. 113(46):12980-12984. 2016. PMID: 27799553 [free full text]

This is a fun little study. They looked at the activity of 12 million facebook accounts in California and compared that to state health records. Having more friends on facebook was associated with decreased risk of mortality. However, you can’t just have friends – they have to request you. If you send out more friend requests, you are actually more likely to die. Being tagged in photos means you will live longer whereas posting more status updates means you will die sooner. (They guess that this might correlate with real life social interactions: being tagged in a photo means you were probably with someone else, whereas constantly adding more status updates means you are probably are currently alone, and might stay that way forever.) So, get more friends on facebook, but make sure they add you and not vice versa, and go tag yourself in some photos. (Do I really need to comment on association not being causation?)

Bottom line: I wish I had more friends

---

Cheesy Joke of the Month

I recently decided that I want to be cremated…

It’s my only chance to have a smoking hot body.