American ER Doc Gone Walkabout Episode 029
One of the strange things about SAEM: I spent a full day in an Evidence Based Medicine course. We spent much of our days analyzing studies in depth, tearing apart their choice of statistical measures and analyzing why a purportedly positive result was in fact an inconclusive or even negative result.
Other sessions were more traditional reviews: one speaker went through about 15 slides, each containing nothing but copies of the title and author lists of 2 or 3 articles per slide, concluding that we had lots of inconclusive, poorly done, studies with mutually incompatible results. Many bits and bytes gave their lives needlessly in support of insupportable conclusions.
Numerous sessions dealt with teaching techniques and methodology, including how not to give boring lectures (a topic not understood by all the lecturers). There was a subtext (and sometimes, a main text) through the topics covered, of attempts to rearrange the way ERs function to improve efficiency and “throughput” (physicians at triage or intake seemed to garner some attention). In one session where the virtues of high volume, high speed throughput of patients was discussed, a question was asked about the effects that the emphasis on efficiency and throughput might have on resident education (if the ultrasound has been ordered at intake, and perhaps already completed, why should a resident perform a bedside ultrasound that might take non-productive time? Or listen to the lungs if a chest x-ray will be ordered? Or examine the optic fundi if a head CT has already been done?). The discussants emphasized that in the new paradigm of Emergency Medicine, the residents would gain education and experience primarily in efficiency, organization, and use of metrics. No mention was made of becoming a physician nor a doctor. (I might have something to say about that in future LITFL posts. Last night I had a resident who was absolutely clueless about findings on lung auscultation, but did pay close attention to which of his patients had exceeded the length of stay guideline. I spent the rest of the shift on suicide watch.)
A session on cognition and decision making discussed the important (sometimes counterproductive) role of non-clinical context in our perceptions of clinical problems. Discussing “decision fatigue”, he summarized a study from the Israeli judicial system (http://www.pnas.org/content/108/17/6889) where petitions for parole were granted about 65% of the time when heard as the first case of the day, or the first case after a break. The percentage granted parole fell successively to near zero for the last case before a break, then the pattern was repeated.
I concluded that I shall take a tea break after every other patient.
More seriously, in the paradigm of ED management where efficiency and throughput are stressed (see above), our residents are considered to be wimps if they take a break – their job is to keep their heads down and turn the crank. The literature on decision fatigue suggests that this may have adverse effects.
Of course no Emergency Medicine meeting would be complete without multiple discussions of that topic which seems to have no (current) answer: pulmonary embolism.
Discussions ranged from incidental and peripheral: it is statistically inappropriate to use a continuous variable (d-dimer) as a dichotomous variable (i.e. simple positive if greater than 500 – we all “know” that a value of 2000 has more likelihood of being a true positive than a value of 501).
To the diagnostic issues of d-dimer: Specificities and therefore LR’s are dependent upon the comparison group and perhaps we should use different cutoff’s when our patient is in a risk group known to have high baseline D-dimers: active cancer, older age, pregnancy. And, if our pre-test (gestalt or calculated by formal score) probability is low, using the standard cutoff of 500, we will do too many CT studies to find one PE (by most people’s implicit criteria for Number Needed to Test) and here, too, a higher cutoff may be appropriate.
One discussant, who gets daily reports of his institution’s CT PE performance, reported that on the prior day, 38(!) CT PE studies at the two hospitals were all negative – but, in the US, when I ask residents how many negative CT PE studies they are willing to do to find one more PE, some answer 30 to 50, but a surprising number are willing to do 100 or 200. Wow! (I wonder what the answer would be if the question was: What would you accept as a testing threshold, if you were the patient and had a twinge of pleuritic lateral chest pain but a d-dimer of 501?)
On the other end of the diagnostic testing, we have a “one size fits all” problem of treatment for a wide variety of PE’s. While standard anticoagulation may be insufficient for the massive and even the submassive PE, thrombolysis may be excessively hazardous for even the submassive clot. (One discussant presented a picture of 2 large clots laid out on a surgical drape – which clots had been removed surgically from her own personal pulmonary arteries a few years before. Jeff Kline pointed out that despite his high level of interest in the topic of PE, he was unlikely to develop his own massive PE to further his depth of involvement in the topic.)
Discussions covered the fact that merely using mortality rates for PE might be too blunt an instrument (Kline pointed out that in patient surveys, more concern was raised over survival with severe disability/activity limitation than over non-survival). Thus, the import of using outcome measures such as pulmonary hypertension to evaluate treatment effects – when applied to those PEs that are not immediately life threatening.
And, potential approaches (half dose thrombolytics and interventional approaches) to minimizing potential adverse events shows promise in allowing safer treatment of the less than massive PE – note that there were no intracranial hemorrhages in the MOPETT trial. (I expected to see Cookie Monster as the principal investigator.)
I fantasized that eventually we would be able to categorize the treatment of massive (full dose aggressive thrombolytic or surgical), submassive to prevent pulmonary hypertension (half dose thrombolytic or interventional approaches – but don’t delay, that clot is cross-linking and organizing), small (not likely to cause long term adverse outcomes, so just prevent further clots with an anticoagulant), and super small (congratulations Mrs Jones, your lungs just filtered out a clot that would have been horrible if it had gotten into your brain). But, then I woke up from my reverie and went on to the next session.
While it was unsurprising that there was lots of coverage of PE, I was surprised that I saw not a single Ketamine session, abstract, nor poster. Although we can surmise that we’ve now grown accustomed to ketamine for sedation, agitation control, and analgesia, where’s the interest in ketamine as antidepressant?
For hospitals (common in the US, if not the world with functioning medical care systems) where arranging safe treatment for severely depressed and suicidal patients is a prolonged and challenging problem (multi day or longer stays in US ERs are not unusual), an antidepressant that is effective within hours, and durable for weeks, would be pretty useful. As best I know, there is only a single abstract presented at ACEP, four or so years ago, in the Emergency Medicine literature. The plastic surgical literature noted the effect (ketamine in the office while we do a little nip and tuck also makes the depression go away), as did the psychiatric literature (ketamine for electroconvulsive therapy fixes the depression even when the shock machine doesn’t work), followed by some bench research (not really sure what the mechanism is, maybe an effect on tau protein synthesis), followed by lots of psych interest in a treatment effect, even in the depressed person resistant to other therapies, but one infusion of ketamine works within hours, and lasts for days to weeks. Come on ER Docs lets get some research going on this! (If you’d like to see either the best or worst, not sure which, of American for profit medicine, take a look at: http://www.ketamineinfusioncenters.com ).
On a side note including both ketamine and American electron medical records: A depressed person left a suicide note and drove her car at high speed into a bridge abutment. She forgot that air bags work, and sustained only a pericardial tamponade which was drained and did not reaccumulate. She was a bit hypotensive while intubated and sedated with propofol, so was switched to ketamine sedation. She was extubated the following day, and evaluated by psychiatry. American EMRs generate many thousands of lines that are there only for billing (“I spent 38 seconds counselling the patient, and if they were present, the family” – you might guess that this is a “macro” that appears automatically – “if they were present” – gimme a break, can anyone who has completed nursery school believe that is a meaningful statement?), or legal purposes (“If I did a procedure, I washed my hands” – IMHO , that statement, when it appears automatically, merely makes us look stupid and naive, not “clean”), and only a tiny number of lines are actually clinically meaningful. Searching for clinically meaningful information is a challenge (generally, you aren’t even sure which of the many “notes” will contain useful information – consultant notes often state only that “I agree with” whatever the resident said, or may have “cloned” the resident note). In that kind of GIGO world, the psychiatry consultants had no chance of finding the recorded information that a suicide note did indeed exist – and when they interviewed the patient, she had been “cured” of depression (perhaps by the ketamine for sedation), and psychiatry concluded that there was no depression, nor suicidality – must have been an accident.
Unsurprisingly, the session on post-arrest resuscitation care had lots of interest. In the question and answer afterwards, someone reinforced for me the adage: “Perfect is the enemy of good (enough).” A questioner pointed out that the TTM trial was one of “equivalence” not of “non-inferiority.” Therefore, did not answer the question – I’m actually not sure which question. So, now we need to go back, develop a validated rationale for a new study, publish the protocol, then develop the study itself, then do a “proper” trial and publish it, before we can draw “any” conclusions as to proper depth of post-arrest hypothermia. By 2050, we should know whether our post-arrest temperature goal is 33, 33.5, or some other number, or whether non-febrile is adequate.
OK, mates, I’ve delved a little into the previously, to me, unknown world of academia. I understand that research and statistics and education is critical. But, it is a bit deeper than a dive for which I can hold my breath long enough, and now the hypoxia is making my head hurt. Time to come up for a breath, go for a bike ride, and go back to some seat of the pants medicine (I was pleased to note that in February, even the NEJM applied science to the bedside exam and published an article on auscultation of the lungs. Yes, the bedside, use your ears, auscultation of the lungs! If you missed it, you should go back and read it. It’s great.)
Later mates, if I survive my bike ride this time.
Next time: Let’s string together a few rules of thumb
American ER doc