Try to Avoid tPA When Already Bleeding

Coming to us from the Department of Common Sense: don’t give tPA to stroke patients who already have intracranial hemorrhage.  There’s a little more subtlety here, of course, because in this instance, we’re dealing with cerebral microbleeds – tiny foci of angiographic damage visualized only on MRI.

These authors performed a pooled and individual-patient meta-analysis of those undergoing MRI prior to treatment with intravenous thrombolysis.  When stratified by CMB burden, arbitrarily divided into “none”, “1-10”, and “>10”, the obvious is … obvious: patients who are already bleeding are more likely to continue bleeding.  In the unadjusted raw numbers, patients with no CMB had a symptomatic intracranial hemorrhage rate of 4.3%, those with 1-10 CMB had 6.1%, and those with >10 had 40.0%.

There are many technical limitations inherent to the retrospective nature of their study, as well as likely other confounding variables – but, the basic gist: our current practice relying only on non-contrast CT likely misses an important safety indicator in the setting of tPA use.

“Risk of Symptomatic Intracerebral Hemorrhage After Intravenous Thrombolysis in Patients With Acute Ischemic Stroke and High Cerebral Microbleed Burden”
https://www.ncbi.nlm.nih.gov/pubmed/27088650

Trauma is Still Trauma the Next Day

Acute closed head trauma is easy enough – and challenging enough.  There are validated decision instruments and guidelines, yet still plenty of CTs performed absent sound indications.  However, the question this study addresses is slightly different: what to do with those who present in a delayed fashion following minor head trauma?

The authors probably sum it up best in a reasonably concise fashion:

“Patients presenting after 24h of injury are a potentially distinct subpopulation. They could be at lower risk, as there is evidence that patients with mild/minor head injury who have injuries requiring neurosurgery will deteriorate within 24h. Alternatively, they could be a self-selecting higher-risk group attending due to the worsening or persistence of symptoms.”

These authors reviewed 2,240 patient encounters resulting in a CT scan of the head, with a goal of winnowing it down to just those performed for a traumatic indication.  Of those, 549 were performed within 24 hours of injury and 101 were delayed presentations.  There were 46 (8.4%) CTs positive for traumatic injury in the acute presentations and 10 (9.9%) in delayed, while 5 and 3 patients each underwent neurosurgical intervention, respectively.  So, the answer to their research question, at least in pragmatic terms, may be that the two forces balance each other out.

These authors also present “sensitivity” statistics regarding the utility of guidelines at predicting the presence of an important TBI, and quote a sensitivity of 70% based on chart review.  The denominator for sensitivity would more appropriately the entire population of presentations for trauma, not simply those who underwent CT scanning.  It is also probably more likely, given these patients had important TBI on CT, there may have been undocumented, guideline-compliant, indications not abstracted by chart review.

While our decision instruments for closed head injury were derived in typically an acute population, I would not yet draw any conclusions refuting their generalizability to delayed presentations.

“CT head imaging in patients with head injury who present after 24 h of injury: a retrospective cohort study”
https://www.ncbi.nlm.nih.gov/pubmed/27076439

The Unmagical Checklist

The checklist has reached ascendant status in medicine.  As introduced into the mainstream by Atul Gawande, they have begun to permeate every nook of healthcare delivery.  However, evidence of benefit when applied to one particular problem in one particular setting is no guarantee of universal utility.

These authors performed a study in Brazilian intensive care units, using a cluster-randomized pre/post design to evaluate the effect of a quality improvement effort built around a checklist.  Each element on the checklist represented a consensus or evidence-based practice associated with improvement in surrogate markers for patient outcomes.  The combined intervention was hoped to improve overall in-hospital mortality for ICU patients at the intervention hospitals.

It didn’t – mortality increased similarly for both intervention and control ICUs.

In fact, for all secondary clinical outcomes – catheter-related infections, ventilator-associated pneumonia, urinary tract infections, ICU days, etc. – there were no significant improvements over the baseline period, and no difference compared with controls.  There were small improvements in processes of care, such as VTE prophylaxis, catheter use, and appropriate tidal volumes during ventilation – but without corresponding clinical outcome improvement.

Interestingly, clinicians working in the intervention ICUs typically felt as though their ICUs were safer.  They were more likely, sometimes significantly so, to provide answers reflecting positive associations regarding their working conditions and safety climate.  Indeed, the intervention was perceived as so likely to be beneficial even prior to the start of the study that a short duration was mandated for the trial so all ICUs could eventually start using the checklist.

These authors have several justifications for why their checklist did not function appropriately, focusing on various details regarding the trial.  I think the simplest expression regarding the effectiveness of a checklist relates to the magnitude of effect and the baseline frequency of adherence.  Unless a significant magnitude of effect is seen by improving compliance with an intervention, and the intervention itself is infrequently performed, returns will diminish dramatically.  A checklist such as this, with multiple low-yield elements, is unlikely to return substantial patient-oriented outcome improvements.  Indeed, the resources devoted to checklist rounding and adherence may even dilute the focus on important clinical considerations.

“Effect of a Quality Improvement Intervention With Daily Round Checklists, Goal Setting, and Clinician Prompting on Mortality of Critically Ill Patients”
https://www.ncbi.nlm.nih.gov/pubmed/25928627

Putting Children to the Flame

Many readers here are students, trainees, or otherwise academic-affiliated, and have limited exposure to the world of community practice.  In these settings, frequently, our pediatric exposure is supervised by clinician-educator sub-specialists in Pediatric Emergency Medicine.  We see the very best evidence translated into acute care of children in the Emergency Department.

The real world is a little different.

These two articles describe the shortcomings of advanced imaging practice in community pediatric settings – in the diagnosis of appendicitis, and in the evaluation of closed head injury.

In the appendicitis article, the authors compare two settings both staffed by PEM physicians – an academic medical center with in-house pediatric surgical coverage, and a community center with consultation available only by phone.  Each site had similar rates of appendicitis diagnoses – 4.7% vs. 4.0% at the academic and community site, respectively.  The academic site, however, evaluated fewer patients with abdominal pain with blood work, and then fewer still of those went on to advanced imaging.  Then, of those receiving advanced imaging, the rates were 10.8% CT at the academic center vs. 28.1% CT at the community center.  Ultrasound however, was employed in 16.6% of cases at the academic center versus 6.5% at the community practice.  Nearly all this difference, however, seemed to be made up of patients admitted to the hospital without any operative intervention.  The obvious reality, then:  radiation in lieu of observation.

The second article here describes the neuroimaging (CT or MRI) of patients evaluated following trauma, along with their ultimate disposition.  Of 2,679 patients reviewed, there were 94 patients with important non-surgical, trauma-related diagnoses, and an additional 16 patients who required neurosurgical intervention.  These authors, however, based on GCS estimates recorded and the distribution of outcomes in the PECARN study, estimate the prevalence of entry criteria into appropriate scanning would have obviated >2000 of these scans.  While I believe they are probably mis-applying the evidence and overstating the inappropriateness of CT, the rarity of serious diagnoses suggests at least a majority of these CTs probably could have been avoided.

In short, we’re still doing too many CTs on children.  Some of the contributing issues are systems based, and some are related to practice re-education.  More ultrasound and more observation, please – and less nuking of children.

“Imaging for Suspected Appendicitis: Variation Between Academic and Private Practice Models”
https://www.ncbi.nlm.nih.gov/pubmed/27050738

“Neuroimaging Rates for Closed Head Trauma in a Community Hospital”

Hello, Have You Heard of NEXUS/CCR?

In the same vein as my previous post inappropriate imaging despite the presence of PERC, this next article takes our evaluation of minor trauma to task.

This brief retrospective series looked only at presentations to the Emergency Department following “ground level fall” leading to a CT of the cervical spine.  These authors defined a “ground level fall” as fall of fewer than 3 feet or 5 stairs.  These authors then reviewed the documentation associated with each case for criteria specifically excluding the case from NEXUS or CCR and appropriate for CT imaging.

Of the 760 patients with ground-level falls included in this study, there were 7 cervical spine fractures – 6 stable, and 1 unstable.  All patients with a cervical spine fracture had documented criteria supporting appropriate CT imaging.  However, based on their retrospective review, 22.0% and 20.7% of encounters specifically documented criteria meeting NEXUS and CCR, and should not have led to CT imaging.  An additional 9.3% and 29.9% of patients had insufficient documentation of NEXUS or CCR criteria needed to determine appropriateness.

These authors posit there may be substantial cost savings to the healthcare system if these decision instruments are appropriately applied.  I tend to agree – although, there are obvious limitations to this sort of retrospective review.  It does, at least, back up my own anecdotal experience witnessing clinically questionable use of advanced imaging in minor trauma.

“Utility of computed tomography imaging of the cervical spine in trauma evaluation of ground level fall”
https://www.ncbi.nlm.nih.gov/pubmed/27032009

Amiodarone, Lidocaine, or … Nothing

The prehospital game has always been muddy, particularly when it comes to the various pharmacologic interventions fixed in the constellation of Advanced Life Support.  You could – and some have – go as far as to say virtually nothing in the armamentarium of prehospital care has been proven to improve meaningful survival following cardiac arrest.

This latest evidence drop is the ALPS trial from the Resuscitation Outcomes Consortium – amiodarone, lidocaine, or placebo in patients with shock-refractory ventricular fibrillation/ventricular tachycardia.  The underlying physiologic theory would be that there’s a period of time between onset of VF/VT and death where drugs are the answer.  Early on, an appropriately delivered shock solves the disorganized electrical problem.  Later on, after the intracellular rigor sets in, nothing will be reliably beneficial.  In some intervening period, there is hope that amiodarone or lidocaine provides a little extra stabilization to help a shock take hold.

These authors performed a prospective, double-blinded, placebo-controlled, randomized trial with three arms to compare each antiarrhythmic therapy to placebo.  With a 90% power to detect a 6% survival to hospital discharge between arms, these authors enrolled 3,026 patients.  Groups were generally well-balanced for important prognostic features, adjunctive treatments, and post-admission care.  But, unfortunately, the survival advantage seen was 3.2% for amiodarone and 2.6% for lidocaine – meaning the confidence intervals each cross unity and the p-values are 0.08 and 0.16, respectively.

So, we have, yet again, a study that provides more to argue about than to inform practice.  Delving into the secondary outcomes and the supplementary appendix, it is clear that both antiarrhythmic drugs are doing something.  Patients receiving the active arms in the trial required fewer shocks, received fewer doses of alternative antiarrhythmic drugs (e.g., magnesium, procainamide), and had significantly higher rates of hospital admission – 45.7% vs. 47.0% as. 39.7% for amiodarone, lidocaine, and placebo, respectively.  However, all these advantages on the front-end decayed into smaller and smaller absolute differences on the back end – where mRS 3 or better discharge status was only 18.8% vs. 17.5% vs. 16.6%.

The glass half-full look at this is: even though it’s not statistically significant, even a couple percentage points of life vs. death represents a couple thousand additional neurologically intact survivors each year.  The glass half-empty look at this is: 26.9% of the amiodarone group required immense resource outlay and ICU care and was still ultimately dead or disabled, along with 29.5% of the lidocaine group, but only 23.1% of the placebo group.

These data suggest many reasonable choices may be made.  The signal may not be strong enough, and the downstream costs high enough, that a case could be made to dramatically curtail use of both active drugs.  Or, specific instances of use or delivery improvements could be proposed, based on other survival signals hidden in the secondary data.  Finally, costs, ease of use, and other secondary signals could make dueling cases to discontinue use of one of the two active drugs.

I think these drugs probably have value – but, their value won’t be maximized until in-hospital care produces either a better yield of neurologically intact survivors or better prognosticates resuscitation to reduce ultimate resource utilization.

“Amiodarone, Lidocaine, or Placebo in Out-of-Hospital Cardiac Arrest”
https://www.ncbi.nlm.nih.gov/pubmed/27043165

Where Acute Otitis Media is Born

Is it 3 AM in your Emergency Department?  Is there a febrile infant with their still-awake parents straggling in the door?  Do you hear the first few bars of the “it’s just a virus” song start playing over Spotify?

This little study prospectively enrolled healthy infants at birth and followed them to their first episode of acute otitis media or 12 months of age.  They were followed specifically to determine predictive clinical and epidemiological factors influencing the first diagnosis of AOM.  Additionally, as they aged and during illness, nasopharyngeal swabs were taken to evaluate viral and bacterial flora.

Based on a sample of 367 infants followed for a total of 286 child-years, there were 887 presentations for viral URIs and 180 presentations for AOM –and all but two of AOMs were preceded by a URI.  The median time from URI presentation to AOM diagnosis was 3 days.  These authors also present a fair bit of microbiologic data regarding specific risks for URI and AOM, although these are not specifically modifiable and of lesser clinical relevance.  From a modifiable environmental outlook, however, there are a few interesting tidbits tying into what we already suspected to be true:  breastfeeding is good, the new PC13 vaccine is good, and daycare is a cesspool.

Overall, this would tend to support our typical advice to parents to have their children present for a recheck 48-72 hours following Emergency Department visit, particularly if there has not been clinical improvement or in the context of apparent clinical re-worsening.

“Acute Otitis Media and Other Complications of Viral Respiratory Infection”
https://www.ncbi.nlm.nih.gov/pubmed/27020793

The EndOfLitOfNote [April Fool’s]

Over the last five years, Emergency Medicine Literature of Note has grown from quite inauspicious origins to something more – with a readership greater than just my mom.

Of course, as is constant in nature, all things must end.

Today will be the final post of EMLitOfNote in its current form.  While my academic and other non-profit affiliations have either encouraged or tolerated the blog, my newest endeavor means my public-record skeptical combativeness must cease.

In partnership with Genentech, Covidien, and the American Stroke Association, I will be joining a new project – the SMart StrOKE IniTiative (SMOKE-IT).  Rather than focusing on the flaws and conflict-of-interest in the new stroke literature, I’ve been asked to consult on pre-publication messaging to prevent such naysayers as myself from nitpicking the inconsequential details.

Even though EMLitOfNote provides commentary on all disciplines of work, the legal team associated with this new partnership feels this site is best discontinued.  While it seems, frankly, unfair, this new project provides me with generous enough compensation that I am somehow able to cope.

You can read the press release here: http://bit.ly/1acPEXp

The Biomarker for Burnout

I’m tired.  You’re tired.  We’re all tired.  Importantly – performance suffers with exhaustion, unhealthy behaviors at work increase, and cognitive errors at work rise.  Burnout.

And now there might be a test for it.

This is a small study of resident trainees in Turkey, correlating the levels of neurotrophic factor S100 calcium binding protein B with symptoms of Burnout Syndrome – emotional exhaustion, depersonalization, and personal accomplishment.  S100B is a marker of glial activation and brain injury, and seems to fluctuate with stress and depression, although the associations have not been shown to be reliable.

Each resident trainee was asked to complete a questionnaire regarding burnout prior to, and following, a night shift, along with concomitant blood draw.  Unfortunately, the results are primarily grim, and not on account of the primary outcome: 37 of 48 participating residents scored in the severe depression category on the burnout questionnaire.  The remaining 11 scored in the moderate range.

Looking at the actual purpose of the study, however, they did find S100B levels were significantly different between severe and moderate depression, even accounting for the small sample.  The pre- and post-night shift levels were not appreciably different.  Overall, S100B seemed to correlate best with the overall burnout score, in particular the subscore for emotional exhaustion.

It’s a little hard to interpret these data, or envision how they might be applied in a real-world situation.  It does seem a reasonable biomarker to pursue as an objective measure of the stresses of training, and, frankly, it may be the on-shift changes were not detected as a result of most residents already exhibiting features of high stress and burnout even before starting their night.  Then, even assuming S100B were proven valid, the “gold standard” in this case – the burnout inventory – is probably less expensive and certainly less invasive to deploy.

I am not certain the way forward for this line of burnout biomarker research, but it is rather interesting.

“Serum S100B as a surrogate biomarker in the diagnoses of burnout and depression in emergency medicine residents.”
https://www.ncbi.nlm.nih.gov/pubmed/27018399

Hospitalization or Home After TIA

In the pursuit of “value-based care”, innovators are consistently looking for ways to deliver similar outcomes without the risks and resource utilization of inpatient hospitalization.  One of these realms is the evaluation of transient ischemic attack.  Most of the recommended follow-up tests are only relatively urgent in nature, and with medical management the typical mainstay of therapy.  As serious considerations go, it seems ripe a candidate for outpatient management.

This retrospective look at outcomes from Canada, however, suggests there may be pitfalls to such a strategy.  These authors reviewed the outcomes of 8,540 patients presenting with TIA or minor stroke, and compared those either admitted to the hospital at the index visit with those discharged, and among those discharged, referral to a specialized follow-up clinic or not.

Patients admitted to the hospital, by all measures, had more severe cerebrovascular disease – as evidenced by duration of symptoms, ABCD2 scores, diagnosis of minor stroke, and other comorbidities.  However, despite this, following hospitalization, these patients had the lowest risk or recurrent stroke or TIA within one year.  The benefit, presumably, comes from increased likelihood of undergoing risk stratification and treatment – carotid imaging, echocardiography, appropriate anticoagulation, appropriate antithrombotic therapy, and the like.  Then, among the discharged, various adjusted and propensity matched analysis demonstrated a protective effect of referral to specialty outpatient follow-up against death, but not for stroke or TIA.  These data do not have the granularity to fully describe whether the excess deaths were in some fashion related to cerebrovascular disease.

Most of the absolute differences in outcomes between groups are small – a few percentage points each, and smaller after adjustment.  That said, it’s probably clearly superior care, as configured in Ontario during this time frame, to have been admitted to the hospital.  As TIA evaluation, and other similar conditions, move to outpatient pathways rather than traditional hospitalization, this represents an important reminder of potential risks of degradation in thoroughness and quality.

“Association between hospitalization and care after transient ischemic attack or minor stroke”
https://www.ncbi.nlm.nih.gov/pubmed/27016521