What Have We Learned About Mobile Stroke Units?

To date, not much.  We’ve learned it’s physically possible to put a CT scanner in van.  It’s technically feasible to infuse an expensive medication into a vein.  A small amount of time is saved.

The presumption is, the few minutes saved, cumulatively, will outweigh any harms associated with the lack of evaluation in a comprehensive center by a fully-trained Emergency Physician.  And, this study – BEST-MSU – aims to measure this, comparing patient outcomes in a week-on/week-off fashion.

This is their initial report, covering a 10-week run-in phase.  During this time, the MSU was in service for 57 of those days – and there were 130 activations, or, approximately 2.7 per day.  Those activations resulted in 24 patients who were potentially eligible for the study.  Of those, 12 were treated with tPA pre-hospital.  And one of those had a non-stroke final diagnosis.

So, 130 activations for “appropriate” 11 administrations of tPA.

As the authors correctly report, very little can be concluded regarding the effectiveness of the therapy.  What does seem to be clear – this is a substantial resource expenditure associated with repeated deployments for the smallest handful of treatment eligible individuals.

“Benefits of Stroke Treatment Using a Mobile Stroke Unit Compared With Standard Management – The BEST-MSU Study Run-In Phase”

How Many LPs Does It Take to Find SAH?

In this observational series: 204.

This study describes the findings of 2,248 patients with “headache, suggestive of subarachnoid hemorrhage” who underwent computed tomography, followed by lumbar puncture.  CT is quite sensitive for aneurysmal hemorrhage in the few hours following onset, but fades with time.  Spectrophotometric analysis of CSF is the gold standard for diagnosis in the United Kingdom, and was used by these facilities to describe positive, negative, and indeterminate results on LP.

Of these 2,248 patients, LP were negative in 1,507, uninterpretable in 350, indeterminate in 299, and positive in 92.  All patients with positive LP had follow-up CTA or MRA, and there were 9 vascular abnormalities.  Then, a subset of the uninterpretable and indeterminate patients underwent CTA or MRA, as well, resulting in an additional two vascular abnormalities.

Thus, 204.

There are a few surprising bits regarding these data.  The ~4% rate of “SAH” detected in this study is lower than most other observational series.  Some of these, however, were neurosurgical referral centers, with much higher rates.  Then, of course, the incidence will further depend on various definitions of positivity and the rate at which patients with lower pre-test likelihood are evaluated for SAH.  Additionally, the rate of vascular abnormality in those with “SAH” was also quite low.  In such a small sample, the confidence intervals are quite wide, so it’s hard to estimate the generalizability of these findings.

Regardless of the precise numbers, their overall conclusion is reasonable: CT + LP is extremely low yield for true pathology.  There is a definite need for decision instruments with improved specificity to prevent such extensive resource utilization.

“An Observational Study of 2,248 Patients Presenting With Headache, Suggestive of Subarachnoid Hemorrhage, Who Received Lumbar Punctures Following Normal Computed Tomography of the Head”

Don’t Hospitalize Warfarin & Minor TBI

There have been a few retrospective, observational studies evaluating the outcomes of anticoagulated patients with minor head injury.  The incidence reported by such series ranges from 6% in observation with mandatory repeat CT, to 0.6% with discretionary CT.

This series from Singapore reports: 0.3%.
In this series, all patients taking warfarin and having minor head trauma underwent protocolized initial CT and hospitalization for observation.  Repeat CT was performed, however, only at the discretion of the treating physicians.  Of the 295 patients hospitalized, only 11 underwent repeat CT, and only 1 abnormality was identified.  That one patient underwent neurosurgical intervention.  Of the remainder, no patients returned within two weeks of the initial incident with a further episode of delayed bleeding.  Thus, 0.3%.
Is it the difference between mandatory and discretionary repeat CT resulting in the wide range of reported incidence of delayed hemorrhage?  But, if there aren’t any symptomatic changes, are the extra hemorrhages detected clinically important?
Interestingly enough, they also reported three deaths from nosocomial pneumonia.  So, yes, the risk of delayed hemorrhage is non-zero – but likely lower than the risks associated with hospitalization.
“Outcomes of warfarinized patients with minor head injury and normal initial computer tomographic scan”

Ah, Is Choosing Wisely Futile?

… or is it just too early to tell?

Announced with much pomp, the ABIM Choosing Wisely campaign was picked up by many specialties.  Each chose at least 5 “low value” tests reflective of unnecessary resource utilization, and such have been disseminated via press release, social medial, and web presence.  The Annals of Internal Medicine further supports the endeavour by publishing original research in the same vein.

This retrospective analysis of patients enrolled in Anthem health plans evaluated the utilization trend for seven of these services over the past 3 years.  And, ultimately, the conclusion is: nothing has reliably changed.  Two of seven “low value” services had decreasing trends on the order of a fraction of a percentage, while two more had increasing trends on the order of a fraction of a percentage.  Just to tie in with one of my all-time favorite themes, antibiotics for acute sinusitis remained stable between 84.5% and 83.7%.

So, the early report is: no one’s making wise choices.

What’s the solution?  If you look around medicine for initiatives with the most robust implementation, it seems tied to either reimbursement or public shaming via quality scorecard.  Is it time for a value-based care rating?  Can we take all the resources devoted to Press-Ganey and repurpose them for good instead of evil?

“Early Trends Among Seven Recommendations From the Choosing Wisely Campaign”

A Mostly Uninformative Back Pain Trial in JAMA

There were almost 3 million visits to U.S. Emergency Departments last year for low back pain – and, yet, it is fair to say we manage these visits poorly.  Patients typically continue to have pain at discharge, at long-term follow-up, and we inject a vast number of opiates into circulation in the course of treatment.

This trial, appropriately, looks at a few of our most common prescriptions: scheduled NSAID (naprosyn), an anticholinergic smooth-muscle relaxant (cyclobenzaprine), and an opiate (oxycodone-acetaminophen).  Patients presenting to Montefiore hospital in the Bronx were randomized into three arms, and medications distributed in placebo-controlled, blinded fashion.

However, the primary outcome was not exactly what you might expect.  These authors used a questionnaire describing functional impairment, and used the change in impairment from enrollment to 1 week as their primary outcome.  This is an interesting choice as, while it doesn’t encounter some of the subjectiveness of pain scales, level of function is a technically surrogate outcome for pain relief.  Then, since it can be reasonably suggested the simple passage of time is the curative element in most cases of acute low back pain, it reasonable to expect such disability to regress to the mean, regardless of therapy, by one week.  I wonder if these authors did not inadvertently choose an outcome likely to show no difference.

And, that is precisely what they found.  At 1 week and – in another odd timeframe choice – 3 months.

However, that’s not the entirety of the story.  They measured many different outcomes, including adherence, utilization of the as-needed study medication, desire for same medication, days of return to work, and self-reported pain.  None of these secondary outcomes can be parsed reliably, but the general signal throughout is one of increased relief with the use of opiates.  It should be noted the authors’ conclusion is worded carefully – simply stating these data “do not support the use” of the therapies tested.  I am all for the avoidance of opiates, but the outcome measurement in this study probably obfuscates any potential benefit, and may rather mask their utility.

Lastly, these authors screened 2,588 patients with a complaint of low back pain in order to identify 390 meeting their inclusion criteria.  Radicular pain, traumatic pain, long-standing back pain, and elderly patients were all excluded – and, clearly, make up the bulk of visits to the Emergency Department.  These data apply to only a very small subset of patients, and they were enrolled at a single center.  The generalizability of their findings is not ideal.

It’s not a sexy topic, but its prevalence certainly makes it an important one.  If a fraction of the hundreds of millions of dollars devoted to the development of me-too blockbuster copycat drugs were devoted to such common issues, we would have far better data to guide the bulk of our practice.  I love that these authors did this trial – I just wish their measurements and timeframes differed.

“Naproxen With Cyclobenzaprine, Oxycodone/Acetaminophen, or Placebo for Treating Acute Low Back Pain: A Randomized Clinical Trial”

NEXUS Chest (Again)

This isn’t the first go ‘round for these authors.  In 2013, they published their first attempt – which demonstrated a sensitivity of 98.8% for thoracic injury, but a specificity of 13.3%.  The rule was, essentially, minimal improvement over clinical judgement and reliably decrease resource utilization.

In another massive effort, these authors enrolled 11,477 patients in 8 trauma centers – 6,002 in the derivation phase and 5,475 in the validation phase.  Based on their prior work, they evaluated each for the presence of 14 potentially predictive features.  Following data collection for the derivation phase, recursive partitioning was used to develop two decision instruments: one for “major” injuries requiring intervention, and one for “major and minor” injuries requiring a minimum of observation.

Their new criteria, comprised of about half of the 14 tested, were much improved over the prior study.  In the validation phase, the “major and minor” decision instrument had sensitivity of 99.2% for major injuries and 95.4% for major or minor injury.  Specificity was still poor at 25.5%, but improved from the prior study.

This validation phase, however, included only those undergoing CXR and CT chest – only 2,628 of the 5,475 enrolled.  The authors state “we have previously demonstrated that the clinically significant injury rate in this non-imaged group of patients approaches zero and is therefore negligible” – which certainly applies to the “major” injuries, but far less reliably to “major or minor”.  However, the more important implication is for the context of use for this decision instrument.  This instrument should not be applied to all trauma presentations – but, rather, to all trauma presentations for which CT chest imaging is initially judged necessary.  Only then, in the subset of patients who were otherwise on-their-way to CT, would this decision instrument potentially reduce imaging.  The authors do not provide its test characteristics in an all-comers population, nor provide comparative details regarding the characteristics of those imaged versus those not, so any further speculation is simply that.

As with the previously proposed NEXUS Chest, the same limitation applies – this is a one-way decision-instrument intended only to obviate CT, not inform clinicians for whom CT is needed.  As such, it is subject to the cognitive biases resulting from one-way decision-instruments, paradoxically leading to increased resource utilization.  It is, perhaps, the PERC rule for trauma – and do you think incorporating PERC increases, or decreases, CTPA?

“Derivation and Validation of Two Decision Instruments for Selective Chest CT in Blunt Trauma: A Multicenter Prospective Observational Study (NEXUS Chest CT)”

Not Much to Say About SPLIT

The Emergency Department is the land of fluid resuscitation.  The most typical resuscitation fluid tends to be 0.9% Normal Saline.  At simple face validity, this makes little sense as a volume expander – a poor mimic of basic physiology, and associated with an iatrogenic acidosis in large volumes.  Questions abound regarding the risks of renal failure associated with excessive saline administration, as compared to a more balanced or buffered solution.

This SPLIT trial, performed in intensive care units across New Zealand and Australia, provides little additional insight.  It is, happily, a lovely randomized, controlled, double-blind trial of fluid administration, comparing 0.9% NS with “buffered crystalloid”, also known as Plasma-Lyte.  Various outcomes measured changes in renal function, need for renal replacement therapy, and in-hospital mortality.

No differences.

But, also, not terribly generalizable.  Over 70% of these patients arrived to the ICU from the operating room, most of which was after elective surgery, most of which was cardiovascular.  Only 15% arrived from the Emergency Department, and only 4% carried a diagnosis of sepsis.  Patients also received only a median of 2 liters of fluid during their median of 1.5 days in the ICU.

It’s of mild interest to see no difference, but it does very little to further inform the sort of large-volume, rapid resuscitation routinely performed in the Emergency Department.

“Effect of a Buffered Crystalloid Solution vs Saline on Acute Kidney Injury Among Patients in the Intensive Care Unit The SPLIT Randomized Clinical Trial”

Also: for an excellent review of the “buffered crystalloid” solutions, visit PulmCrit.org.

STONE Score Close, But Not Quite

The STONE score is a lovely idea for a common problem – ureteral stones are a frequent source of low-value imaging in the Emergency Department.  An adequately sensitive or specific decision-instrument would improve diagnostic accuracy while reducing imaging.  Unicorns would bloom from the frozen tundra.

STONE, however, has a couple issues.  Firstly, it was derived and validated at a single institution in similar cohorts.  Secondly, it has a few variables that probably lack face validity, i.e., three points for non-black ethnicity?

This report is from the large, multi-center trial comparing CT vs. ultrasound for the diagnosis of ureterolithiasis.  Of the 2,759 patients randomized from the original trial, 845 underwent CT as gold standard and had the data available for evaluation of the STONE score.  This constituted their retrospective validation cohort.

So, STONE is OK.  Not great, but OK.  The AUC of STONE was 0.78, which is a step up from 0.68 of physician gestalt.  However, it was only 87% specific at the high end in the validation to rule-in stone, and, then, 96% sensitive at the low end as stone rule-out.  Phrased otherwise, there were about 10 to 20% fewer cases of ureterolithiasis identified in the moderate and high cohorts in this validation, compared with the original.  There were also, unfortunately, a greater number of significant alternative diagnoses in the low and moderate cohorts, as well.

The authors come to very reasonable conclusions.  First, the STONE score isn’t perfect.  It can be a useful adjunct to risk-stratification and shared decision-making, but the positive and negative likelihood ratios are inadequate as a standalone rule.  As a corollary, the score definitely requires refinement or reinvention, based on their analysis of the predictive contribution of different variables in the original score.  It is probably still fair to use this score to supplement gestalt in the context of pursuing judicious use of resources, but it must be considered in the context of other predictive features to determine the appropriate imaging strategy.

“External Validation of the STONE Score, a Clinical Prediction Rule for Ureteral Stone: An Observational Multi-institutional Study”

Again, One Troponin is Enough

As we’ve seen suggested in prior work, an undetectable high-sensitivity, high-precision troponin at presentation is a powerful predictor of minimal risk in the short-term.

In this prospective study based at four hospitals in Scotland, 6,304 eligible patients with suspected acute coronary syndrome were captured for analysis across three prospectively gathered cohorts.  These patients were split across a derivation cohort to establish a threshold with an appropriate negative predictive value, an internal validation cohort, and an external validation cohort.  Their target for a 30-day MACE of MI or cardiac death was a negative predictive value of 99.5%.

Making essentially a long story short, they found a threshold of 5 ng/L established this minimal-risk population, with only 9 patients meeting the primary outcome out of of 2,302 patients with an initial troponin less than the threshold.  And, if you want to be even more airtight, 6 of these presented within 2 hours of the onset of chest pain, and 3 had apparent ischemic changes on their initial EKG – which may have been picked up by a rapid repeat testing protocol in the ED for some patients.

There are a few holes in this paper, of course.  Not all patients were hospitalized and had such repeat testing, so some small nSTEMI or vasospasm events could have been missed.  Patients in this study required a median of 54 minutes for blood sampling after presentation to the ED, so some caution should be exercised regarding repeat troponin testing if your department is efficient regarding phlebotomy at presentation and onset of symptoms was just prior to arrival.  But, on the whole, this greatly adds to the body of evidence we’ve been building – that cutting edge troponin assays alone can provide powerful prognostic information.

Now, what to do with all the “intermediate” cases ….

“High-sensitivity cardiac troponin I at presentation in patients with suspected acute coronary syndrome: a cohort study”