Head Injury Showdown: PECARN Wins!

Most are familiar with the Pediatric Emergency Care Applied Research Network (PECARN) decision instrument for children with mild traumatic brain injury.  While they enrolled the largest number of patients in their derivation, they’re not alone:  the Canadian Assessment of Tomography for Childhood Head Injury (CATCH) and Children’s Head Injury Algorithm for the Prediction of Important Clinical Events (CHALICE) address similar clinical questions.

And, now that you know about them, you might as well forget about them.

This is a prospective validation of each of the three decision instruments at Denver Health, enrolling 1,009 children with blunt head injury and GCS 13 or greater.  52 patients had head injuries on CT, 21 of which were judged clinically significant, and 4 required neurosurgical intervention.  Based on the 90% of their cohort for whom they had complete outcome data, the PECARN rule was 100% sensitive and 62% specific, CATCH was 91% sensitive and 44% specific, and CHALICE was 84% sensitive and 85% specific.  Therefore it is most defensible to use the PECARN decision instrument in a setting concerned with maximal sensitivity.

However, what’s most interesting in this study – only 188 children underwent CT, and physician practice had 100% sensitivity.  All told, the PECARN instrument classified 47 as “high risk” and 335 as “intermediate risk”.  The original derivation publication states this intermediate cohort may be eligible for observation vs. CT, depending on provider comfort level.  Ultimately, the management of “intermediate risk” is the key to this instrument’s role in reducing resource utilization.  In many settings, such as this one, if the “intermediate risk” group predominantly undergoes CT rather than observation, resource utilization will increase, rather than decrease.

However, the Denver Health expertise is not generalizable to most institutions – but provides an interesting perspective on the performance of PECARN to expert clinical judgment.

“Comparison of PECARN, CATCH, and CHALICE Rules for Children With Minor Head Injury: A Prospective Cohort Study”

Skull On, Skull Off, Disabled or Dead

What is a “good outcome” for stroke patients?  Is it “alive”?  Or is it “alive & independent”, as in most of the tPA trials?  Through what lens ought we interpret the findings of some of these highly intensive interventions for stroke?

This is DESTINY-II, which enrolled elderly patients with malignant intracranial swelling following significant MCA territory infarction.  In this study, patients were randomized either to usual care or hemicraniectomy, a potential life-saving intervention that relieves intracranial pressure and reduces cerebral herniation.  The untreated cohort had awful outcomes – at 12 months, zero patients were free from disability, zero had mild disability, and 5% had moderate disability.  The remainder were severely disabled in dead.  The hemicraniectomy cohort also had awful outcomes – at 12 months, zero patients were free from disability, zero had mild disability, and 6% had moderate disability.

So, of course, this study was stopped early because of overwhelming benefit to the hemicraniectomy cohort.

The key difference – hemicraniectomy patients survived to be severely disabled, while control patients died.  76% of patients in the control group died vs. 43% of the hemicraniectomy group.  Most of the difference was made up by patients with mRS 4: “Moderately severe disability; unable to walk without assistance and unable to attend to own bodily needs without assistance.“

Determining the value of survival with complete dependence vs. death is clearly a challenging ethical decision.  Should this therapy be more widespread, given the resource intensive care and the ultimately dismal disability outcomes?  Those questions remain to be answered – but at least this study helps us better share the prognosis of either option with patients and families.

“Hemicraniectomy in Older Patients with Extensive Middle-Cerebral-Artery Stroke”

tPA Means: Blinders On, Full Power!

IST-3, despite its pro-tPA bias, was a negative trial for the primary endpoint.  But, that hasn’t stopped folks from looking at secondary outcomes and slicing it into subgroups and claiming – victory!

The root of the issue is:  in a negative trial, when you pull out a subgroup with positive outcomes, naturally there must be a counterweight subgroup with worse outcomes.  The trouble with the true believers/sponsored representatives is they focus only on the positive subgroup for hypothesis generation, and neglect the negative bits.

This latest rehash of IST-3 focuses on patients with high predicted risk of sICH or poor functional outcome.  These authors reviewed all the published prognostication tools and developed their own from the patient data in IST-3.  For what it’s worth, most of the tools performed approximately the same:  fair to poor.  The best AUC for prediction of sICH was 0.68 and the best AUC for prediction of post-tPA poor outcome was 0.80.

Then, these authors stratified the groups into low-, medium-, and high-risk for sICH or poor functional outcome with tPA, and compared the outcomes to the placebo group in IST-3.  For these subgroups, these authors found patients stratified as high-risk by the various scores tended to have better outcomes when treated with tPA compared with placebo.  Thus they conclude: “These data suggest that intravenous rtPA has an absolute beneficial effect in patients at a high predicted risk of sICH or poor functional outcome.”

… which is a fine and reasonable hypothesis given their baseline assumptions.  However, they never acknowledge the biases of IST-3 or the greater context of a negative trial – the counterweight that tPA is thereby futile or harmful in those at low- or medium-risk for poor outcome.

Of course, when most authors are funded by Boehringer Ingelheim to fly about the world and describe the wonders of tPA, it’s no surprise they choose to focus only on the conclusions that might lead to more tPA use, rather than narrowed, more appropriate tPA use.

“Targeting Recombinant Tissue-Type Plasminogen Activator in Acute Ischemic Stroke Based on Risk of Intracranial Hemorrhage or Poor Functional Outcome: An Analysis of the Third International Stroke Trial”

Procalcitonin in Serious Bacterial Infection: Spoiler Alert – It Doesn’t Help Here Either

A guest post by Anand Swaminathan (@EMSwami) of EM Lyceum and Essentials of EM fame.

Over the years, numerous studies have been published attempting to show the benefit for serum markers in diagnosing sepsis or other infections. These markers include ESR, CRP and more recently, procalcitonin (PCT). Despite the reams of literature published, no study has shown a true patient centered outcome benefit to using these markers. Instead of doing an in depth review here of the literature on PCT, I recommend reading Rory Spiegel’s post here.

This recent article from Academic Emergency Medicine attempts to use PCT as an indicator of serious bacterial infection (SBI) in children under 3 years of age. They basically compared PCT with WBC, absolute neutrophil count (ANC) and absolute band count. PCT had the largest area under the curve (0.80 vs. 0.76 for WBC, 0.73 for ANC and 0.67 for absolute band count). Overall, the study found that all of these tests suffered from poor sensitivities but that specificity for PCT (92.7% at a cutoff of 0.6 ng/ml) coupled with its sensitivity (51.6% at the same cutoff) yielded the best positive likelihood ratio of any of these tests (+LR = 7.04). Based on this finding, the investigators conclude that PCT is a “more accurate marker than white blood count, absolute neutrophil count or absolute band count in identifying young febrile infants and children with serious bacterial infections.”
But, are we asking the right question? This study, as with many of the others, tries to use PCT to identify patients that we would otherwise miss as having a serious infection. However, they don’t compare PCT to physician clinical judgment. Or, more importantly, they do not investigate if PCT adds to clinical judgment. Instead, they compare it to markers we know are seriously lacking in their ability to predict (WBC, ANC and absolute band count).
Additionally, the investigators focus on the positive likelihood ratio and the high specificity. But we aren’t concerned about overworkup in febrile kids. As with all bad diseases, we want high sensitivity to make sure we miss as few SBIs as possible and a low negative likelihood ratio to aid in risk stratification. With a strong negative likelihood ratio (-LR < 0.10) we could use a PCT < 0.5 ng/ml to risk stratify patients to a low or very low risk of SBI and potentially send them home with follow up. Here, a PCT < 0.5 ng/ml had a – LR = 0.52. In this study, 13.3% (30/226) patients ultimately had an SBI. If you started with a pretest probability of 13.3% and apply a – LR of 0.52 using the Fagan Nomogram (below) you’d get a post-test probability of around 10%. This is nowhere near low enough for us to stop our workup.
Where does this leave us? Biomarkers will continue to be pushed since there are strong industry interests. Additionally, we want something concrete, objective and tangible to help us with our clinical decision-making. Future studies, though should focus on the additional benefit of markers to the clinician’s assessment and gestalt instead of looking at the biomarker in a vacuum. Show us this and we’ll all sit up and take notice. Until then, procalcitonin is simply another test without a clear indication.

Special thanks to Rory Spiegel (@CaptainBasilEM) and Mike Mojica for the help with this post.

Focused Evaluation for “Lethargy & Poor Feeding”

As these authors note, infants are evil.

Well, more specifically, they note infants with non-specific complaints as benign as “crying” can be harboring serious pathologic diagnoses.  Therefore, the diagnostic work-up for such complaints as “lethargy” or “poor feeding” varies widely by clinician and comfort level.

These authors retrospectively reviewed charts for 352 infants 0 – 6 months with presenting complaint of “lethargy” or “poor feeding”.  They exclude the chronically ill/premature, abnormal vital signs, and those with recent trauma, and review the laboratory testing and ultimate diagnoses for each remaining patient.  Of the 272 remaining, 34 patients ultimately had a diagnosis requiring intervention or monitoring.  These included hematologic disorders, dehydration, intracranial bleeding and SBI.  Of these 34, 26 were otherwise well-appearing.  However, these authors note each of the well-appearing patients had some obvious focal finding on physical examination – mostly jaundice, leading to treatment for hyperbilirubinemia – leading to directed testing.  They conclude, therefore, a well-appearing infant with a reassuring examination does not need any specific testing or monitoring.

This study is limited by its retrospective nature, as well the lack of comprehensive follow-up.  That said, their algorithm for focused evaluation of “lethargy” and “poor feeding” is probably reasonable.  Fishing expeditions in the otherwise well infant are certain to be costly and low-yield, with continued caregiver observation and follow-up a more prudent plan.

“Diagnostic Findings in Infants Presenting to a Pediatric Emergency Department for Lethargy or Feeding Complaints“

Azithromycin, the World’s Most Effective Antiviral

The only thing better than providing one mostly useless treatment for influenza: providing two.

Sponsored by Pfizer and overseen by authors with Pfizer COI, this study randomizes patients between oseltamivir (Tamiflu) monotherapy and oseltamivir + azithromycin dual-therapy for influenza.  The theory behind the madness is azithromycin modulates anti-inflammatory processes and decreases the susceptibility to secondary bacterial pneumonia.  Thus, the primary endpoint of the study was … well, there wasn’t one.  “The primary endpoint was defined as variations in the levels of inflammatory markers” – 20-odd co-primary endpoints – while patient-oriented, symptom-oriented endpoints were secondary.

Of the 107 patients enrolled, baseline characteristics were similar – although the dual-therapy arm had significantly more cough.  And, as far as could be possibly conceived as relevant, all the outcomes were identical – although the dual-therapy arm had 16.1% incidence of possible drug-related adverse events, compared with 7.8% in the monotherapy arm.  As far as the “primary endpoint”, the authors data-dredged ten different inflammatory cytokines and serum markers for changes in levels between day 2 and day 5 – and also could not find any clinically significant positive findings.

Sadly, the authors were undeterred in their desire to support their initial hypothesis – and thus conclude in their abstract “combination therapy showed an early resolution of some symptoms.”  Specifically, on day 2 of therapy, there was a statistically significant difference in improvement in sore throat symptoms that evaporated by day 5 – and, using ANOVA, they found “sensation of heat” was decreased in the azithromycin group.  Considering this was an open-label study and the authors performed at least 60 different statistical comparisons, it’s simply tragic science they bothered to make any substantial note of these outcomes.

This is simply junk.  The pre-study likelihood of finding a difference must be considered low, so even the “trends” they observe in secondary endpoints should not encourage anyone to adopt this treatment strategy.  Please, please – don’t use either of these treatments for influenza in the absence of any sort of reliable evidence for benefit.  We have enough waste and harm in the world from these medications already.

“Efficacy of Combination Therapy with Oseltamivir Phosphate and Azithromycin for Influenza: A Multicenter, Open-Label, Randomized Study”

Alas, EGDT, We Hardly Knew Ye

Twitter and the usual accelerated knowledge translation sites have been abuzz with the release of several important articles regarding resuscitation in severe sepsis.

The one garnering the most press is ProCESS, a 1:1:1 randomization of patients with severe sepsis into Early Goal-Directed Therapy, protocol-based aggressive fluid resuscitation, or “usual care”.  Many detailed analyses and sub-analyses will likely be written, but, the basic gist – it appears the critical innovation coming out out of Rivers’ EGDT is awareness of the importance of any aggressive early recognition and treatment.  The primary outcome – in-hospital mortality at 60-days – was similar across each group.  And, the minor variations in secondary outcomes probably support simply paying close attention to individual patient physiology.

This is not specifically practice-changing in many critical care settings – there has been plenty of skepticism regarding the specific interventions in the Rivers’ algorithm.  The search will certainly go on regarding ways to improve upon the 20% modern mortality in severe sepsis, but it is now easily defensible to eschew the Edwards’ catheter, blood transfusions, and dobutamine from Rivers’ specific protocol.

“A Randomized Trial of Protocol-Based Care for Early Septic Shock”

The “Standard of Care”

A guest post by William Paolo (@paolomd1), the Program Director of Emergency Medicine at SUNY Upstate.
“Standard of care” is a legal term whose colloquial medical usage, outside of tort law, has been unfortunately adopted by the medical infrastructure into its cultural lexicon.  The implications of its usage, when related amongst physicians, is the suggestion that there is an accepted, established, and parsimonious rendering of medical care that all reasonable providers would, under similar circumstances, judiciously employ.  It serves as an idealistic touchstone resting upon the foundations of summated evidence via which clinicians measure their individual and collective performances.  Actions that deviate from the collective wisdom are deemed inappropriate, negligent, and worthy of derision for failing to practice within the established evidentiary parameters of the authoritative collective guild.  Undermining this concept are the radical disparities of an agreed upon standard among clinical specialists and varying geographical norms that disrupt the foundations of a standardized standard of care.   The very term itself is normative, proposing what ought to be rather than what currently is, based upon a leap of logic that has never been fully supported by medical empiricism as expressed within the evidentiary literature.  The standard therefore may be determined by the collective, but more often it is determined by a scant few individuals utilizing the argument from authority to prescribe practice patterns.  The difficulty lies in prospectively determining what current “standard of care” actually results in patient harm, as the medical story is replete with examples of injury obvious only in retrospect.
The PROWESS study was released in 2001 in which activated protein C as manufactured and distributed by Eli-Lilly under the name Xigris was evaluated for the treatment of severe sepsis.   1690 people with septic shock requiring vasopressors were randomized to receive either activated protein-C or placebo.  The primary end point was death from any cause 28 days after infusion. Because of the results the phase 3 trial was stopped early having demonstrated an absolute mortality reduction of 6% yielding a number needed to treat of 17.   As is now widely known there were multiple issues with the original study and the subsequent 2012 PROWESS-Shock study demonstrated no benefit and potential harms of Xigris.  In 2014 it is easy to appreciate the issues of harm and need for reproduction and verification of PROWESS to overcome equipoise however physicians in 2001 had a very well done study (as most industry supported research is—though it is also done well to bias in their favor) that was stopped early due to patient benefit.  One could not fault a 2001 physician for referring to activated protein C as the new standard of care for sepsis—or can we?
Standard of care forces physicians to adopt an intellectually closed approach to evidence presuming that science has settled particular questions regarding clinical conundrums.  Retrospectively the foolishness of this position is obvious as the inexorable progress of empiricism wrought through experimentation recurrently dismantles accepted evidentiary norms.  The “standard” of current epochal standard care has no more underlying claim to absolute truth-value than previous erroneous medical misadventures exemplified by the various theories of humorism.  The problem, as it were, is one of perspective as it is difficult discern objective truths when temporally related to the perpetuation of often faulty ideas and attitudes.  Only the march forward of time and accumulated wisdom is able to dismantle that which seemed once intuitively and evidentially obvious in a given medical period.  The reasonable intellectual position to therefore adopt, as a profession, is one of radical agnosticism towards absolute truth claims and delineations of care as defined by standards.  This is not to say that we should fall into nihilism and presume that all of our current care will one day be proven mistaken and therefore be paralyzed by the knowledge of transformation.   The story of medical science, as all of science, is replete with advancements and misadventures with a clear arrow of progression. “Standard of care” adopts a position of unsupported truth-value without the reason necessary for its nuanced interpretation.  Though we may continue to utilize it as a profession it would be preferable to hand it, in its entirety, back to the lawyers who endowed us with it at the beginning.

“Efficacy and safety of recombinant human activated protein C for severe sepsis.”

Will Twitter Ruin Your Diagnostic Abilities?

Medical errors, by some estimates, are associated with cognitive biases up to 75% of the time.  Given the oft-quoted 98,000 deaths per year as a result of medical error, recognition of these biases seems prudent.  Knowing is, after all, half the battle.

One of these is “availability bias”, the tendency to conflate the likelihood of disease depending on whether the details are readily present in memory.  Essentially, if you don’t think of it – you’ll never diagnose it – but if you think of it too frequently, you might test or treat for it with greater frequency than appropriate.

These authors subjected 38 internal medicine residents to a simulation where they read Wikipedia entries on two diseases.  Six hours later, they were asked to review and submit diagnoses for eight cases – two of which superficially resembled the disease descriptions from Wikipedia.  Finally, the residents were asked to use a structured methodology evaluating signs and symptoms in order to systematically create and winnow a list of potential diagnoses.

I’ve probably already clued you into the end result – but, basically, in the initial case review, residents had a 56% correct diagnosis rate for the “availability bias” cases and a 70% correct diagnosis rate for the others.  Then, by simply re-reading the cases in a systematic fashion, they subsequently were able to bring their rate of correct diagnosis up to 71% on the bias cases.

So, the next time you discover something novel and interesting on Twitter – try not to take it with you to work unchecked ….

“Exposure to Media Information About a Disease Can Cause Doctors to Misdiagnose Similar-Looking Clinical Cases”

There’s No Telling What Patients Want

“Shared decision-making” has become a frequent watchword of sorts, encompassing participatory concepts in which patients are better involved in their own care.  I, and many others, have espoused this sort of paradigm in medicine.

Unfortunately, there’s a bit of a problem.  On the physician side, we probably don’t have good mechanisms through which to translate evidence to individual patients.  Most information derived from clinical studies describes outcomes from aggregated cohorts – so, usually, the best we can do is inform our patients how the “average” person performed with a specific treatment.

Then, on the patient side – as this study demonstrates – their risk-taking behavior is heterogenous, irrational, and extreme.  These authors report on 234 surveys of patients presenting with low-acuity chest pain in a Veterans Affairs cohort, trying to get a handle on hospitalization preferences given a certain pretest likelihood of disease.  Their basic model:  hospitalization reduces the risk of bad outcome by 10%.  Then, they asked if the patient would like to be hospitalized for base likelihood of poor outcomes ranging from 1 in 2 to 1 in 10,000.

Half the patients wanted to be hospitalized, even when the benefit to hospitalization reduced the event rate from 1 in 10,000 to 1 in 11,000 (an NNT of 110,000).  Then, another 10% of patients wanted to be discharged in all circumstances, even when the risk of poor outcome was improved from 1 in 2 to 5 in 11 (an NNT of 22).  And, depending on how the risks were communicated, and whether visual or numeric scales were used, also affected how the patients chose.

So, ultimately – yes, we’d like to involve patients in their decisions.  But, unfortunately, it looks as though it’s going to be quite the challenging proposition – and we might not like (or have the capacity to abide by) their preferences.

“Measuring Patient Tolerance for Future Adverse Events in Low-Risk Emergency Department Chest Pain Patients”