Monday, February 6, 2012

Would Free Medications Help?

It's too bad this study doesn't actually look at what I would have hoped it would - but it's interesting, nonetheless.  One of my hospitals is a true safety-net hospital and we see, repeatedly, repeatedly, repeatedly, the complications of neglected chronic disease.  One of our frequent laments is whether the costs of recurrent acute hospitalization wouldn't be prevented a hundred times over if we'd simply sink some costs into preventative maintenance care, free medications, etc.

This study almost looks at that.  This is from the NEJM which compared the outcomes of patients following myocardial infarction, and they follow a group which receives completely free medication and a group that does not.  Unfortunately, the group that does not receive free medications is still receiving heavily subsidized medication support, and is only responsible for a co-pay.

Despite only needing to come up with a co-pay, there's a significant difference in medication compliance, with an average absolute difference in full adherence with medications of ~5-6%.  With this minimal absolute difference in adherence, the full adherence group had significantly fewer future vascular events - mostly from stroke and myocardial infarction - approximately a 1% absolute decrease.  There was a non-significant decrease in total costs associated with the patients who were on the full-coverage medication plan.

Now, they don't follow-up any medication-related adverse events, so this is the most optimistic interpretation of benefits of full-coverage, but it would seem that it is overall cheaper and more beneficial to supply medications for free.  And, it makes me wonder what the results of a similar cost/health-benefit study would show in our safety-net population.

"Full Coverage for Preventive Medications after Myocardial Infarction"

Saturday, February 4, 2012

Safety-Nets & ED Length of Stay

This is a relatively intriguing public policy article in JAMA following up in a timely fashion regarding the new CMS Emergency Department quality measures.  These new measures include various time-to-X measures, including length of stay, length of time to admission from bed request, etc.  There is some concern that these quality measures may be tied to federal funding, unfairly targeting "safety-net" hospitals that are not at baseline provided with the resources to address patient flow issues.

This article is a review of the NHAMCS database, a national probability sample survey of patient visits, looking at independent predictors of increased length of stay in patients admitted and discharged from the Emergency Department.  Based on the review of this sample, they do not see a significant difference in ED length of stay - and conclude that these quality measures should not be of concern to "safety net" EDs.  However, these general time-based measures mask most of the problems encountered in "safety net" institutions.

There are some baseline differences in patient characteristics between the safety-net and non-safety-net hospitals in their sample, and they tend to work in favor of safety-net hospitals.  The safety net hospitals in this sample tended to have younger patients with lower triage acuities, which should work in favor of reduced ED overall average length of stay.  My anecdotal experience suggests that, once the quality measures track more detailed ED transit times, I believe we will see more significant deficiencies drop out in the safety-net group.

"Association of Emergency Department Length of Stay With Safety-Net Status"

Thursday, February 2, 2012

Half of Fractures Received No Analgesia

One of the new CMS quality measures involves measuring time to receipt of pain medication for patients diagnosed with long bone fractures.  While this isn't the most exciting quality measure in terms of outcomes, it is probably a reasonable expectation that fractures receive pain control, and it might be a plausible surrogate marker for overall Emergency Department operations - at least, until the powers that be focus solely on these few measures at the expense of other clinical operations.

This article is a retrospective review of all pediatric long bone fractures evaluated at their facility.  They used the electronic medical record to track the timing of any "adequate" pain medication.  They have a specific weight-based definition of "adequate" for IV narcotics, PO narcotics, and non-narcotic analgesics, and they specifically break down pain medication received within 1 hour of arrival.

They identified 773 cases in their records, and by their definitions, 75 patients received an "adequate" dose of pain medication within 1 hour.  One can quibble with their definition of "adequate" because there is a range of pain needs that don't necessarily require maximal dosing.  But, you cannot quibble with the fact that 353 children received no pain medication at all within an hour of ED arrival (or prior to ED arrival).  Certainly, some individual factors at play would result in reasonable delays to pain medication, but definitely not nearly half.

"Analgesic Administration in the Emergency Department for Children Requiring Hospitalization for Long-Bone Fracture"

Tuesday, January 31, 2012

Congratulations Michelle Lin!

One of the prominent medical education bloggers - who is really much more than just a great blogger - has been awarded an endowed chair by the University of San Francisco School of Medicine to support her medical education efforts.  This is notable to me because, in the press release, they specifically mention part of the mission of the award specifically notes "keep up her active 'Academic Life in Emergency Medicine' blog".  

It's fascinating to see how alternative publication sources and online media are influencing the perception of "academic achievement".  For instance, my JAMA commentary - a journal with Impact Factor of 30 - has been viewed as full text or downloaded as PDF ~2000 times in the last six months.  This blog, on the other hand, exceeds 400 views per day.  There's no question which has been more rewarding to my brief career so far.

Again, congratulations to Michelle!  Now she has to do, not just great things, but insanely great things!  (also, go Stanford!)

"Inaugural Academy Chair in Emergency Medicine"

Dosing Errors With IV Acetaminophen

As a follow-up to the recent posting regarding IV acetaminophen, this recent article in Pediatrics highlights a few case reports regarding overdose.

According to the authors, the most frequent error in administration when the order is written in milligrams, but the medication order is administered in milliliters - a 10-fold overdose.  All of the patients in this series received n-acetylcysteine infusion, and none appeared to suffer significant liver injury specifically attributed to the overdose.

Another lovely demonstration of the potential for iatrogenic injury in healthcare.  Even the most apparently benign orders can have unanticipated harmful consequences, and a demonstration how intravenous administration is at higher risk.

"Intravenous Acetaminophen in the United States: Iatrogenic Dosing Errors"

Monday, January 30, 2012

Scattering Tacks In The Road

I might be the only one who finds the irony in this, but, at long last, we have a rapid assay to estimate the activity of the new oral direct thrombin inhibitor, dabigatran.

Just to recap, with coumadin, we can measure PT/INR; for heparin, PTT; and for enoxaparin and its brood, (less rapidly) Factor Xa levels.

Now, we have the HEMOCLOT test.

Created and marketed by Boehringer Ingelheim, the manufacturers of dabigatran. (Edit: sorry!  This is not manufactured by Boehringer - they only published this study.  Boehringer is, however, working on a FAB antibody to dabigitran to use as an antidote, however.)

It's a beautiful piece of business to put a dangerous medication on the market, and then sell the only practical means of monitoring levels.

"Using the HEMOCLOT direct thrombin inhibitor assay to determine plasma concentrations of dabigatran."

Sunday, January 29, 2012

Further Harms of IV Contrast

Radiation: cancer.  Iodinated contrast: renal injury.  Now, iodinated contrast: thyroid dysfunction.

This is a retrospective, matched, case-control study performed in Boston to evaluate any association between CT administration of IV contrast and hyper- and hypothyroidism.  They gathered 178 new-onset hyperthyroid and 213 new-onset hypothyroid cases and statistically matched them in their patient database to euthyroid "controls".  There were no significant differences between the groups at baseline - although, they don't match between terribly many clinical variables.

In the end, they find the patients who developed thyroid dysfunction had higher rates of iodinated contrast exposure - primarily from cardiac catheterization, but also from CT scans.  For hyperthyroidism, 6.1% of controls had contrast exposure, whereas 10.7% of their hyperthyroid patients had received contrast.  For hypothyroidism, the numbers are 8.5% controls vs. 12.2% hypothyroid.

It's a bit of a backwards way to approach it - ideally they'd compare a group receiving iodinated contrast against a group that did not, and observe the incidence of thyroid dysfunction - but it seems that's not the format of data to which they have access.  In any event, the physiologic basis is reasonable for the association - more data needed to confirm these findings.

Just in case you needed another reason to not order a contrasted CT.

"Association Between Iodinated Contrast Media Exposure and Incident Hyperthyroidism and Hypothyroidism"

Friday, January 27, 2012

The Harmful Rush To Hypothermia

Mild hypothermia seems to be a clinically useful therapeutic modality for improving neurologic outcomes following return of spontaneous circulation in cardiac arrest.

However, like any emerging therapy, the precise details regarding which patients are most likely to benefit and how to best apply it are still in flux.  This is an Italian registry study that gathered prospective data on all individuals at 17 hospitals who underwent therapeutic hypothermia following cardiac arrest.  The specific question asked by these authors is regarding the optimal time for initiation of hypothermia - using 2 hours after ROSC as their cut-off.

Turns out, they found an association between "early" (< 2 hours to initiation) therapeutic hypothermia and worsened mortality - 47% mortality vs. 23% mortality in the ICU.  This ~20% absolute difference in outcomes holds up over the 6 month follow-up period.  No difference in cerebral performance category is observed between the two groups, although there is a nonsignificant trend towards better CPC in the "early" group.

Hard to know what to actually do with data.  Is early hypothermia truly harmful?  Because of the observational design, it's hard to say whether there aren't confounding baseline differences in the "late" population that produces selection bias for higher survival rates.  Or, are the mortality rates higher in the early group because patients are incompletely resuscitated before initiating hypothermia?

More questions, no answers.

"Early- versus late-initiation of therapeutic hypothermia after cardiac arrest: Preliminary observations from the experience of 17 Italian intensive care units"

Wednesday, January 25, 2012

Helping TPA Help Patients Bleed

TPA for stroke, the miracle therapy that has your Emergency Department shoving people out of the way to drag someone to the CT scanner within 10 minutes of ED arrival, isn't good enough.  After all, TPA, a "clot-busting" drug that saves dying brain cells by restoring flow, only completely opens up the occluded target vessel within 2 hours in 20 to 30% of the cases, with partial recanalization occurring in up to 60%.  So, the "Texas Biotechnology Corporation" and their equity stakeholders at The University of Texas Health Science Center at Houston have undertaken a project to add additional anticoagulation - argatroban - to TPA in the interests of actually delivering on the "clot-busting" part of the promise.

This is an open-label, pilot safety study enrolling 65 patients.  It was stopped after the first 15 patients for safety review after two experienced intracranial hemorrhage.  After review, it was restarted with additional restrictions on only giving it to milder stroke patients with NIHSS score < 15 (right hemisphere) and < 20 (left hemisphere).  All patients subsequently underwent vascular imaging to assess for recanalization, and the authors reported safety outcomes for events within seven days.

The good news: sorry, no good news.  14 had sustained complete recanalization at 2 hours - 30%.  An additional 12 patients had sustained partial recanalization at 2 hours - 25%.  Of course, this isn't a controlled trial, so comparison to the recanalization rates demonstrated in existing literature is flawed - but it's certainly not an order of magnitude better.

But, this wasn't an efficacy trial, this was a safety trial.  And seven patients met the ultimate safety endpoint of death - 10%.  For intracranial hemorrhage, 19 (29%) patients had ICH, 3 of which were symptomatic. Because NIHSS score predicts bleeding, we can compare to the NINDS trial TPA group, whose median NIHSS score of 14 compared with this trial's median of 13.  The NINDS trial showed a 10.8% rate of ICH and about 4% mortality at 7 days.

Seems like a treatment with triple the ICH and double the mortality, and that isn't proven superior, shouldn't support the conclusion of "potentially safe" or that "Further study of this treatment combination appears warranted."

"The Argatroban and Tissue-Type Plasminogen Activator Stroke Study : Final Results of a Pilot Safety Study"

Monday, January 23, 2012

Progress In Combating Publication Bias

...if from abysmally terrible to embarrassingly bad represents progress.

A certain subgroup of trials registered with are required to report their results within one year of conclusion of study.  These mandatory-reporting requirements include clinical trials of FDA-approved drugs, devices, or biological agents that have at least one study site in the U.S.  In the future, this will expand to include unapproved drugs.  These requirements, ideally, should help reduce publication and sponsorship bias by ensuring result availability regardless of ability to obtain publication or the desire of a pharmaceutical corporation to publish negative results.

And, so far, these authors discover that it is a tremendous success - trials subject to the mandatory reporting complied with the requirement in twice as many of identified trials as compared with registered trials that were not required to report results.

Unfortunately, twice as many was only 22% compared with 10%.  So, there's still quite a ways to go before we have full transparency in clinical trial reporting - but it's "progress."

"Compliance with mandatory reporting of clinical trial results on cross sectional study"

Saturday, January 21, 2012

Observation For Anticoagulated Head Trauma

Coming in a future issue of Annals, the Editor's capsule summary: "Delayed intracranial hemorrhage is common after minor head injury when patients are receiving warfarin. A minimum protocol of 24-hour observation followed by repeated scanning is necessary to detect most such occurrences."

Now, this isn't a terribly management agnostic statement.  It does not specifically state this is something we need to start doing - but it rather implies that, if you don't, you'll be missing this "common" phenomenon.  It isn't an alien concept - since 2002, the European Federation of Neurological Societies has recommended admission for observation after minor head trauma - but it's certainly not the standard of care here.  So, for the Annals editors to state that observation and repeat scanning is "necessary", they must obviously have excellent evidence.

Or they have an observational case series consisting of 87 patients from Italy.

These authors present a prospective case series of all patients at their institution who were admitted for observation specifically for minor head trauma while on oral anticoagulation.  At the time of repeat CT scanning 24 hours later, the authors report five of them had new bleeding detected.  In addition, two patients who were discharged after two negative CT scans returned with symptomatic bleeding, one at two days, and one at eight days.

So, should we be observing and rescanning every anticoagulated minor head trauma patient as these authors suggest (and as they do in Europe)?  If you practice in a zero-miss litigation environment, this article and ACEP's apparent embrace of the results will hamstring your decision-making.  This data is completely inadequate to change clinical practice, and inconsistent with prior literature documenting delayed hemorrhage in only 2 of 137 patients.

Clearly, some patients will have delayed bleeding - a subset of which will be clinically significant.  However, we simply cannot expose all anticoagulated patients with minor head trauma to the harms and costs of hospitalization.  Better studies are required to prospectively determine the risk profile of patients who require further observation in a hospital setting, rather than a watchful discharge home.

"Management of Minor Head Injury in Patients Receiving Oral Anticoagulant Therapy: A Prospective Study of a 24-Hour Observation Protocol"

Friday, January 20, 2012

Correspondence in Nature Reviews Neurology

After publishing an article that alluded to the "antipathy" of emergency physicians towards TPA in acute ischemic stroke, the editors were nice enough to publish my correspondence defending the reasonableness of a cautious attitude.

"Skepticism about thrombolytics in stroke is not unreasonable."

Thursday, January 19, 2012

EMLit on ERCast

It's like podcasting with Jack Buck - always a pleasure.  Rob Orman does great work, and he's got a huge archive of invaluable wisdom accumulated through his podcasts.  I hope you enjoy our most recent episode together.

"Decision Tools: PERC, NEXUS and CURB-65"

Lies, Damned Lies, and Tamiflu (oseltamivir)

Receiving quite a bit of press yesterday, and rightfully so, the Cochrane Collaboration published their analysis of oseltamivir - the miracle influenza antiviral that (at great cost) is part of our nation's strategic stockpile for an influenza pandemic.  The story is interesting both regarding what they found, and what they didn't find.

As for the data from the review, the numbers are similar to what we've been basing our practice upon - oseltamivir significantly shortens the length of time until symptom improvement from 160 hours to 139 hours.  However, it did not demonstrate any difference in hospitalization rates.  Additionally - whether through study bias or by direct medication effects - the oseltamivir groups were significantly less likely to have a confirmed diagnosis of influenza.

So, this suggests that it's a little troubling that we've gone to all the expense to stockpile this expensive medication that does not appear to reduce hospitalizations from influenza - and it remains an individual decision whether that extra day of symptom improvement is worth exposure to the side effects of the medication.  But the reason this is national news is that Roche pharmaceuticals refused to supply the promised clinical data requested by the Cochrane Collaboration; the published analysis is based on 15 oseltamivir studies with complete information, and excludes 42 other studies with discrepancies in the data.  This sort of behavior is just another representative sample of the unethical, but completely understandable, profit-motivation of pharmaceutical corporations protecting their financial interests.

I would be greatly surprised if the clinical data Roche is holding onto supports oseltamivir efficacy.

"Neuraminidase inhibitors for preventing and treating influenza in healthy adults and children - a review of clinical study reports"

"Flu Drugs: Search for evidence goes on"

Tuesday, January 17, 2012

100,000 Incorrect TIA Diagnoses Every Year

...if we extrapolate the results from this single-center study to the entire United States.

In Annals this month, a chart review of the "discordance" between the final neurologist "gold standard" diagnosis and the provisional Emergency Department diagnosis.  Apparently, in Cleveland, 36% of the patients receiving an initial diagnosis of transient ischemic attack in the Emergency Department are subsequently evaluated by a neurologist and given an alternative diagnosis.  As the authors note in their introduction, the diagnosis of TIA is made 300,000 yearly - and if 36% of those cases are made incorrectly, then we're theoretically admitting 100,000 patients for extensive and expensive evaluation.

So, are they right?  Well, if the three neurologists responsible for 93% of their 427 evaluations are representative of the entire country, perhaps.  Or, if their chart review methods are adequate - as the authors note, one of their chart audits changed an abstracted diagnosis of TIA to "right hip pain" - then, perhaps.  If you ignore that ED physicians have a few minutes of history, examination, and limited imaging available at their disposal - compared with the neurologists that can subsequently perform any manner of inpatient studies that might uncover an alternative diagnosis mimicking a TIA - then, perhaps.

If neurologists are walking into the ED and evaluating patients under the same constraints as we are and producing this level of discordance, then we have a problem.  But, I don't think this study tells us anything we can use to evaluate ED physicians' ability to appropriately include or exclude TIA in the differential for neurologic complaints of a transient nature.

Somewhat disappointing that a small, retrospective chart review with results that might not be internally or externally valid are in the premier, #1 Impact Factor journal of our profession.

"Variables Associated With Discordance Between Emergency Physician and Neurologist Diagnoses of Transient Ischemic Attacks in the Emergency Department"

Sunday, January 15, 2012

PERC - Still (Mostly) Useless

...except, perhaps, in a risk-management sense - but, only if we keep beating it down into its narrowest application due to its terrible specificity.

This most recent Annals publishes a systematic review of the Pulmonary Embolism Rule-Out Criteria, a decision instrument recommended in ACEP's pulmonary embolism clinical guidelines as a reasonable tool to risk-stratify a patient into a so-called "zero-risk" population that does not require any testing - not even a D-dimer.  And, I think they do a reasonable job including studies and summarizing the data, especially considering the width of the error bars on a lot of these studies.

The key points - pooled sensitivity is 97% when applied to a low-risk (Wells, Geneva, gestalt, whichever) population with a negative LR of 0.18.  This means, if you had someone who you already didn't think had a PE and they meet PERC criteria, it helps you with your medicolegal documentation, since it's in ACEP's guidelines.  The negative LR is strong enough to be helpful - but when you're already looking at single-digit percentage risk for PE, the absolute reduction in risk is quite small.

The important point to hammer home is the positive LR is only 1.23, which makes it the D-dimer of decision instruments.  Please don't justify further work-up just because they fail PERC - it barely moves the needle with its terrible specificity.  You need to have another clinical justification for further work-up in pulmonary embolism.

As an aside, in this era of over-testing and over-diagnosis of PE, the diagnosis of PE isn't necessarily the ideal endpoint - what we should be following are patient-oriented outcomes such as death/heart failure in untreated PE in PERC-negative patients to truly make it a valid tool.

"Diagnostic Accuracy of Pulmonary Embolism Rule-Out Criteria: A Systematic Review and Meta-analysis"

Friday, January 13, 2012

Heart Failure, Informatics, and The Future

Studies like these are a window into the future of medicine - electronic health records beget clinician decision-support tools that allow highly complex risk-stratification tools to guide clinical practice.  Tools like NEXUS will wither on the vine as oversimplifications of complex clinical decisions - oversimplifications that were needed in a pre-EHR era where decision instruments needed to be memorized.

This study is a prospective observational validation of the "Acute Heart Failure Index" rule - derived in Pittsburgh, applied at Columbia.  The AHFI branch points for risk stratification described below, in this extraordinarily complex flow diagram:

Essentially, the research assistants in the ED applied an electronic version of this tool to all patients given by the Emergency Physician a diagnosis of decompensated heart failure - and then followed them for the primary outcome(s) of death or readmission within 30 days.  In the end, in their small sample size, they find 10% of their low-risk population meets the combined endpoint, while 30.2% of their high-risk population meets their combined endpoint.  Neither group had a very high mortality - most of the difference between groups comes from re-admissions within 30 days.

So, what makes this study important isn't the AHFI, or that it is reasonable to suggest further research might validate this rule as an aid to clinical decision-making - it's the progression forwards of using CDS in EHR to synthesize complex medical data into potentially meaningful clinical guidance.

"Validating the acute heart failure index for patients presenting to the emergency department with decompensated heart failure"

Wednesday, January 11, 2012

Can We Stop Placing NG Tubes?

One of the worst-tolerated procedures in Emergency Medicine - placement of the NG tube.  Unfortunately, when I call my GI fellow on-call for any upper GI bleeding, the first question is invariably - what did the NG lavage show?

There is good evidence demonstrating that positive NG lavage tends to identify the presence of high-risk lesions found on subsequent endoscopy.  There is also evidence that endoscopic treatment of high-risk lesions decreases rebleeding and mortality.  So, if NG lavage identifies high-risk lesions, and endoscopic treatment of high-risk lesions decreases mortality, then patients who undergo NG lavage for their upper GI bleeds should have lower mortality, right?

This is a retrospective review of all the patients admitted to the West LA VA with a diagnosis of upper GI bleeding - a sample of 632 meeting inclusion criteria.  Of these, 255 did not undergo NGL and 378 did.  What's interesting in this article is that the authors attempted to statistically create two identical cohorts using propensity scoring.  They ended up with two nearly identically matched cohorts of 193 patients from the original 632 based on demographics, triage, lab values, physiologic characteristics, and medical interventions.

Between these two groups, they found no significant difference between mortality, hospital stay, emergency surgery, and blood transfusion requirements.  There was a statistically significant difference in the number of patients who underwent endoscopy - patients who didn't receive NGL had 60% endoscopy vs. 72.3% in the NGL group.  This is mildly interesting - considering that, in theory, the identification and endoscopic treatment of high-risk lesions is associated with increased survival - and if you're doing less endoscopy on an identical patient cohort, you should be missing the opportunity to treat those lesions.  Yet, there was no significant difference outcomes between cohorts.

So, yes, if you wanted to stop placing NG tubes because they're uncomfortable for patients and apparently don't change ultimate outcomes - certainly, that may be reasonable.  Some gastroenterology literature suggests patient-specific risk factors are more important in predicting the impact of endoscopic intervention on outcomes, rather than the limited information derived from the NG lavage.

However, this is just statistical calisthenics in an attempt to replicated a randomized-controlled trial and doesn't give us the prospective evidence needed to change practice.  Or argue over the phone with the GI fellow.

"Impact of nasogastric lavage on outcomes in acute GI bleeding"

Monday, January 9, 2012

So, NEXUS Is Invalid?

Another doom and gloom trauma article that wants to take one of our most cherished tools away from us regarding the evaluation of the blunt trauma patients.  Certainly, nothing is sacred, but these authors want to take NEXUS out to the woodshed and make sure every trauma patient gets a CT of the c-spine.

The premise of their argument is reasonable - NEXUS was derived in an era of plain films for radiographic clearance of the cervical spine, and now, many studies have observed that CT with 3D reconstruction picks up potentially significant injuries that could be missed by plain x-rays.  Therefore, the gold standard incorporating plain radiography for NEXUS renders it invalid due to missed injuries.

These authors performed a prospective evaluation of the NEXUS rules by applying them to 2,606 adult trauma patients, all of whom underwent 16 multidetector CT scanning with 2mm thick axial cuts.  They found 157 patients with a total of 258 fractures - and note that 26 patients had fractures identified despite meeting NEXUS criteria.  Of these 26, 16 were managed in a c-collar, 2 underwent operative stabilization, and 1 had a halo placed.  Therefore, they simply conclude that NEXUS is not externally valid to their trauma population and everyone should receive a CT of the c-spine based on mechanism.

Finding flaws with NEXUS - excellent, let's identify the subset at higher-risk so we can prevent missed injuries.  However, this article doesn't help us at all.  They don't do any sort of descriptive analysis of the NEXUS-negative patients who end up with significant injuries with which to educate our practice.  They simply conclude with the blanket statement that the dollar cost of performing all the CTs is less than the dollar cost of potential malpractice payouts.

In an era where we're trying to cut healthcare costs and reduce the practices of defensive medicine, this is precisely the sort of article that we don't need.  This is fantastic data presented in a non-constructive fashion that will likely, as the authors seem to intend, ensure the 97% of NEXUS-negative patients who had no injuries get their CT of the c-spine.

"National Emergency X-Radiography Utilization Study Criteria Is Inadequate to Rule Out Fracture After Significant Blunt Trauma Compared With Computed Tomography"