The Opiate Overdose Train

There is a certain inalterability about trains.  Their travel is predictable and linear.  Slowing and stopping are extended affairs.  It’s hard for a train to make a sharp turn.

Apparently, opiate prescribing is like that.

This study reviewed administrative data from a health insurer to identify patients receiving long-term opiate therapy.  Patients were then included for analysis if they had a visit to an Emergency Department or a hospitalization related to heroin or opiate overdose.  These same patients were then followed for up to 2 years following the index overdose, and their opiate prescribing tracked.

With a median follow-up of 299 days, opiates were dispensed to 91% of patients following opiate overdose – 7% of whom went to to repeated overdose, many of whom had multiple overdose events.  Some patients had their opiate quantities curtailed, but the majority received the same – or even more – opiates after the overdose event.

Certainly, some of this prescribing is still appropriate – our tools for managing severe pain are grossly inadequate, and in hospice settings, inadvertent overdose is an acceptable hazard of control of malignant pain.  But, just as certain, there is a cohort suffering the harms of shocking irresponsibility.

We’ve been getting bombarded with information regarding the harms of opiate prescribing for several years now; why are we still inflicting such great harm on the healthcare-seeking public?

“Opioid Prescribing After Nonfatal Overdose and Association With Repeated Overdose”

Santa “Danger” Claus

If you’re like me – and have drawn the Christmas Holiday straw this year – perhaps you’re lamenting your choice of profession as you drive into work.

Of course, you could have Santa’s job.

Night-time flying, a sleigh-landing hazards, mental health stressors, cardiovascular risks … it’s a wonder Santa survives and returns year after year!

“The occupational health of Santa Claus” (free fulltext)

What’s the Big Deal About Meclizine?

When I was growing up, I used to go fishing with my Dad.  As we sat, in the dark and wet, freezing in Tillamook Bay in November, he would offer sage words of advice:  “Vertigo?  Meclizine, of course.”*

And, so, through residency and even as teaching faculty, the standard antihistamine treatment for vertigo has always been meclizine – until this past month, when all the meclizine was pulled from our Pyxis due to manufacturing contamination.  So, what would we use as an alternative?

I will not take credit for having performed the most exhaustive of possible literature searches, but, by far, the most fantastic piece of literature I found was from 1968.  This comes from the era of the Cold War and the Space Race, and the folks most interested in effect anti-motion sickness was NASA.  So, among other things, early aerospace medicine was interested in preventing astronauts from vomiting their brains out as they were spun and shaken under otherworldly conditions.  To do so:

“… subjects, male and female, were rotated using the step method to progressively increase the speed of rotation (+2 rpm) after every 40 head movements to a maximum of 35 rpm. The end point for motion sickness was the Graybiel Malaise III total of symptoms short of frank nausea.”


Regardless, what I found interesting – and potentially practice-changing, was that meclizine stunk.  It was, in fact, the worst performing antihistamine of their trial:

What worked best?  Scopolamine plus amphetamine was far superior to any other monotherapy or combination therapy.  However, this isn’t a realistic treatment option.  From these data, at least, probably the best option may actually be promethazine (Phenergan).  And, for what it’s worth, promethazine also comes in suppository form – meaning it can be used as a rescue when the vomiting has already taken hold.

Probably any other sedating antihistamine has some value in meclizine’s absence, but I’m going with promethazine.  In fact, I’m not entirely certain I’ll ever go back ….

“Evaluation of sixteen anti-motion sickness drugs under controlled laboratory conditions.”

* This never happened.

The New “Standard of Care” for Appendicitis

Effectively, that is the question raised by this study – regarding antibiotics vs. surgery for acute, uncomplicated appendicitis.

This is a pragmatic, prospective study conducted in the pediatrics population at Nationwide Children’s Hospital.  The intervention and comparison were simple: qualifying patients with objectively mild appendicitis were offered a choice between hospitalization and intravenous antibiotics alone, or appendectomy with 12 hours.

And, as generally expected, not every patient choosing antibiotics successfully completed follow-up without crossover.  Of the 37 patients and families choosing antibiotics, 2 failed initial hospitalization, another 2 failed within 30 days, and 5 more failed within a year.  Median follow-up was 21 months at the time of article submission, and no further patients had undergone appendectomy.  Of 65 children undergoing surgery, 5 had post-operative complications, two of which were major (re-hospitalization, re-operation).

However, as I stated above, the question raised regarding antibiotics and appendicitis – is it now necessary it be discussed?  Have we reached a critical mass in the literature where all patients with suspected uncomplicated disease be offered antibiotics-only?  It is certainly unreasonable foundation for a complaint if an informed consent for surgical treatment of appendicitis did not include an antibiotics-only strategy as a legitimate alternative.  It was, in this cohort, much less disabling, in general, and substantially cheaper.

If not now – the not-to-distant future.

“Effectiveness of Patient Choice in Nonoperative vs Surgical Management of Pediatric Uncomplicated Acute Appendicitis”

The Disutility of Acetaminophen for Influenza

Wait … what?

This headline: “Acetaminophen Appears To Be Ineffective Against Flu.”

Hopefully, when you saw that either in the ACEP member daily news e-mail, or on the Reuters distribution, you had the same befuddled reaction as me.  Even better, the coverage includes this fascinating fearmongering from one of the authors:

“What this study does is raise some very serious questions about the real evidence base for using acetaminophen routinely for anyone that has the flu.”

Actually, it doesn’t.

This is a randomized, double-blind, placebo-controlled study of patients with suspected influenza, treating with five days of either scheduled acetaminophen or placebo.  The primary outcome was the area under the cure for quantitative PCR influenza log(10) viral load from baseline to day 5.  Patients were admitted to a clinical trials unit for the first 48 hours, then followed up in clinic at day 5 and day 14.  Secondary outcomes included symptom scores for health and temperatures of each patient.  The authors based their power calculation on the standard deviation for PCR viral load, and estimated to have 80% power to detect a difference greater than that standard deviation, they would need 80 patients with influenza.

Over the course of two influenza seasons, 2011 and 2012, they were able to enroll: 46

More specifically, they enrolled 46 patients with PCR-confirmed influenza.  The other 34 patients had false-positives on the rapid assay used for enrollment, and could not be included in their primary outcome analysis.  Because of their inadequate sample size, it is therefore mostly nonsensical to comment on their failure to find statistical differences in their outcomes.  However, there do not appear to be any potentially clinically important clues hiding in plain sight.

Bafflingly, one of their discussion points notes “These findings raise questions about the anti-pyretic efficacy of paracetamol in influenza and other respiratory infections.”  The mean temperature in the 80 patients randomized to placebo never even exceeded a 38°C “fever” threshold.  Then, the mean temperature was actually normal on day 2-5.  To suggest acetaminophen fails to normalize fever in respiratory infection compared with placebo, fever ought be relatively ubiquitous in the study population.  Otherwise, there simply isn’t any clinical feasibility of detecting a difference.

Finally, what makes the questions about the use of acetaminophen most bizarre, the study was actually undertaken to confirm the safety of acetaminophen.  Their clinical trial registration states:

“Paracetamol is recommended for the routine treatment of fever and systemic symptoms in influenza. However, this recommendation is contrary to evidence which suggests that the presence of fever is protective in infections. This study aims to investigate whether the regular use of an antipyretic (paracetamol) during acute infection with influenza may prolong viral shedding and clinical symptoms.”

Setting aside the inability of their sample size to actually address their outcomes – their conclusions are based on a failure to detect a difference.  Therefore, their original concerns over potential harms from the anti-pyretic nature of acetaminophen ought to have been allayed!

Quite a long way from the headline at the top of this post.

“Randomized controlled trial of the effect of regular paracetamol on influenza infection”

Nitrate Use in Aortic Stenosis Nonsense

This is another instance in which the “Editor’s Capsule Summary” and my interpretation diverge significantly:

“This study provides some support that nitrates do not cause relevant hypotension and may be judiciously used in patients with aortic stenosis and pulmonary edema.”

Rather, as was the consensus last night at OHSU’s journal club, these data provide essentially no support.

These authors provide a retrospective regarding the incidence of “clinically relevant hypotension”, identifying three groups of patients presenting with acute pulmonary edema – severe, moderate, and no aortic stenosis.  There were 65 patients in each cohort, and all patients received nitrates in either sublingual or intravenous form.  Regarding their primary composite outcome, there was no difference between arms.  Thus, the dogmantic avoidance of nitrates in aortic stenosis is challenged.

I am quite enthusiastic, as is Newman’s editorial, regarding the refutation of dogmatic (and erroneous) medical practice – he cites the example of the “dangers” of ketamine following head injury.  However, these retrospective data are not of appropriately high quality to inform practice in the absence of stronger prospective or randomized, controlled trial data.

The retrospective selection of patients only with prior discharges from the enrolling hospitals, as well as the inclusion of multiple presentations of some patients, likely amplifies selection bias with the effect of overstating the apparent safety of the intervention.  The study is powered to 80% to detect a 20% absolute difference between arms, which certainly exceeds the clinically relevant safety margin when comparing outcomes between cohorts.  The primary composite outcome for “clinically relevant hypotension” was mostly comprised of outcomes for which nitroglycerin needed to be terminated.  Termination of potentially helpful therapy is not a patient-oriented outcome or even a physiologic surrogate for one.  Sustained hypotension for >30 minutes is a better physiologic outcome for judging potential adverse effects of nitrate use, and, finally, intubation or mortality are better clinical outcomes.  The severe AS cohort had much higher incidence of these more serious outcomes, as opposed to the primary outcome in this study.  Finally, as a catastrophic oversight, these authors fail to provide any sort of control group – there’s no way to compare these patients’ outcomes to any similar cohorts managed without nitrates.

If you place this study in context with Ioannidis’ “Why Most Published Research Findings Are False” it is easy to see, rigorously, why these data should have little impact.  This study is underpowered, replete with bias, and the pre-study odds – for what “tradition” is worth – speak against the safety of nitrates in aortic stenosis.  Now, importantly, I am not saying you should not use nitrates in aortic stenosis – only, rather, doing so requires acknowledging the profoundly limited clinical evidence guiding such a strategy.

“Complications Associated With Nitrate Use in Patients Presenting With Acute Pulmonary Edema and Concomitant Moderate or Severe Aortic Stenosis”

Thanks to Ran Ran, Ben Sun, Kavita Gandhi and the other OHSU residents for their excellent contributions to the discussion!

Welcome to Yesterday, Have You Heard of PERC?

I usually like these sorts of articles regarding the yield and utilization of CT pulmonary angiograms.  They’re fun to dissect, useful to marvel at the inefficiency of our usage, and finally to feed my editorial hyperbole.  But, not this time.

This is a retrospective study from the University of Michigan comprising six months of CTPA data from 2013.  These authors reviewed charts on 602 consecutive patients and calculated modified Wells and PERC for each, and describe the appropriateness and yields of various cohorts.

Rather than detail these statistics and outcomes – other than to note their overall yield of 61 positives reported out of 602 scans – I’d rather just focus on the 108 patients scanned who were PERC negative.  PERC has been around since 2004, and it’s been percolating into various guidelines and evidence-based algorithms since.  Hello, it’s 2015: why are almost 20% of CTs at an academic medical center PERC-negative?

The authors state two PERC-negative patients had positive CT findings; given the pretest probability, I wouldn’t be surprised if one or both were ultimately false-positives.  Come on, man.

“CT Pulmonary Angiography: Using Decision Rules in the Emergency Department”

Hunting for Strokes in Vertigo

The vast majority of presentations for “dizziness” in the Emergency Department are benign diagnoses – various dysequilibrium syndromes, vestibulitis, neuritis, and other disorders of the otologic canalicular system.  But, then, some are strokes – and it’s quite challenging to balance diagnostic accuracy and MRI utilization.

This is an observational study performed at the University of Michigan evaluating patients presenting with complaint and exam findings consistent with acute dizziness – requiring either nystagmus or demonstrated gait instability.  Of these patients, all underwent a standardized neurologic examination, including HINTS examination, and subsequent MRI.  A stroke was the presumed diagnosis if an MRI was performed within 14 days and diagnostic of hemorrhage or acute infarction.

There were 320 patients enrolled, and 272 completed the clinical evaluation and an MRI.  Overall, 29 (10.7%) of patients had positive imaging – and, unfortunately, little was strongly predictive.  The few predictive features that shook out of their logistic regression model were the ABCD2 score, a positive HINTS exam, a central pattern to nystagmus, or the presence of other neurologic deficits.  Those individuals hitting all the risk factors ended up with ~20% risk of stroke, while those with none were <5%.  The remaining patients were simply in the intermediate risk group, reflecting the overall baseline level, and approximately one-third of the cohort fell into each category.

I do like their objective criteria for enrollment, based on a minimum of nystagmus or gait abnormality.  I tend to feel many patients receive MRI for similar complaints absent any of these features.  There is, unfortunately, no accounting of any general background rate of diagnosis of posterior circulation stroke at their institution, so there’s no way to estimate the miss rate or added value of their inclusion criteria.  I think their general observations are fair starting points for shared decision-making, although there’s still not quite enough information here to dramatically improve imaging yield.

“Stroke risk stratification in acute dizziness presentations: A prospective imaging-based study”

You Can Make Unnecessary Care Go Away

Low-value care is such a pervasive problem, ABIM developed the Choosing Wisely initiative.  However, Choosing Wisely is, unfortunately, a disengaged and toothless activity.  And it hasn’t worked.

But, as this study shows, you can eliminate unnecessary care with a more proactive and involved approach.

This is a quality improvement collaborative across 21 hospitals aimed at reducing the use of unnecessary or ineffective care relating to bronchiolitis.  As we’ve seen time and time again, if anything works at all for bronchiolitis, the ambiguity over its effectiveness probably means the effect size is clinically meaningless.  To this end, these hospitals banded together to deploy a QI program targeting reductions in bronchodilator use, steroid use, chest radiography, and other process measures.  Across all measures, the pre- and post-intervention measures demonstrated pooled meaningful and statistically significant improvement.  Bronchodilator use dropped from 46.2% to 32.7%, steroids from 10.9% to 2.2%, and CXRs from 12% to 6.7%.  A secondary effect of these interventions was a reduction in length of stay by 5 hours, from 49.6 to 44.6.


Of course, the QI intervention did not have the same effect at all participating hospitals.  Some, clearly, were on the ball, and almost entirely eliminated some unnecessary care (steroids).  Others, however, had no change from baseline, or, even, an increase – like two hospitals demonstrating 150%+ increase in bronchodilator use, and three hospitals with 100%+ increases in CXR use.  It would be interesting to see some qualitative analysis regarding the lack of improvement at certain hospitals.

But, in general, widespread improvement in unnecessary care can be realized.  In contrast to Choosing Wisely, it requires motivated agents of change and constant feedback.  The Choosing Wisely lists and their elements, unfortunately, seem adrift.

“A Multicenter Collaborative to Reduce Unnecessary Care in Inpatient Bronchiolitis”