Category: Quality

Computer Says: Discharge that Pulmonary Embolism!

We’ve learned a couple important things about pulmonary emboli for the past five or so years. First, we diagnose too many of them. Second, all pulmonary emboli do not need to be hospitalized. Knowing, as they say, is half the battle. That’s a start – but it’s not enough.

This study involves important thing number two above, the hospitalization of PE. Kaiser Permanente, in its endless quest for value, has already published several studies demonstrating the safety of discharging patients with PE. However, hidden in the descriptive statistics from those studies are the unfortunate still-low percentages of patients discharged.

In this prospective, multi-center, “convenience-assigned” trial, a computerized decision-support tool was rolled out to support risk-stratification for patients diagnosed with PE. Based on the pulmonary embolism severity index (PESI), patients scoring in class I or II were encouraged to be discharged, while those with higher scores were nudged towards hospitalization. In their pre-post design, little change occurred at the control hospitals, while the percentage of patients with PE discharged from the intervention hospitals jumped from 17.4% to 28.0%. No issues regarding untoward 5-day recidivism or 30-day adverse events were detected.

This is a great step forwards, and, frankly, one of the most prominent examples of decision-support being actually useful to implement practice change.   That said, in the intervention hospitals, there were “physician champions” associated with the roll-out of the CDS intervention, which almost certainly increased update.  Then, 41.2% of patients were PESI class I or II, so there’s even further room for improvement above these topline results – but this is an at least solid effort.

“Increasing Safe Outpatient Management of Emergency Department Patients With Pulmonary Embolism”

http://annals.org/aim/article-abstract/2714293/increasing-safe-outpatient-management-emergency-department-patients-pulmonary-embolism-controlled

Clinical Policy: Sanity Returns to ACS

This may be the most important recent sentence in modern emergency medicine:

“… based on limitations in diagnostic technology and the need to avoid the harms associated with false-positive test results, the committee based its recommendations on the assumption that the majority of patients and providers would agree that a missed diagnosis rate of 1% to 2% for 30-day MACE in NSTE ACS is acceptable.”

It’s no longer the domain of rogue podcasters and throwaway magazine editorialists to declare our zero-miss culture destructive and self-defeating – it’s finally spelled out in black & white by our speciality society. This is not a license to kill, of course, but it is now utterly reasonable to feel as though the wind is at your back when sending an appropriately-evaluated patient home.

This clinical policy statement does not address terribly many questions, but it does jam a lot of evidence into one document in their review. Specifically, these authors ask:

1. In adult patients without evidence of ST-elevation ACS, can initial risk stratification be used to predict a low rate of 30-day MACE?

In short, yes. These authors recommend HEART as their decision instrument du jour, but also acknowledge other scores that simply do not yet have enough diverse evidence to support their use. Interestingly, they also note clinical gestalt may be just as good as any decision instrument, at least when the ECG and troponin are negative for new ischemia. Again, more prospective evidence would be required to formally enshrine such a recommendation into a clinical policy statement.

2. In adult patients with suspected acute NSTE ACS, can troponin testing within 3 hours of ED presentation be used to predict a low rate of 30-day MACE?

Here the authors have only Level C recommendations, which means their recommendations are based on low levels of evidence. Overall, they are weakly in favor of using of high-sensitivity troponins alone, or repeat conventional troponin testing as part of a risk-stratification or accelerated diagnostic pathway.

3. In adult patients with suspected NSTE ACS in whom acute MI has been excluded, does further diagnostic testing (eg, provocative, stress test, computed tomography [CT] angiography) for ACS prior to discharge reduce 30-day MACE?

Please no: “Do not routinely use further diagnostic testing (coronary CT angiography, stress testing, myocardial perfusion imaging) prior to discharge in low-risk patients in whom acute MI has been ruled out to reduce 30-day MACE.”  Take that, CCTA proponents.  They give an expert consensus recommendation of 1 to 2 week primary care follow-up when feasible, or consideration of observation when no follow-up is possible.

The fourth question posed deals with use of P2Y12 and
glycoprotein IIb/IIIa inhibitors in the ED, and is met basically with a shrug.

So!  Go forth and provide good medical care – specifically, high-value medical care, further freed from the mental oubliette of zero-miss.

“Clinical Policy: Critical Issues in the Evaluation and Management of Emergency Department Patients With Suspected Non–ST-Elevation Acute Coronary Syndromes”
https://www.ncbi.nlm.nih.gov/pubmed/30342745

All Hail the Female Resuscitationist

There has been more than one instance recently of observed associations between female gender and improved outcomes. Female physicians have lower rates of 30-day mortality and readmission rates for hospitalized elderly, and have better outcomes among female patients with acute myocardial infarction.

Now, another set of data showing improved survival after in-hospital cardiac arrest.

This is a retrospective review of 1,082 in-hospital cardiac arrests between 2005 and 2017 in which the gender of the code team leader could be ascertained. The minority – 30.2% – were led by a female physician. Location within the hospital, shockable rhythm, time of day, and patient age were similar between the male and female physician-led cardiac arrest cohorts. With male physicians, ROSC was 71.7% and survival to discharge was 29.8%, bested by female physicians with 76.8% ROSC and 37.3% survival. In a sample size this small, there are many potentially unmeasured confounders regarding the underlying health and type of arrest that may have contributed to the baseline likelihood of ROSC and survival – but this is still quite the interesting association.

Unfortunately, this brief analysis cannot tweeze out specifically why the female physician-led cohort had better outcomes. Their data set recorded compression depth and rate, and these were effectively the same – but they do not have medication use, timing, and other relevant attributes for evaluation. They make some further associations between physician and nurse gender, but the confidence intervals simply explode regarding whether any observed survival advantage may have occurred by chance alone.  I expect other inpatient cardiac arrest registries or databases may have more granular data to either confirm or refute this association – and, hopefully, if such an association continues to be observed, to better determine the practice patterns associated with any increased survival.

Lastly, it is reasonable to be concerned regarding publication bias relating to these such reports of gender-based outcomes.  It is probably editorially more interesting – and certainly seems more likely to get picked up by the lay press – to report associations favoring the female gender than the other way around.  Perhaps a bit more research seems warranted before condemning men to the scrap heap of history.  I hope, for my own sake!

“Female Physician Leadership During Cardiopulmonary Resuscitation Is Associated With Improved Patient Outcomes”
https://journals.lww.com/ccmjournal/Abstract/onlinefirst/Female_Physician_Leadership_During_Cardiopulmonary.96124.aspx

Five-Stars is Bad Medicine

In modern medicine, the patient is the customer. Medical services are customer services. Measures of patient – nay, customer – satisfaction are tied to reimbursement and, by association, contracts and employment. We’ve often remarked this perceived or overt emphasis on satisfaction is an incentive for bad medicine – specifically the “Where’s my Z-pack variety?”, and this is one of the few studies to actually show such an effect.

These authors reviewed three years of data from their direct-to-consumer telemedicine program and assessed the correlation between receiving a 5-star patient rating and various physician-related features. There were 85 physicians included across 8,437 patient visits for respiratory tract complaints, mostly sinusitis, but also pharyngitis, bronchitis, and “other” categories. While adjusted ORs showed a variety of small associations with 5-star service just barely clearing statistical significance, there were clear ORs favoring those who gave out candy. Antibiotics were provided in 66% of all visits, and the aOR for a 5-star rating was 3.23 (2.67-3.91) as compared to no antibiotic, and a non-antibiotic prescription bestowed an aOR of 2.21 (1.80-2.71). No other aOR exceeded 1.30, except the “free coupon” visits at 1.58 (1.31-1.90). They also noted it was not possible to be in the 90th percentile for patient satisfaction unless you were basically in the top half of antibiotic prescribing.

There were a couple physicians who were above the 50th percentile for patient satisfaction while maintaining some semblance of antibiotic stewardship. The authors do not provide any qualitative evaluation of those physicians but – thank you good sirs, please share your wisdom with us all.

“Association Between Antibiotic Prescribing for Respiratory Tract Infections and Patient Satisfaction in Direct-to-Consumer Telemedicine”
https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2705078

So Many Strokes, Oddly

There a fairly steady stream of occasional articles describing the evolution of stroke care in the U.S. These are typically pieces praising improvements tied to “Get With The Guidlines-Stroke”, describing decreases in overall stroke mortality, and the like. “Never better!” would probably be how most neurologists describe the current state of stroke care.

These impressions might be a bit of a rose-colored view of the elephant. This is just a simple descriptive analysis in trends for stroke, TIA, and ICH care in U.S. emergency departments and hospitals from 2006 until 2014.  A couple things of curiosity and/or concern stand out:

  • Annual in-hospital mortality from stroke has declined from ~5.8% to ~4.4%. This looks good until it’s noted annual total stroke admissions increased from 353,000 to 415,000. So, in an absolute sense, mortality hasn’t changed much – we’ve probably just been adding additional cases that wouldn’t otherwise have been diagnosed as stroke.
  • Costs of hospitalization for all diagnoses have almost doubled. For stroke, hospitalization charges have risen from a mean of $27,000 to $48,000. I’d love to chalk up the cost increase solely to the accompany increased frequency of use of our favorite clot-buster, but the relative cost increases are similar for TIA, as well.

The overall gist I get from these data is the value, overall, of our care for acute neurologic emergencies is diminishing. I’m certain we’re doing a much better job of post-stroke care these days for those who would truly benefit, but clearly we’re also sinking a lot more money into an expanding population where the average benefit is probably lower.  It’s shaping up to be an interesting race to see which aspect of healthcare can bankrupt our economy first.

“National trends in stroke and TIA care in U.S. emergency departments and inpatient hospitalizations”

https://www.ajemjournal.com/article/S0735-6757(18)30648-X/fulltext

Urgent Cares (and Emergency Departments and Medical Offices) Are the Worst!

This small research article has been making the rounds in the news over the last couple days. In theory, these findings supposedly surprising and enlightening – although to anyone in medicine, or who follows this blog, they are hardly profound.

This is a simple retrospective, cohort analysis of the Truven Health MarketScan Commercial Claims and Encounters Database, which pools de-identified data from patients with employed-sponsored health insurance. In this study, they simply chopped up claims for office, urgent care, retail clinics, and emergency department visits. They publish rates of antibiotic use for various coded discharge diagnoses, again, chopped into categories of “antibiotic almost always indicated” (e.g., urinary tract infection), to “antibiotic may be indicated”, to “Antibiotic-inappropriate” (e.g., influenza, bronchitis).

The numbers get ugly in this latter category, and reflect least favorably on urgent care clinics. Rates of antibiotic prescribing for viral upper respiratory infection and bronchitis, for example, were 41.6% and 75.8%, respectively. This is obviously pathetic, and urgent care centers are rightfully taking heat for this, but neither the ED nor the medical offices deserve much credit, either. The ED was at 18.7% and 56.6%, and offices were at 29.9% and 73.1%, for viral URI and bronchitis, respectively. Retail clinics were not great, but certainly better, at 10.5% and 31.1%.

Of course, these are coded diagnoses and do not always fairly reflect the underlying clinical presentation or diagnosis. And then there’s this:

“We used facility codes but could not validate whether facilities were actually urgent care centers, retail clinics, EDs, or medical offices.”

When the crux of the study pits these different types of facilities against each other, that’s probably somewhat important.

“Comparison of Antibiotic Prescribing in Retail Clinics, Urgent Care Centers, Emergency Departments, and Traditional Ambulatory Care Settings in the United States”

https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2687524

tPA in Under 20 Minutes is Recklessness

In my book, “safe” translates to a lack of attributable harm. Therefore, going as fast as possible while still claiming safety – should mean no excess harms resulting from the rush.

There’s no way to precisely tell whether or not this is the case here in Helsinki, where the stroke neurologists have cut their door-to-needle time for thrombolysis to under 20 minutes. The results as described here, however, are not promising, and the authors agree with my impression:

“Our findings support the safety of highly optimized door-to-needle times.”

Ha ha! Of course they don’t.

This is a retrospective review of 1,015 stroke code patients arriving over a two-year period between 2013 and 2015. This institution, incorporating elements of pre-hospital assessment into their initial evaluation, have had door-to-needle times below 20 minutes since 2011. How do they perform?

Of the 1,015, there were 150 (14.8%) patients with misdiagnosis on the initial assessment. Of these, 90 were ultimately diagnosed with a stroke mimic, 59 were eventually diagnosed with a stroke or TIA, and one small basal ganglia hemorrhage was missed. These initial misdiagnoses led, as you might imagine, to both unnecessary treatment and delays to the correct treatment. The most profound effects of these delays were in the context of stroke mimics, whose median delay until a correct diagnosis was 39 hours. Thirteen stroke mimics received thrombolysis, and diagnostic inertia from the initial misdiagnosis led 13 more to have median delays of up to 56 hours for the initiation of condition-specific treatment.

Now, there are limitations here that likely tilt these statistics in favor of the institution. There is no described standard follow-up evaluation to confirm cerebral ischemia, and likely some of those with TIA (146 patients) or who received tPA (331 patients) and improved could further be lumped in with the stroke mimics based on their clinical evaluation and whether they ever underwent MRI. Conversely, even though these authors are speeding headlong in order to give tPA, we can’t actually attribute all these misdiagnoses to their rushed evaluation. It is likely some of these cases would remain clinically challenging, even with a few extra minutes of careful consideration.

However, if they are trying to prove their implementation is safe, this comparison group is exactly what is necessary. They’ve shown their protocol is results in a substantial number of misdiagnosis and documented patient harms; the onus is on this team to prove their pursuit of a handful fewer minutes to tPA is not a contributing factor.  Finally, any possible advantage to shaving a handful of minutes off door-to0-needle times pales in comparison to these obvious misses.

“Diagnosing cerebral ischemia with door-to-thrombolysis times below 20 minutes”
http://n.neurology.org/content/early/2018/07/11/WNL.0000000000005954

More Snapshots of Awful Antibiotic Use

Is there ever any good news these days? Geopolitical disasters, unwarranted pharmaceutical price increases – and physicians can’t even manage to get the evaluation for group A strep right.

This is a “successful” quality improvement paper wrapped around depressing and embarrassing data from a typical primary care pediatrics practice. These authors, primarily pediatric infectious disease specialists, were dismayed by the rate of guideline-non-compliant group A streptococcal testing and treatment in their group.

How bad?

The base rate of unnecessary GAS testing was 64% of all rapid strep tests performed. The base rate of inappropriate antibiotic prescribing – driven primarily by treating positive results in those who should never have been tested (e.g., likely non-pathogenic colonization) – was 49%.

After their multifaceted year-long intervention, they were able to achieve the amazing results of: 40% unnecessary testing … and the same, inappropriate 49% for antibiotic prescribing. When restricted to selection of antibiotic, at least, first-line antibiotics used 87% of the time.

Is this really the best we can possibly do, even after intent focus on practice improvement? And for a disease entitiy with such limited benefit for antibiotic in most modern settings?

“Improving Guideline-Based Streptococcal Pharyngitis Testing: A Quality Improvement Initiative”
http://pediatrics.aappublications.org/content/early/2018/06/18/peds.2017-2033

Comparing Unnecessary ED Chest Pain Testing

Since my last post regarding the disutility of coronary CT angiograms for the evaluation of acute chest pain in the Emergency Department was so popular, here’s more: a randomized trial testing CCTA versus stress echocardiogram.

The problem: no specific functional or anatomic testing is routinely necessary in the ED.

And, hidden in this comparison are broad results typical of testing in a “low to intermediate” population with a Diamond-Forrester pre-test probability of 28% and a TIMI score of 0 to 1. Effectively, it’s unimportant to describe their comparison because the incidence of their safety outcomes over a median follow-up of 733 days is so low its virtually impossible to detect a difference. Any Major Adverse Cardiovascular Event occurred in only 4.5% of all patients – but, restricted to the endpoints relevant to the testing performed, only 3.0% had a nonfatal myocardial infarction or cardiac arrest. Any differences in throughput and resource utilization between arms will be related to specific process and protocol implementation unique to the trial institution, so even their findings with respect to their primary outcome are not likely to be generalizable.

But, back to the futility of any test – only 10% of those enrolled were referred for cardiac catheterization, and only half of those received an intervention as a result. Another 10% or so received new or increased pharmacotherapy – likely overlapping with the population undergoing catheterization. Therefore, probably 85% of patients enrolled clearly received no specific benefit from these tests, 5% probably had some benefit, 5% were harmed (excess revascularization, serious complications), and 5% are equivocal.

And, all of this on the foundational premise these tests need to be performed widely, and in the ED. True disease is rare in this population, and the timeliness of diagnosis of CAD does not need to be made at the index visit. This is not high-value medical care.

“Coronary Computed Tomography Angiography Versus Stress Echocardiography in Acute Chest Pain: A Randomized Controlled Trial”

https://www.ncbi.nlm.nih.gov/pubmed/29909113

Treating Influenza with Antibiotics & Other Stories

Every time I review an article espousing the benefits of a protocol based on the use of procalcitonin to improve antibiotic stewardship, I usually say something along the lines of: “We don’t need this test, it only looks like we need it because our baseline antibiotic prescribing is hysterically shameful.”

Well, here’s another piece of evidence describing the basis for that statement.

This is a secondary analysis of observational data collected from the Influenza Vaccine Effectiveness Network. All patients were eligible for inclusion in the study if they presented with an acute cough of duration fewer than 7 days. Patients all received influenza testing as part of disease surveillance, as well as any other testing indicated.

Of 14,987 patient visits analyzed, 6,136 (41%) were associated with an antibiotic prescription. Of these, 2,494 patients (52%) received diagnoses for “potentially indicated” antibiotics – pharyngitis, sinusitis, and otitis media – while 2,522 (41%) fell into a category of “antibiotics not indicated” – viral upper respiratory infection, bronchitis, allergy or asthma, clinical influenza, or “other”. So, as far as the coded diagnosis is reliable, it is likely half of prescribed antibiotics are simply unnecessary.

Then, of the 14,987 analyzed, 3,381 had laboratory-confirmed influenza. Excluding those receiving a diagnosis of pneumonia, there were 945 who received a prescription for antibiotics. Finally, there were an estimated 860 patients with a diagnosis of pharyngitis and a negative test for Group A Strep, 327 (38%) of whom received antibiotics.

And, let’s not even get into whether patients received an appropriate narrow-spectrum antibiotic (44%).

There are limitations to the precision and clinical context of using diagnosis codes to classify antibiotic prescribing as appropriate or not, but these results are broadly consistent with the prior literature.  Before we start deferring our prescribing decisions to something like a PCT assay, there’s a huge opportunity to simply Do The Right Thing, first. Once the low-hanging fruit has been resolved, then we can worry about tweezing out the uncertain cases in a narrow cohort with potential limited application of PCT or other infectious disease differentiation engine.

“Outpatient Antibiotic Prescribing for Acute Respiratory Infections
During Influenza Seasons”
https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2683951