Please note that the text below is not a full transcript and has not been copyedited. For more insight and commentary on these stories, subscribe to the This Week in Cardiology podcast, download the Medscape app or subscribe on Apple Podcasts, Spotify, or your preferred podcast provider. This podcast is intended for healthcare professionals only.
In This Week’s Podcast
For the week ending June 9, 2023, John Mandrola, MD comments on the following news and features stories.
Impella Class I Recall
Abiomed has recalled nearly 500 of its Impella 5.5 left ventricular (LV) assist devices due to purge fluid leaks. I am no expert on the inner workings of the LV assist device, but it sounds like a severe issue because it can lead to loss of LV assistance.
The US Food and Drug Administration (FDA) labels it a Class I recall, which is the most serious type because of the potential to lead to injury and death. Abiomed has said they have received 179 complaints, including three patient injuries and no deaths.
Comments. The most common uses of Impella are in high risk percutaneous coronary intervention (PCI) and cardiogenic shock. It's only fair to say that many medical devices, and probably most medical device makers have had recalls.
What's notable about an Impella recall is that it's an unproven device in RCTs and its signals in observational studies are not reassuring.
The PROTECT II randomized controlled trial (RCT) tested Impella vs intra-aortic balloon pump (IABP) in high-risk PCI. The trial was stopped for futility and did not show a benefit in the primary endpoint at 30 days. Notable here is that the IABP has never been proven beneficial in an RCT.
The IMPRESS trial studied Impella vs IABP in cardiogenic shock after myocardial infarction (MI). There was no difference in mortality at 30 or 60 days. Bleeding rates were substantially higher in the Impella arm. In 2020, I reported on a University of California San Francisco-led observational study which found, after propensity matching, that the rate of death in the Impella arm was 45% vs 34% for IABP. The rates of major bleeding were 31% vs 16%, respectively.
At the same time, a Washington University-led group used an insurance database to report an ecological (meaning before and after) observational study that found no improvement in outcomes in hospitals with high vs low Impella use. The authors also found more adverse outcomes in the Impella era compared with the earlier time period: Death was 24% higher, bleeding 10% higher, and stroke 34% higher.
There are a number of larger ongoing trials testing this device. Particularly DanShock — which is a large European trial looking at Impella vs “conventional” support in patients with cardiogenic shock. The primary endpoint is all-cause mortality and it is scheduled to be completed January 2024.
That will be about 15 years after FDA clearance. Impella gained approval based on the much lower bar of 510(k) clearance. this doesn't make sense to me because my understanding is that this means the device had to be shown similarly equivalent to a marketed device. And 510(k) pathways are typically used for lower risk devices.
This is another area that tempts me to be cynical. The device is expensive. It's a money-maker for the company. US doctors earn extra productivity for its use. A PCI can be declared “high-risk” and an Impella placed. Boom. Extra money.
But I will close with a somewhat positive comment. Here's the thing with these invasive expensive technologies: I strongly suspect there are small numbers of patients that benefit from them. CS and truly high-risk PCI are tough things to study. Average effects may not reflect effects in specific patients.
So, there is a tension between having such a device available to use in highly selected patients by highly skilled, non-conflicted doctors or not approving it all, which may not be a good idea. I don't know the right answer. But lean strongly towards having re-imbursement tied to evidence generation. If there was such a thing, we would have known its average effects many years ago.
I have made Powerpoint slides taking the skeptical side of digital health. The seductive idea was that with enough monitoring of bodily functions we could achieve better health. But this belies the obvious definition of health as a state of complete physical, mental, and social well-being. The World Health Organization even adds the phrase that health is not merely the absence of disease.
The late Petr Skrabanek, in his wonderful book — free in PDF form on the Internet, the Death of Humane Medicine, writes that the search for health is a sign of unhealth. Rare is the person connected to a wearable health sensor who has complete physical and mental well-being.
The core problem with wearables is that they are made by companies interested in profit. And the numbers of healthy young consumers far outnumber older infirm people.
JAMA Network Open has published a nice observational study from a Yale-led group, documenting the patterns of wearables in adults.
The group specifically studied sociodemographic patterns and those with or without cardiovascular (CV) risk.
The data source is Health Information National Trends Survey, designed as a serial cross-sectional survey supported by the National Cancer Institute. It collects data on the knowledge, attitudes, and use of cancer- and health-related information.
This is a good use of observational data—Science tells us what we can do, trials tell us what we should do, and registries tell us what we are doing.
About 9000 participants were surveyed and these were believed to represent large swaths of the adult population.
In nationally weighted assessments, an estimated 3.6 million US adults with CV disease and 34.5 million at risk for CV disease used wearable devices compared with an estimated 29% of the overall US adult population.
These factors were associated with lower use of wearable devices:
Lower educational attainment;
Those with CV disease;
Those at risk for CV disease.
Among individuals with or at risk for CV disease, fewer than one in four use wearable devices, with only half of those reporting consistent daily use.
The authors second statement pushes the envelope a bit:
As wearable devices emerge as tools that can improve cardiovascular health, the current use patterns could exacerbate disparities unless there are strategies to ensure equitable adoption.
Comments. This is reasonable work to do and publish. I say it that way because it comes as no surprise that digital health is used less often in people who stand most to benefit from it. But without data, this would just be a podcaster assuming something was so. That's very low-level evidence. Observational data is good for telling us what is actually happening.
This data confirms that wearable sensors are being used by people least likely to benefit.
But that assumes wearables provide benefit over standard measures. And that is where I take some issue with the authors' second conclusion that says, “as devices emerge as tools that can improve CV health”
I am not sure that is a good assumption. Take atrial fibrillation (AF) screening with the absolute best device — the implanted loop recorder (ILR). The LOOP RCT could not show a significant benefit in stroke reduction. And a follow-up study found higher rates of bradycardia detection and pacers without differences in syncope or sudden cardiac death.
ILRs are always on. There are no better AF screening devices, and still, it couldn't pass muster in an RCT.
Thus, I would propose an alternate hypothesis: Perhaps the fact that wearable use is lower in people with lower socioeconomic status is not a causal factor in health disparities. Consider a counter-factual case wherein a policy idea was to provide smart-watches to people with lower socioeconomic status, worse CV health, and more risk factors.
Is the problem of worse CV health in disadvantaged urban and rural neighborhoods really a problem of smartwatch deficiency?
I don't think so. I think it is far upstream from that. And policies to improve wearable use in disadvantaged groups is a lot like spending energy getting a coronary artery calcium scan paid for in a 300-pound patient who still smokes.
Disparities in healthcare stem from much larger societal issues. Sadly, improvement lies far outside the realm of the medical professional.
The Lowest of Low-value Spending
JAMA has published an important observational study from a Johns Hopkins team of researchers. This study transcends cardiology into all of patient care. While it's an American study, it also applies to European healthcare.
The question: What are the costs to hospitals of measuring and reporting quality metric data?
My priors here are probably similar to yours: It costs a lot. Tons are spent complying with metrics that purport to measure the quality of care.
Before I tell you about this retrospective, time-driven, activity-based costing study, we should set out that everyone wants good quality healthcare. But, the moment I joined a practice more than two decades ago, I was stunned by the degree of dubious care delivered in the real world.
When I was young and trying to build a referral pattern, I longed for ways to distinguish myself on merit. but as I aged and lived through one foolish quality measure after another, I have to come to realize how hard it is to measure something that is so variable, so qualitative, so...impossible to measure, and that is good doctoring.
Yet, no one in health policy school seems to have read Goodhart's law, which states that when a measure becomes a target it ceases to be a good measure.
The most famous example of this is the story – perhaps apocryphal — is about Soviet factories given targets of numbers of nails produced. This led to factories making tons of tiny useless nails. Then when given nail targets based on weight, the factory made giant — also useless — nails.
The nanosecond you say that you are doing “risk-adjusted” outcome measures, hospitals hire an army of people to make sure every patient is coded as sick as possible. Dr. Mandrola, your patient has a low body mass index. Is she underweight? No. She is a retired Belgian bike racer who excelled at climbing mountain passes. That's why she's little.
Whenever you make atorvastatin use after a stroke a measure of quality, 90 year-olds who barely weigh their age in pounds, don't eat, and don't move are discharged on 80 mg of atorvastatin. By the way, the SPARCL trial excluded patients with AF-related stroke, so it's ludicrous to put post-AF-related stroke patients on a treatment for vascular disease because of a quality measure.
Whenever you make mammogram uptake a quality measure, women get breast imaging without a thorough understanding of the pros and cons. Remember that everyone cares about disparities of care and outcomes. Money spent on low value care is less to spend on helping the less fortunate.
The Johns Hopkins study set out to estimate the cost of data collection and reporting, and did not include any of the quality improvement projects.
This was a single center study. They identified 162 unique metrics. Pause there. Not 10, or 20, or even 100. 162!
60% were claims-based.
66% were outcome metrics.
62% related to patient safety.
Preparing and reporting required 108,000 person hours.
Personnel costs: $5 million in 2022 plus $602,000 in vendor fees.
Claims-based metrics cost $37,000 per metric per year.
Chart-abstracted metrics cost $33,000 per metric per year.
Significant resources are expended exclusively for quality reporting, and some methods of quality assessment are far more expensive than others.
Claims-based metrics were unexpectedly found to be the most resource intensive of all metric types.
“Policy makers should consider reducing the number of metrics and shifting to electronic metrics, when possible, to optimize resources spent in the overall pursuit of higher quality.”
Comments. One way to interpret these findings is that $5.6 million per year is only a fraction of the $2.4 billion budget at Johns Hopkins. But the authors urge us to extrapolate to the more than 4000 acute care hospitals in the United States. That means that billions of dollars are spent simply collecting and reporting data.
The authors also emphasize that even claims-based and chart-abstracted metrics are costly. Of course, they are.
Some may criticize me for covering this story, because it isn't directly clinical. How can working stiff doctors do anything about this? And I get that.
But I would argue that studies like this are important because they highlight important principles of evidence translation. That is, all of these metrics are well-intentioned.
Everyone wants quality, but imagine a world, a frame of reference, wherein everything that affected care had to be studied. You want to introduce five things that require extra work? Study it. Show me in an RCT that this improves outcome.
We don't do that. We do stuff that makes sense. Like ablating AF early. Like using LV assist devices in cardiogenic shock. Like adding on quality metrics.
Why not assume most stuff doesn't work and require it to be studied before it's implemented?
One reason, and another reason to cover this is that it is good for doctors to learn some basic economics. An important concept is rent-seeking. That is the behavior of people or institutions to gain profit without creating value.
The quality improvement industry, much like accreditation agencies, such as the American Board of Internal Medicine, are classic rent seekers. They extract wealth but cannot point to solid RCT-level evidence that they improve outcomes.
I dare you to Google “hospital readmission reduction program and unintended harm.”
Percutaneous Tricuspid Valve Interventions in Patients with Pacemakers
Michael Lloyd from Emory University gave a talk at the Heart Rhythm Society meeting about leadless percutaneous tricuspid valve (TCV) interventions. The first question is how much damage is a pacer lead to the TCV?
Some data suggest a worsening of tricuspid regurgitation (TR) with pacing leads. In fact, a late-breaking study from Dr. Leon comparing TR rates after subcutaneous implanted cardioverter-defibrillators (ICD) vs transvenous ICD — a silly comparison because subcutaneous ICDs don't deal with the TCV — but the analysis did show that the rate of any TR with TV-ICD is 42% and the rate of severe TR is about 7%.
I think the point is that, yes, sometimes, a lead through the TCV harms the valve. But recall that we have decades of evidence that standard devices improve outcomes, at least with the ICD and cardiac resynchronization therapy. And we don't have any comparative trial-level data suggesting that pacing or ICD systems that don't use right ventricular endocardial leads improve outcomes.
Nonetheless, the era of transcatheter intervening on the TCV is just beginning. I am worried. But my worry will not stop the march toward clipping and replacing that valve. It's already a IIb in the European Society of Cardiology guidelines.
Industry is doing direct to consumer advertising, raising awareness about fixing leaky TCV.
TRISCEND 1 (using the EVOQUE device) enrolled 56 patients total; one-third had jailed RV leads meaning a new valve was placed when there was an existing RV lead. At 30 days there were no lead issues.
In the VIVID registry 329 had percutaneous TVR; 31 patients had TCV leads.
Three underwent lead extraction before the valve was replaced.
28 had jailed leads with the valve and three of those had lead malfunction (10%).
Unpublished Emory experience:
Seven of 30 patients had transcatheter TV replacement (tTVR) with pacing leads, two with preemptive lead extraction.
Five had jailed leads. Three of the five had big lead issues; one patient died, two received urgent revisions.
Of the two of the five who did not have electrical issues, one died of valve dehiscence, and the lead was implicated.
Options. When the structural team says, yes, we need to do tTVR:
Leadless pacing before the device.
If there is pre-existing hardware, lead extraction first.
Triscupid sparing system: Put pacing leads into the coronary sinus, which avoids the tricuspid valve.
Comments. In an ideal world, electrophysiologists (Eps) and structural docs should be working together. I am surprised that at Emory, valves were placed with pacing leads in place and jailed. I can only assume this happened because no one talked with the EP.
At my shop, EPs and structural people talk all the time. Before and after TAVI, mitral clips, and now with TCV interventions.
The first take home is TALK.
Another take home is to consider that triscuspid interventions have a very dubious evidence base.
TR is common and it varies. Primary TR is rare. That is when TR is the only problem.
Most patients with TR have other problems – valve issues, heart failure, and other things— and when these other problems are corrected, TR lessens.
In other words, TR is often a sign of other things rather than a primary issue. You know, like AF.
The New England Journal of Medicine published TRILUMINATE and it was technically positive, but it is one of those cynic-inducing trials. Industry sponsored it and designed it to be positive.
The trial enrolled 350 patients with severe TR. One group gets the intervention, the other group doesn't. No sham.
Endpoint is a composite of many things — d eath, need for TCV surgery, heart failure hospitalizations, or quality of life (QOL) by Kansas City Cardiomyopathy Questionnaire (KCCQ).
No difference in any the hard outcomes, but positive on KCCQ. Again. QOL is important, but it belies science to measure a subjective endpoint when one group gets an invasive procedure and the other gets none.
I mention TRILUMINATE again because it's important — especially for young doctors — to remember the pivotal trials on how patterns are established.
Given the financial rewards to docs, hospitals, and the direct-to-consumer ads, the commonness of TR, the doctors' love of procedures, tTVR stands to grow into a common procedure. But we should always remember it was a broken study that got it all established.
The worry I have is: Were the fatal flaws of the TRILUMINATE trial (the choice of a QOL endpoint in an unblinded study) done knowingly or purposefully? Because if they were...oh man...I don't want to go there. I want to believe in science.
Yet again, the same tension of all overused lucrative procedures applies. That is, I bet there are rare patients with isolated TR that may benefit from the new valves. So for that rare patient, it is nice to have this technology.
But new valves don't get made to treat a rare patient. They get made to use in lots of patients. And boy, there are a lot of TR jets to find on echocardiograms.
© 2023 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: Jun 09, 2023 This Week in Cardiology Podcast - Medscape - Jun 09, 2023.