May 22, 2020 This Week in Cardiology Podcast

John M. Mandrola, MD


May 22, 2020

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.

In This Week’s Podcast

For the week ending May 22, 2020, John Mandrola, MD comments on the following news and features stories.

Brief COVID Update

When I recorded last week, there were 1.5 million cases in the United States. Today there 1.6 million. The rate of growth last week was 1.15x but is now down to 1.06x. United States deaths last week were 86,000 and today about 96,000; growth is about the same at 1.12x. That’s now about 5 weeks with these slight rises.

At my hospital in Louisville, Kentucky, we have seen a very slight uptick of COVID patients, almost exclusively from nursing homes, with very few patients in the ICU or on ventilators. Elective procedures have started and the hospital census is at about 70% —up from half empty. Some furloughed workers are back.

Globally there are now 5.1 million COVID cases with 2 million recovered. Last week I worried about Brazil, and that worry looks well placed, as Brazil now sits third in the world behind the United States and Russia. I also keep an eye on Sweden, with their more lax social distancing. Their COVID deaths of 380 per million is less than Spain, the United Kingdom, Italy, France, and Belgium. I still don’t know what to make of India, with its rate of death of only 2 per million.

COVID and Medical Education

My editor, Tricia Ward, had an excellent interview with two ID fellows from Harvard who are behind the YouTube COVID-19 literature updates. Drs. Eric Meyerowitz and Aaron Richterman have been doing this amazing 1-hour lecture series. They spend up to 80 hours reading, prepping, making slides and doing the presentation. Of course, the educational power of digital media is not exactly a new story, but what is new is that the social distancing policies of COVID have greatly accelerated the trend for free open access medical education or #FOAMed.

Other groups doing amazing COVID education are the FLARE group at Massachussetts General Hospital and Josh Farkas of PulmCrit blog and podcast. Statistician Professor Andrew Gelman often writes on COVID topics on his blog.

Collider Bias

Medscape Internal Medicine published an excellent summary by Prof Tim Morris from the University of Bristol. He and colleagues had a beautiful preprint on collider bias, which he called the banana peel of COVID-19 research.

Collider bias is complicated, but if you don’t want to be fooled by so-called paradoxs, it’s a super-important concept to learn.

In the obesity paradox, say for heart failure, researchers sample only patients admitted to the hospital for heart failure. The collider is something else that makes leaner patients admitted for heart failure sicker, for example, COPD, cancer, etc, and thus obesity looks protective. As Professor Morris explains, the same issue of non-random sampling of retrospective studies is running rampant in COVID-19 research. One study even found smoking was protective for COVID.

A brilliant study by Steven Stovitz is a good example. He and his colleagues purposely sampled a hospital database selectively to show how obesity in children can look like it protects kids from getting diabetes.

Watchman Registry

At ACC, the NCDR registry of approximately 38,000 Watchman implants over a 2-year period between 2016 and 2018 was presented and published. As background, in two pivotal RCTs (PROTECT and PREVAIL), Watchman did not meet non-inferiority to warfarin in the first primary endpoint (stroke, systemic embolism, or cardiovascular death) because ischemic strokes were higher in the Watchman arm. In the second co-primary endpoint, which excluded the 7 days after the procedure, the device barely met non-inferiority. Based on this dubious evidence, Watchman was approved for use in patients who had high stroke risk and a reason not to be on long-term oral anticoagulation. But since these patients were excluded from the trials, we have no RCT-level evidence.

Now to the registry. The data were collected by people in hospitals (abstracters) who are often off-site or working for third parties. While this is surely better than a voluntary registry like EWOLUTION, it’s hardly super reliable. For example, the authors write that data quality is assessed by random audit. But only 5% of the 495 hospitals were audited, and this found 93% were in agreement with the source document. What’s more Watchman complications can occur post hospitalization. Things like post-hospital stroke, bleeds, and device related thrombus will be missed by this paper.

The median HAS-BLED score was 3, which would predict a modest rate of 3.74 bleeds per 100 patient years—a risk level not usually considered high enough to warrant discontinuation of oral anticoagulation. But nearly a third of patients had HAS-BLED scores of 0-2. So, what was the contraindication to oral anticoagulation in these 10,000 patients?

On the surface, the paper’s report of 69% rate of prior bleeding looks reasonable. But only 12% of these were for intracranial bleeding; 42% were GI bleeds, but we have no breakout of these: did they require a transfusion, was it hemorrhoidal bleeding, was it a polyp that was cauterized? And... 6% of bleeds were nosebleeds; 15% were other.

Now to the “impressive” complication rate. Counting only in-hospital complications, those often recorded by an off-site abstracter, major complication rate was 2.1%. Remember, you have to balance that against the possibility of no benefit. In all, 74 patients died from the procedure—0.2%, but again, if a trial shows no benefit in this patient population, that would be terrible.

Final point, among operators, the median number of left atrial appendage occlusion procedures performed annually was 12, but most physicians perform less than 20 per year. This is American healthcare at its absolute worst. Most procedures are done by low-volume operators.

Ventricular Tachycardia

Dr Phil Cuculich, Dr. Cliff Robinson, and colleagues have published a five-patient case report in the New England Journal of Medicine, and a larger series in Circulation, called the ENCORE-VT study of 19 patients. At the Heart Rhythm Society meeting, Dr Cuculich presented longer-term follow-up of the patients they have done.

Of the patients in the study, 78% had a reduced VT burden (no shocks) at the 24-month follow-up compared with the 6 months prior to their procedure. The corresponding rate 6 months post procedure had been 94% in the study. Cuculich added that very few patients were actually cured of the VT but most had only small amounts of VT. At the end of two years, 52% of patients were alive. That seems like low survival but these are patients with advanced heart failure, most deaths attributable to heart failure, none related to the procedure.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.