Doctoring ICD/Pacemaker Reflection

Unlike Watson, Human Docs are Influenced by Life

My friend called to tell me what happened. His voice cracked.

Jim thought that it was a normal Sunday morning ride. He was meeting his friends, and they were riding their normal route, on roads called Covered Bridge, Sleepy Hollow and Wolf Pen Branch. It was one of those Sunday mornings in Kentucky that makes cyclists feel as if they are in church. It had rained for days, but this Sunday morning there was an orange sun rising through the haze of the morning dew. Serene. Spiritual. All this.

For fifty-eight years, Jim had lived right. He was winding down a highly successful professional career. He was physically fit—in the normal way, not the obsessive bike racer way. He had plenty of friends and family. There were no significant medical problems. Jim could not be called inflamed.

Out of the blue, he told his mates that he didn’t feel well. And then, he fell off his bike and died—in seconds, on a serene Sunday morning with his lifelong friends at his side.

Jim died of sudden cardiac death (SCD). Sadly, this is how one in every five people find out they have heart disease. Jim was a physically-fit Tim Russert.

Sudden death is what heart rhythm doctors are trying to treat with implantable defibrillators (ICDs).  If Jim had had an ICD, he may still be riding those Sunday rides with his buds.

But Jim wouldn’t have met the CMS guidelines for an ICD.

I am not saying he should have had an ICD. His heart was normal–until that Sunday morning. No test exists that can predict when and if a human heart will fatally quiver. That’s the point: most patients who die of SCD don’t meet the government’s narrowly-defined ICD criteria. Heart doctors tasked with preventing sudden death know this fact all too well.

Jim highlights the human side of the problem that I would like the Department of Justice investigators to know about as they embark on their inquiry of ICDs. That is, when doctors are wrong about withholding an ICD the consequences are grave.

Doctors do not want to overuse ICDs. We know they have significant risks, and that they are very expensive.  Most good doctors strive to practice evidenced-based medicine. I do. Not just because the government mandates it, but because medical practice is best when backed up by science.

There are ICD guidelines: 40 days after a heart attack, not 39; 35% ejection fraction, not 36%; and 30 seconds of a rapid heart beat, not 29.


Randomized clinical trials.

All stuff that Watson understands well.

And then there are real people like Jim.



7 replies on “Unlike Watson, Human Docs are Influenced by Life”

Humans are continuous, not discrete variables. There is a relative risk, a hazard risk, for all of these data. Does clinical judgement prevail? That is why 70% confidence limits are more logical than 95.

I don’t get it, John. This case has nothing to do with a failure of guidelines. It has to do with a failure of our medical knowledge. Is there ANY scenario that you would seriously recommend that would have saved this guy’s life? Yes, this was a very sad event, and we all hope that medical knowledge will improve to the point that in the future this sort of event can be routinely prevented. But unfortunately we don’t have that ability today, so why kick the guidelines around? I think that will only set medicine back, rather than forward.


As always, you make excellent points.

It’s true, Jim’s death, like so many who die because of sudden cardiac death was not preventable–a lack of medical knowledge.

By using his example, I had hoped to make three points:

1. Though we squabble over how best to treat patients at high SCD risk (those with established heart disease), these represent only a small number of people that die suddenly. The tip of the iceberg. This fact is not well known.
2. The recent vigor with which guidelines get considered to be absolute, (black and white, right or wrong) stimulates me to emphasize the tremendous grey area in how best to risk stratify patients at SCD risk. Clinical trials help us when our patients are similar to the study cohort, but such is not the norm. Moreover, CMS guidelines differ substantially from the ACC/AHA/HRS recommendations. More grey area.
3. As humans, doctors are influenced by our life experiences. I am not saying that sad, unexpected, atypical cases should trump science-based practice. Rather, let’s just say that we we aren’t computers. Being wrong about an not recommending an ICD recommendation is quire serious.

Thanks for helping me clarify my points.

Your sadness for your friend is completely understandable, and your point that guidelines cannot be absolute is true. But — there are too many ICDs, not too few, and they’re a burden as often as they’re a lifesaver. They’re not a magic bullet, alas. Sorry about your friend.

Can ICDs be burdensome? Absolutely.
Do patents need to carefully selected? Yes.
Should a pre-ICD meeting where the patient comes to understand what to expect, including the possibility that the ICD becomes a burden happen? Yes.
Thanks for your comment.

CMS does not understand clinical studies; they make guidelines based on physician expert recommendations. Case in point: the dinamite and iRis studies that is used to say that ICD is not effective within the first 40 days did not require inducibility of VT if NSVT was present post MI. However the MUSTT and MADIT proved and showed the greatest benefit from ICD implantation if NSVT was present and sustained VT was induced at EP study. Dinamit was infact a neutral study it showed nothing. It left more unanswered questions, which is why MADI 1 and MUSTT is so important in this population. Where are our experts? MADIT II did not annul MADIT1 or MUSTT or did it? Scary? and why was this not part of the guidelines? Experts speak up.

Comments are closed.