Fighting Hubris in Medicine


The weekend started with a tweet about an elderly man with atrial fibrillation.  Atrial fibrillation is an arrhythmia of the heart that predisposes those who suffer with it to strokes.  The strokes are a  result of clots being thrown from the heart into the brain.  The typical treatment for this condition in those deemed high enough risk is to thin the blood to help prevent these clots from forming, and thus reducing the risk of stroke.

The problem with thinning the blood is that the risk of bleeding increases, and it does so especially as one advances in age.  It doesn’t help matters that the risk of having a stroke also increases with age.

In a 101 year old deciding on the best course of action is thus a challenging one.  It is easiest when patients are adamant about a certain path.  Far be it from me to tell a centenarian what to do.   In this case, the man who had been alive for two world wars chose to come off the anticoagulant he had been dutifully prescribed.

I queried the audience

Most leaned towards stopping Pradaxa, and some responded that there wasn’t a wrong answer.

I asked the question because this was a decision that had been made four years earlier.  The consequences of that decision were playing out now.

No one lives forever.  In a parallel world, the anticoagulant continues and this is a discussion about the brain hemorrhage that resulted.  These aren’t the only two possibilities. He could also have avoided both types of stroke, attended a few more great grand kids birthdays and passed away in his sleep from a cardiac arrhythmia.

Two hours have passed since the symptoms started, and a decision is made to not give  a systemic clot buster because of his advanced age.  No doubt weighing on the minds of the doctors caring for him is the risk of bleeding into stroke damaged brain tissue that is increased with old age.  He is rapidly transferred to a facility that can do a procedure that is a few years old at this point, but still new – fish out clot from an artery of the brain and avoid a stroke.  It had long been dogma that brain tissue, unlike cardiac tissue was not amenable to this treatment because the tissue was so much less forgiving.  Cardiac muscle can survive in the setting of an occluded artery for 60-90 minutes, brain tissue was thought to be irreversibly damaged after only 5 minutes of an occlusion.

But it turns out things are complicated, brain tissue death results when the the rate of blood flow is less than 10ml/100gm/minute of brain.  Blood clots that lodge in arteries may not initially completely occlude the artery, or the area at risk may have collateral blood flow that gives just enough flow to keep brain tissue alive and expand the window of time for an intervention to have success.   Ultimately, trials hit on the right patient population that may benefit from a mechanical ‘thrombectomy’ – those presenting with a significant neurologic deficit, within 6 hours, who had not completed an infarct on brain imaging.

It is a remarkable thing to behold – patients present paralyzed, unable to speak, and after removal of a serpiginous clot walk out of the hospital 48 hours later.  I queried the audience again

The discussion that followed was an interesting one, and I encourage readers to follow the various threads.

The boldest response, unsurprisingly, comes from John Mandrola, who took to his blog to address the case presented.

I tweeted the case initially as a question directed towards statisticians and decision experts who seem to spend careers writing about making better decisions by better analyzing data, but don’t actually make decisions.  The data has limits, even more so in the centenarian class of patients.  John, a guru of using evidence from trials to make decisions, refreshingly starts by agreeing with me that evidence has no place here.  Quantifying risks and harms are a fools mission when there is no data to be had on 100 year olds.. or 90 year olds for that matter.  I couldn’t agree more.

Things go sideways from there.

John goes on to admonish physicians for the hubris of believing they can help patients that are old. In John’s words:

“The struggle exposes our hubris. We are falsely confident that we control outcomes of a person who has lived for decades.

I strongly believe we do not control this person’s outcomes.”

This makes little sense generally.  The 101 year old man in question was independent, living alone at home.  If he developed a urinary tract infection, should we throw up our hands and say we can’t do anything because we don’t control his outcomes?

The first issue here relates to treating a very elderly person with an anticoagulant.  John, to his credit at least  takes a stand.  He says unequivocally: no anticoagulant.

Apparently, when there is true equipoise regarding a treatment and the treatment is associated with potential harm and added cost, the right answer is not to treat.  To support this statement he notes that there is no proven benefit of anticoagulant drugs in patients this old because a trial has never taken place in this group.  He also says there is a well known increased risk of harm related to bleeding, as well as the cost of the drug.

I guess we start worrying about the cost of the drug once patients get on the other side of 80?  And if there is no evidence of benefit of the drug because there isn’t a trial, how can we possibly know there are harms that may result without a trial?  Both contentions are silly.  A medication that thins the blood enough to cause bleeding will also thin the blood enough to prevent a stroke from a clot.  It is true that I would be moved by a trial in 100 year olds who were placed on blood thinners if 80% of them died hemorrhaging into their brain.  I would also be moved by a trial of blood thinners in this group if 80% of them died in a nursing home of sepsis after being paralyzed by a stroke. The point is – we don’t know.

Uncertainty means being unsure of harms or benefits.  John is sure of the harms, and also sure of no benefit.  The underlying ethos  is less is more.  I’m often a subscriber to this ethos when its clear the potential for harms are larger than the benefits.  This is not always so easy to fathom.

The decision with regards to thrombectomy is a complex one.  The stroke in progress involved the inferior branch of the middle cerebral artery.  This is some incredibly expensive real estate.  There are two major speech centers in the brain, one controls the spoken word, the other controls the ability to grasp the meaning of the spoken word.  The latter is a far more devastating sequela.  Where as the ability to comprehend but not speak is frustrating, the inability to comprehend speech renders a formerly functional individual function-less. In this case the area at risk was an infarct in the temporal lobe. An infarct here results in agitation and confusion.

Essentially, if this infarct completed, it was a death sentence.  A thrombectomy was a chance at life.  The CT head done showed no infarct.  The Alberta stroke program early CT score (ASPECT) was 10 – the highest possible score on a 10 point scale that suggested brain tissue had not yet begun to die.  The patient was confused, and agitated, breathing on his own, moving all his extremities.  There was no easy death here, no gentle passage into that dark night.  The stroke wasn’t big enough to kill, but large enough to maim.  After a discussion of the various outcomes with family, the patient went on to get a thrombectomy.   The procedure was technically a success.  Clot was retreived, the vessel opened.  It remains too early to know if a significant stroke was able to be averted.

Even if the outcome here is bad, is it really fair to claim hubris?

John ends by bemoaning my framing of the choice as thrombectomy or doing nothing.  To do nothing as I suggested isn’t the right term when we’re trying to sell that to patients and their families as a viable option.  It’s better to use the term palliation.  A wonderful group of doctors practice the art of  care directed towards symptom relief in patients with terminal conditions.  But the prognosis here is unclear. If a stroke is averted, what palliative care should be offered to the functional, independent 105 year old?  At the moment time is brain, and the minutes that tick by need to be spent deciding on the benefit the therapy may offer.  There is ample time for palliation if the decision is made not to intervene, or if the therapy fails to prevent a stroke.

Yet, hubris is the charge leveled from afar at clinicians at the bedside navigating treacherous waters to find their patients safe harbor.  The careless reader may think I am proposing an interventionist strategy for every 105 year old. I most definitely am not.  I am just as opposed to mindless action as I am to mindless inaction.  John may not realize it, but hubris afflicts those committed to inaction just as much as those committed to action.

Anish Koka is a cardiologist in private practice.  Trolling may be done on twitter @anish_koka

Categories: Uncategorized

3 replies »

  1. We don’t have to prove that this gentleman has had a great biochemistry spectacle going on for over a hundred years. He has proved by surviving that he is a superior being. He should be thought of as being far younger. He is younger. The hubris is in his physiology.

  2. First of all, pradaxa is not the best NOAC available for various reasons. Was it a carry-over for this gentleman from when it was the first best novel anti-coagulant?
    The first question in the case was one of all or nothing. No pradaxa was chosen.
    But what about half dose apixaban, as here?
    Sure, the evidence is not overwhelming, but it looks like a better bet than pradaxa or no pradaxa.
    Hubris is a difficult accusation. Perhaps it applies just as well to the decision to not look further for more appropriate treatment for the elder’s condition. The question put to him was too simplistic.
    To presume that presenting several treatment choices to an elderly patient amounts to abandonment is certainly hubris.