August 6, 2007

How Do You Treat Atrial Fibrillation?

It could be important...

Say you have a patient with a history of two heart attacks.  That's bad.  Now say he has atrial fibrillation.  That's worse.   Now let's say he is on no meds.

What do you do?  You'll treat two distinct problems: the history of ischemic heart disease gets a beta blocker, which has been shown to reduce mortality and improve quality of life.  A-fib gets amiodarone, the generally accepted treatment.   

Atrial fibrillation is a very fast but erratic twitching of the atrium.  While a-fib can be annoying-- palpitations, anxiety, etc-- it is a major cause of embolic stroke; the uncoordinated flapping of the atrium leads first to blood stasis in a less moving part, then flicking off of clots which happily go to your brain.  Hence, anticoagulants are common.

Beyond this, treatment of a-fib can be divided in two parts: rate control, with beta blockers, or rhythm control, with drugs like amiodarone.  The question is, which is better, and which is safer?  Traditionally, amiodarone was thought to be more efficacious.

Before I give you the answer, I want you to be aware of the form of the question. The question requires that you see a-fib as a distinct (disease) state, where risks and benefits of its treatment are considered.  It is, analogously, "what are we going to do about John's left arm?  It's a big jerk, always punching people.  Is what we do going to work?  Is it worth it?"

A study in the NEJM pitted sotalol, a beta blocker, against amiodarone for the effiacy and maintenance of a-fib.

What they found was, as expected, amiodarone was the best:


(The curves are read as "what are the chances you remain symptom free at x days?"  People are all symptom free on day 1, so the chances are 100%.  As time goes on, you get sicker.  So at day 600, you have a 10% chance of being symptom free on placebo, but a 70% chance on amiodarone. )

What was surprising is what's best if the patient has ischemic heart disease along with a-fib: 



If you have ischemic heart disease, sotalol was as good as amiodarone.  And, presumably, a beta blocker is less toxic than amiodarone (though the study did not show this.)

But there's a further point to be made here, about information bias.  Since beta blockers are already standard of care for any patient with ischemic heart disease, it doesn't actually matter whether they have chronic a-fib or not.    The history of ischemic heart disease is more important-- singularly more important-- than the a-fib.

So the guy above only gets a beta blocker.   Think about this. 

Whether he had "regular" ischemic heart disease, or ischemic disease and atrial fibrillation, the treatment is the same, single, drug. 

This is entirely the opposite of what happens in psychiatry.  In psychiatry, we pay lip service to a "disease state," but end up treating each disorder as if it was separate.  So a schizophrenic with negative symptoms (apathy, anhedonia, amotivation) gets an antipsychotic and an antidepressant, even though the antidepressant probably isn't necessary.  It may seem like both are necessary, but that's because you gave them both at the same time.  How can you know which was relevant med?  Bipolar is worse: a "mood stabilizer" and an antipsychotic and an antidepressant (or Lamictal)  each for a different symptom.  The nod to "disease state" however, is the dangerous one: all the meds are continued, even though each was started for a specific symptom that resolved.

And, this does not even begin to account for the overlapping pharmacology of many of the meds (like giving two antipsychotics for different "symptoms." )

In psychiatry, the incidental gets as much attention as the substantial.  But it is so hard for people to step outside themselves and say no to medication overkill.

This impulse towards polypharmacy- shotgun symptom management-- will never change unless there is a financial incentive, or, more accurately, negative reinforcement.  And I have one: give doctors a medication budget per day per person.