Without evidence of benefit, an intervention should not be presumed to be beneficial or safe.

- Rogue Medic

Narrative Fallacy II

Also posted over at Paramedicine 101. Go check out the rest of what is there.

In the comments to Why Can’t Medics Resuscitate II, anonymous finishes up with this question.
 

One question. The sole purpose of Epi during arrest is to act as a peripheral vasoconstrictor, thus increasing preload – correct?

 

I cannot answer this question without writing about narrative fallacy.

How did we get to Narrative Fallacy II, without a Narrative Fallacy I?

I wrote Narrative Fallacy I before Paramedicine 101 was started. You can read it at Narrative Fallacy I. The brief explanation of narrative fallacy is that a narrative fallacy is something that takes the facts, or the information that may be factual, and tries to use this information to explain how that person believes the result is obtained.

Nassim Taleb gives an explanation in his book, The Black Swan.[1] You are given an ice cube and asked to predict what it will look like after sitting on a table at a temperature that will cause it to melt. Not a huge problem. The result will be some kind of puddle.

However, this is not what we are doing. We are presented with the result. From that result (a puddle), we attempt to predict what caused the puddle. We do not know about the ice cube. We do know that an ice cube is one possible cause of a puddle. We also know that many other things could lead to the puddle. This is part of the problem of narrative fallacy. We choose one of many possible explanations that we know about. We choose to ignore the explanations that we think do not fit. We choose to ignore the explanations that we do not know about.

In ten years, or twenty years, we will probably use a different explanation, because we will know more. Does it really matter which erroneous explanation we use, now? The only real use for these explanations is to create more hypotheses to test. The explanation may help in deciding what treatment to give next, although not for a protocol monkey just working his way down the algorithm. Then the explanation does not matter at all.

Some problems:

1. The facts may not be factual. If the result is misunderstood/misinterpreted, then basing the explanation/narrative on that misinformation is not likely to produce an understanding of how something works.

2.The explanation/narrative may fit the facts, but that does not mean that the explanation is the most appropriate for that set of facts. Just about any episode of House M.D. will demonstrate this. As you can clearly see, Dr. House looks up to me. Well, that is one explanation for the picture. It seems to fit the facts, but it doesn’t really fit the facts. After all Dr. House isn’t real, but someday I will be a real boy!
Think of any story with a plot twist. This is based on a narrative fallacy. You have read/watched the story. You have been led to believe that the facts mean one thing, but as the story progresses, you learn that this was due to a misinterpretation. Detective stories are often based on narrative fallacy. Dr. House is based on the fictional detective Sherlock Holmes, so it is no surprise that this is a dominant plot device on the show.

We recognize patterns. Narratives are ways of explaining what might have led to the pattern that we think we are looking at. The pattern may be due to information that we add, due to our biases. The pattern may be due to information that we leave out, because it does not fit with our biases. It is important to recognize patterns. It is important to have ways to evaluate the accuracy of patterns.

3.Now that we have an explanation, we extrapolate from that explanation as if it is a given. Since we all know blah blah blah. Therefore, blahblah blahblah blahblah. blah to blahblah, which everyone knows can be represented as blah2. Clearly, the misinformation is growing at an exponential rate. OK, maybe not so clear and maybe it is not even misinformation, but here is an explanation of narrative fallacy from some of the most accurate observers of narrative fallacy available.
 

[youtube]yp_l5ntikaU[/youtube]
 

If you have trouble understanding what they are saying, the full text is available here. If that doesn’t work (I am having some linking problems.), scroll to the top of the page and click on Scene 5.

So, why do witches burn?

We need to make sure that we are asking the right questions, rather than questions that reinforce our biases.

Were the villagers interested the truth?

No.

Later on, they would be even more resistant to evidence that the witch is not really a witch. After all, they did burn her, so they have to use some Cognitive DissonanceTM to justify their murder.

So, back to the original question:
 

One question. The sole purpose of Epi during arrest is to act as a peripheral vasoconstrictor, thus increasing preload – correct?

 

In the comments, Adam provides the textbook justification of the rationale for giving epinephrine in cardiac arrest. Of course, this is based on the idea that epinephrine works, or that epinephrine behaves as we believe it behaves. This may be true. Or this may be very misleading. Part of the explanation is a major part of why I think epinephrine makes things worse. Too much stimulus. Epinephrine is essentially a heart attack in a syringe. Therefore, it may not be the best choice of treatment for something that is most often the result of a heart attack.

How many of you would knowingly give epinephrine to a patient having a heart attack?

Why are we giving any drug in cardiac arrest? Only one reason – to improve survival to neurologically intact discharge from the hospital and a return to the kind of life the patient had before the cardiac arrest.

When we find a drug that seems to improve outcomes, we feel the need to explain why it helps.

I don’t care.

If the treatment helps, does it matter what the explanation is?

In ten years, or twenty years, we will probably use a different explanation, because we will know more. Does it really matter which erroneous explanation we use, now?

I don’t think so.

What are the chances that in a couple of decades we will be using the same explanation?

Minuscule.

Tiny.

Ridiculous.

Still, we feel this need to come up with some temporary erroneous explanation. Not only that, we criticize those who do not participate in this silliness.

Here is an introduction to an excellent video
 

[youtube]BDbuJtAiABA[/youtube]
 

The full video is over an hour long. If you enjoy it and understand it, the video is far too short.

Nassim Nicholas Taleb: The Future Has Always Been Crazier Than We Thought

Narrative Fallacy –

Narrative Fallacy I

How did this happen? – Research

Narrative Fallacy II

CAST and Narrative Fallacy

C A S T and Narrative Fallacy comment from Shaggy

Some Research Podcasting Comments

Shaggy Comments on Some Research Podcasting Comments.

Spine Immobilization in Penetrating Trauma: More Harm Than Good?

EMS EdUCast – Journal Club 2: Episode 43

Education Problems, Autism, and Vaccines

Footnotes:

[1] The Black Swan: The Impact of the Highly Improbable
By Nassim Nicholas Taleb

A must read book. If you have anything to do with risk management, then uncertainty/randomness/the unexpected are important parts of what you do. He deals with them better than anyone else. Too many misunderstand his writing, perhaps because they cannot abandon their own biases and accept their lack of control of events. While I find his prose to be awkward (perhaps he does not appear to be awkward, when compared to my writing, so maybe it is just me), his conclusions are essential to the understanding of risk management. Risk management people include any of us who treat patients.
Article about The Black Swan.

TM Cognitive Dissonance
Wikipedia
How to harm people with a clear conscience. Fool yourself.
Article

.

Trackbacks

  1. […] Wed, 15 Jul 2009 18:55:53 +0000 by Rogue Medic 7 Comments […]

  2. […] at explaining their results wander into narrative fallacy, which I have written about here, here, here, here, here, here, here, here, here, and […]