Anyway, I found what I was looking for; I had read an excerpt of a book called How Doctors Think, where the author discusses the availability heuristic:
In the theory, "availability" is defined as the tendency to judge the likelihood of explanation for an event by the ease with which relevant examples come to mind. In a clinical situation a diagnosis may be made because the physician often sees similar cases in his practice — for example, the misclassification of aspirin toxicity as a viral pneumonia, or the improper recognition of an essential tremor as delirium tremens due to alcohol withdrawal in an indigent urban setting. Groopman argues that a clinician will misattribute a general symptom as specific to a certain disease based on the frequency he encounters that disease in his practice.
Doctors are never going to be perfect; if you ask me following such a heuristic is probably going to work pretty well almost all of the time. As long as the doctor isn't a dick about it (with a god complex) and is willing to accept that they may be wrong..
In reference to my first comment: Mis-diagnosing "cool, rare diseases" as the flu by definition can only happen very rarely. Coupled with the fact that people are often mis-diagnosed with rare stuff, means that your original statement that "it is far more common..." is obviously untrue.
It must be frustrating if you are the person with the rare disease. But "not looking for zebras" serves the other 99.9% of people quite well..
You must be parsing my sentence differently than I am. Clarified:
It is far more common for a doctor to underdiagnose a patient that presents with flu systems but actually has the "cool, rare disease" [with the flu than it is to correctly diagnose them the first time].
In other words, within the ecosystem of "cool, rare diseases", it's far more likely to be mis- or under-diagnosed the first time than it is to be correctly diagnosed the first time. One reason for that is they're not common diseases (so the docs won't have a lot of experience diagnosing it), but another fascinating reason is that they have already sort of diagnosed you in their heads before getting into the data.
If you're interested in the topic, you should pick up the book I linked above. There's a fascinating study in there that tests this problem: the study takes two sets of identical x-rays with an obvious primary diagnosis like a cancerous tumor and in one set, they add another problem (blocked artery or torn heart valve or something else obviously serious and life-threatening). They give the x-rays to doctors, inform them of the primary diagnosis, have them look at the x-rays, and track their eye movements. Here's the crazy part: every one of the doctors lingered over that second, undisclosed problem, indicating they saw something wrong. Yet none of them reported the problem. (It's been a while so I might have gotten the specifics wrong, but the gist is there. It's really fascinating and I encourage you to read it on your own.)
On a personal note: take it easy with the "by definition" and "obviously untrue" stuff. I know this is the internet and all but it makes you seem hotheaded, especially when we might be misunderstanding each other.
You're right; I'm sorry, I should have given you the benefit of the doubt and parsed the sentence that way.
It is interesting about the 'bias' of doctors towards their own experience. I guess my point is that statistically this kind of 'bias' towards the most common syndromes is a net positive for society; it's an allocation of resources thing--we don't have so many doctors that we can afford them all to be running off on weird tangents investigating possible rare diseases that aren't there...
Again, sucks if you do happen to be the one in X thousand that has the rare disease. Am I being too utilitarian?
Regarding tone: apologies; I agree it was an overly flippant comment. It was more a reaction to some of the other doctor bashing that was going on in this thread that fails to see the woods for the trees. I see now that that wasn't your intention...
Anyway, I found what I was looking for; I had read an excerpt of a book called How Doctors Think, where the author discusses the availability heuristic:
In the theory, "availability" is defined as the tendency to judge the likelihood of explanation for an event by the ease with which relevant examples come to mind. In a clinical situation a diagnosis may be made because the physician often sees similar cases in his practice — for example, the misclassification of aspirin toxicity as a viral pneumonia, or the improper recognition of an essential tremor as delirium tremens due to alcohol withdrawal in an indigent urban setting. Groopman argues that a clinician will misattribute a general symptom as specific to a certain disease based on the frequency he encounters that disease in his practice.
http://en.wikipedia.org/wiki/How_Doctors_Think