More thoughts on the long tail

First, thanks for all the nice comments I have received on this topic.  Several other blogs have linked to this thread, and I have received comments and emails.

One comment deserves discussion –

so what do we do with the pharyngitis patient, negative rapid strep and culture, Centor score 1-2, calls back a week later “my throat still really hurts.” or “my throat hurts worse.” ?

They usually want you to call in antibiotics. Often are frustrated at being told to come back. “nothing’s changed, I was just there a week ago, why do I need to come back? Can’t you just treat it?”

Do you think we should empirically treat those not improving within a week, given your statements on natural history? Do they need to be seen again if “it’s still the same?” And what to do if they really don’t want to come back in?

This represents the true challenge in medicine.  First, thanks for the use of the eponym.  While it is somewhat embarassing, it also makes me smile.

I would probably examine the patient again.  The question here depends on other information.  Does the patient have viral symptoms?  Does the patient have a fever?  Is the patient a smoker?  Does the patient have other symptoms?

I have no easy answer for you conundrum.  I can list many possibilities.  Obviously, the patient might have a persistent viral syndrome.  The patient could have a bacterial infection – the British recognize a syndrome called PSTS (persistent sore throat syndrome.)  Fusobacterium necrophorum as the cause of recurrent sore throat : comparison of isolates from persistent sore throat syndrome and Lemierre’s disease.  Perhaps the patient has early Still’s disease.  Maybe the patient has allergies and post-nasal drip.  If the patient is of the right age, could they have infectious mononucleosis?

My point is not to diagnosis the problem, but rather to postulate how I might approach the problem.  Most times the patient is slowly recovering from a viral syndrome, but once the patient calls back, I would suggest that you re-examine the patient and keep a broad differential in mind.

While most of the patients will have "nothing", some will have significant disease.  Perhaps a repeat exam will give you clues to proceed further.  I doubt that you can figure this out over the phone or over the internet.

Entering the long tail does not imply that the patient definitely has a long tail diagnosis.  Rather, you enter the long tail when the probability of a long tail diagnosis has increased.  I would suggest that the patient calling back with persistent or worsening symptoms represents a good enough reason to consider long tail diagnoses.

I have obsessed over this question, because I believe it makes important points.  The patient should have improved in a week.  The patient is not acting as we would predict.  Thus, we need to re-examine our options.

Some patients will have "nothing."  Some patients will insist on antibiotics, even when they are not indicated.  Some patients will have nothing.  Our value comes in sorting out these groups.  Our biggest value lies in the long tail.  It separates us from algorithms.