Thursday, October 14, 2010

The Medical Code

I spend a lot of my life thinking about language and the way that we use it. In my personal life, I'm a stickler for the precision of language; within my social group I have participated in originating several turns of phrase that cover concepts that are otherwise difficult to describe (in my social group the phrase for these phrases is "consensual language" -- i.e. language that means what it does merely because everyone in the social group consents to the definition.) On my walk to work on Saturday, I was thinking about how groups larger than my social group similarly, but less explicitly, co-opt certain words or phrases to inevitably express a specific idea.

This is particularly true in medicine -- both among doctors and between doctors and patients. Junior doctors present patients to an attending and are expected to use these phrases to paint their conclusions before stating what their opinion is. Similarly, despite the fact that taking a good history is a dying art, specific words that the patient uses can point to a differential. Perhaps the most famous example in all of medicine is the "Levine Sign" -- when a patient, while describing their chest pain, clenches their right fist closed over their mid to left heart (it is considered near-pathognomonic of angina.) In the past week, there were a couple of examples in my life that got me started thinking.

The first was also my very first role-reversal in the (misnamed) Attending Effect -- where a junior doctor, despite taking a good history, fails to be told a key piece by the family, which is then elicited by a more senior doctor (but not necessarily the Attending.) I'd sent my junior medical student to get a history from the family, while I took care of another patient. By the time I'd arrived, my medical student had spent at least 30 minutes very extensively getting the history and had done a very thorough job. Within 5 minutes of my arrival, I asked my first question: "When did your son get his Nissen fundoplication?" The father answered that it had occurred one month prior.

Afterwards, my medical student said to me, clearly frustrated: "I had just asked a past surgical history. The father swore his son had never had surgery. How did you know he had a Nissen?" And the answer was a single word: I'd walked in while the father was describing his cognitively normal 7 year old son "retching." There are only two things that I can think of that cause cognitively normal kids to retch without ever vomiting. The first is having a Nissen. The second is having a non-organic program (such as food aversion, attention seeking for some reason.) Given the rest of the history I'd heard in two minutes, the word "retch" was code for "Nissen" -- the father had essentially told me unasked that a Nissen had been placed, because I knew what to listen for.

The second story that led me to think about this was an event with an attending. I had admitted a little girl whom we suspected had epilepsy, who had been started on an anti-seizure medication, but was having adverse side effects from the med. In the course of picking a new medication, my attending became unconvinced that she even had epilepsy in the first place. Her EEG was negative. He asked me to repeat her semiology (clinical characteristics of a seizure). I started with "she has many types of events..." And he appropriately interrupted me, saying "so, she doesn't have epilepsy." I had used the code incorrectly (the code for seizures involves stereotyped movements, which should fall into one of few similar categories every time.) This time, rather than repeating word-for-word what I had been told, I instead made my case, using selective information: "yes, but one of the types involves being nonresponsive, with her eyes open, followed by rhythmic jerking, during which time she has bowel and bladder incontinence, and then is sleepy for half an hour afterwards." We started her on a new anti-seizure drug.

But perhaps the best story that I have on the code of medicine happened Saturday morning, when I arrived at work. We had admitted a new patient with "seizures" to the service. Having heard the story (by report, she was a teenager who had a history of having "jerks," which lasted seconds for years, but on the day prior to admission had had a "cluster" of "jerks," lasting minutes, which suggests a diagnosis of juvenile myoclonic epilepsy), I put in orders for an EEG and seizure precautions, and went to go see her.

The main differential diagnosis in a teenager who is having seizure-like activity is a new epilepsy disorder (and then there's a diagnostic dilemma within epilepsy, which is beyong the scope of this discussion) and pseudoseizures (also called non-epileptagenic epileptaform events, non-epileptic seizures, psychogenic-seizures and a host of other, confusing, names.) The way you tell the difference is usually with a good history and an EEG. An EEG is tough, because often, events don't get captured on EEG and then if the baseline of the EEG is abnormal, it's easy to call it epilepsy, but if the baseline is normal it's impossible to rule out epilepsy on interictal (between seizures)EEG. Therefore, a good history is the go-to tool in diagnosis.

Therefore, I started with a very careful history of what happened when she gets the jerks. She told me the jerks usually started in her head/neck, but sometimes spread down to her body. She described them as lasting seconds, but often occurring in clusters. All of that is a beautiful history for juvenile myoclonic epilepsy, and the beautiful thing about juvenile myoclonic epilepsy is that it's a type of epilepsy that will show up on EEG, even without a seizure, so I was getting pretty ready to call it a day for history taking. But then, as I was wrapping up the case in my head, she said "the movements begin with this compulsion to move." Carefully, I responded "'Compulsion' means something very specific to me, and I want to hear exactly what it means to you." She defined it: "I get this urge to move and it gets worse and worse and I can put it off, but then it overwhelms me and I have to move." And I realized that I'd also missed a huge piece of the history: she'd told me that right before her clusters started, she'd been playing in the band. Now, with a new diagnosis in hand, I asked "have you ever gotten a jerk while playing your musical instrument?" She hadn't. The diagnosis was clear. She had tics, not seizures. Tics are a different movement disorder, one which rarely gets confused with epilepsy, because the movements are typically not stereotyped and are always voluntary (not voluntary in the sense that the patient wants to do them, but voluntary in the medical code, meaning that the patient can delay doing them, and is giving in to an urge to move by choosing to move in the way that they feel the urge to do so.) The keyword in the whole case had been the "compulsion." I canceled the EEG and her seizure precautions.

We sent her home 90 minutes after admission (a new personal record), having done no diagnostic testing except a thorough history, which is why being fluent in the code is so important.