Dr. Santell's Rounds: Still My Conscience After All These Years

This isn't a story about Dr. Santell, but it illustrates why the memory of Dr. Santell is so important to me, and why sometimes I pretend he's still alive and only a phone call away. I know this is nothing more than childlike fantasy, but it gets me through some long days and--as this story shows--keeps me honest on the job.

Not that long ago I was at the end of a long run of night hospitalist shifts. "One guy is pretty sick," the day hospitalist told me at 7pm sign-out. "You know him, he came in with stroke symptoms but today he got completely obtunded and spiked a temp up to 101." A repeat CT of the brain showed evidence of progressive stroke on both hemispheres of the cerebellum. An earlier CT angiogram had shown extremely small-caliber basilar and vertebral arteries, and the patient's prognosis was considered pretty grim at that point, but nobody thought he'd decompensate so fast.

"Did you tap him?" I asked, referring to a lumbar puncture (LP) to sample cerebrospinal fluid (CSF).

"No. I wasn't thinking meningitis--"

"I mean encephalitis," I said. It was a long shot, but I knew this patient was relatively young, in his mid-fifties, with a relatively fast and unusual onset of symptoms.

"No," said the day hospitalist. "I'm pretty sure this is all due to a bad bilateral stroke." And I agreed with him. We decided I'd check on the patient later in the evening to see how he was doing.

I took over the pager responsibilities, and pretty soon I was swimming in four admissions. This kept me locked in the ER for several hours, but I had a nagging feeling about the stroke patient. He'd been in the hospital for several days, and only today developed fever, but what if he had encephalitis? I thought it was a highly unlikely diagnosis, but still--a relatively young person, a severe neurologic decline, what if we missed the diagnosis?

I did my admissions, but in the back of my mind I was gnawing on the possibilities of an encephalitis diagnosis. A neurologist had seen the patient earlier in the hospitalization, and several serial CT scans of the brain had demonstrated progressive stroke, but the fever was new and no CSF had been sampled during the entire hospitalization. Even if the patient did have encephalitis--and again, the presentation was extremely atypical and I believed it was a vanishingly small possibility--his neurologic decline boded a grim prognosis. So the outlook was bad no matter what I did.

One acceptable strategy would be to treat him empirically with acyclovir to see if he got better. A lot of people do this, especially if an LP is impossible or--let's be honest--inconvenient to perform. No one would fault me for starting acyclovir, except Dr. Santell.

Santell never wanted me to do anything without a good reason. He disapproved of shotgun management, in which multiple different therapies for multiple potential diagnoses are started at the same time. Doctors do this because, a lot of the time, they simply don't know what's wrong with a patient, and they don't want to miss anything, but Santell believed this was sloppy thinking. He taught his residents to follow a rational progress from clinical presentation, to diagnostic work-up, to therapeutic decision. If, as a resident, I had started acyclovir on a patient I hadn't worked up with an LP, he would have shaken his head in that devastating way he did whenever we'd let him down.

But Dr. Santell is dead, my lazy inner voice told me. He'll never know you didn't tap this guy. The competing thought gnawed at me as I tucked in my admissions. I talked to the patient's family, because the patient couldn't give informed consent. I emphasized the low likelihood of the diagnosis, but also the advantage of excluding a potentially devastating infection from the list of possible illnesses afflicting him. The family agreed it was a good idea.

So that was it. At eleven PM, with four admissions to dictate and a profound sense of discontent, I set up an LP tray in the ICU. I hadn't done an LP in several years, although I had placed intrathecal analgesia for laboring patients many, many times. Fortunately, the techniques are similar, and I got that LP as easy as frying an egg. The CSF went to the lab, and I went to the dictation room to rattle off H&Ps. I staggered out of the hospital at two AM, dreaming of other careers.

The CSF analysis did not show evidence of encephalitis and the patient sank into progressively worsening coma resulting from extensive bilateral strokes. He died several days later.

So what's my point? Why did I spend hours on a busy night shift deliberating over the decision to do an LP when most of my colleagues would not have faulted me for starting acyclovir? Because I couldn't stand the idea of Santell shaking his head at my failure to do what, in his eyes, was the right thing. Dead or alive, I stake my reputation on that man's opinion. 

Comments

[Bill]

I need to remember that story when I'm down on doctors, generally.

[sjdmd]

We all have that mentor we venerate. Whether living or dead, present or at great distance, the relationship continues and matters acutely to those of us still in the trenches in whatever form.

Thank you for sharing that story, T. You're all too human and real.