Friday, September 10, 2010

How Doctors Doing their Best Can Get it Wrong

A woman comes to the ER feeling discomfort in her upper chest, and she's frightened. (Fear of death is what brings most people to ER, which in the USA is a bit like giving the doctor free reign with your credit card -- fear of which keeps most people away until life seems at stake.)

The doctor seeing her might get it right, or get it wrong. Let's pretend for the moment that it's actually possible to know the difference (in the heat of the moment, that's less often true than either doctors or patients realize, if only that we can't usually prove our diagnosis instantly). Let's also ignore for now the reasons doctors actually make a mistake -- I'm interested in what might cause a doctor actually doing things right to get it wrong.

Let's get back to our woman, Hilda. Let's pretend she's 40, has never smoked, hasn't bothered to get a cholesterol value 'cause she's so young, has just started a new job that's twice as hard, out of the gate, as she expected. Let's pretend that in her family there isn't anything interesting, like diabetes or breast cancer, except a maternal aunt who died suddenly at 55. She's too busy to exercise, and like most Packer-backers, is about fifty pounds over her best playing weight. The discomfort started while sitting at her desk at work. She's never had it before, or anything like it.

You are the doctor: what should be your priorities, given that you are having to see a dozen other people at the same time, have never met any of them before, and are severely tired?

Or, you are the patient: what do you imagine should come from this visit?

In 2010, we have basically two doctorly approaches: the old way is to try to understand the hidden reality, the unobservable process, that is inside Hilda, to make her feel as she does. The new way is to use Guidelines and Parameters to pigeonhole Hilda's symptoms, and our observations, into the rules that dictate what to do.

Why does this matter? It matters because one approach focuses intensely on the patient: Hilda; the other process focuses intensely on the guidebook. Often, Hilda feels sidelined by this. Always, the following-rules tactic causes more expensive testing than the figuring-her-out tactic. Sometimes, one approach fails the patient by forgetting the rules (which are, after all, carefully thought-out guides), or by ignoring the nuances of Hilda's personal experience and slotting round Hilda into a square diagnosis, to her hurt.

What are we doing when we assess and treat people, anyway?

Thirty-two years ago, a young colleague just slightly older than me said helpfully, "Patients want to know whether they are going to die. You have to tell them they're not." This seemed like a fine joke, and so I started telling every patient, when I told them what was happening to their body (if I could say it truthfully), "First of all, this isn't going to kill you."

I actually expected that some people would laugh. And, truthfully, a very few droll souls did. But it was rarely the delightful tinkling laughter of joy and glee; it was the wan chuckle of 'you're toying with me, aren't you?' What amazed me was how often the patient visibly relaxed, often while saying, "Well, I wasn't thinking that!"

Chastened, I kept that up, and expanded the idea to reassurance in general. After 3 decades of primary care practice, I feel my second-highest priority is to slay as many of the dragons in Hilda's closet as I can. If I, the experienced clinician, imagine the paramedic helicopter swooping in to rescue me and my hypochondriacally bad heart as I crest the hill on my bicycle, panting and in pain, what about the soul who hasn't had any medical training? The truth is rarely as bad as our fantastical fears; and even when it is something really, really bad, such as lung cancer, telling the truth actually reduces the number of dragons to just one. And we usually know how it behaves, which lets us give our patient a clear plan. Thenm, there may be grief, but much less fear or anxiety.

The first priority is to figure out what's actually going on in Hilda's innards that are making her feel the way she does. This takes some skill and knowledge:
1: Communication -- based on medical understanding
2: Forming a mental structure into which to fit the facts
3: Recognizing the difference between a fact and an idea: that is, forming a tentative diagnosis
4: Selecting observations and tests that will confirm or disprove this
5: Educating the patient on what to expect, and why.
6: Arranging what to do next.

During the last century or so, there have been essentially four paradigms for making a diagnosis. I am oversimplifying here
- The "I recognize one when I see one" tradition.
- The "differential diagnosis" tradition.
- The "pathophysiologic thinking" approach.
- The "Practice Guidelines" approach.

Each of these, even if followed faithfully, is prone to particular types of errors. How might these bring grief to Hilda?

More when I have time to write...

No comments:

Post a Comment