Lighthouse Pearls
One insight in clinical reasoning, drawn from the book. A new pearl every Friday.

The chair is a diagnostic instrument.
The most underused instrument in medicine isn't in the supply room. It's the chair. Pull one to the bedside, sit at eye level, ask an open question, and wait through the silence. Patients hand a seated clinician the detail they never volunteered standing up — the medication they quietly stopped, the symptom they were embarrassed to mention, the thing that reframes the whole case. Most diagnostic error isn't a knowledge gap. It's attention, cut short. Sitting down is the cheapest countermeasure we have.

What would make this not what I think it is?
The first diagnosis arrives early, feels right, and quietly begins to defend itself. That sense of a finished story is the most dangerous moment in a workup. So ask the one question the story doesn't need in order to make sense: what would make this not what I think it is? A blood pressure in the other arm. Whether this pain resembles anything the patient has felt before — something that costs almost nothing and changes everything when the answer is the one you weren't expecting. Ask it especially when the case feels settled. Especially then.

The diagnosis you inherit is not the diagnosis you have earned. The diagnosis in motion does more than travel — it recruits. Each handoff adds confidence; each retelling removes doubt. Contradicting findings are absorbed rather than weighed, and the label crosses shifts and floors gathering weight without gathering examination. The questions that stop it are the ones momentum never asks: what came before arrival, and what has the label explained away rather than explained? Every handoff is a chance to reset. Before you carry a diagnosis forward, ask what the momentum has not yet asked — inheriting a diagnosis is not the same as earning it.

The discomfort of not knowing is a finding. There is a particular unease that arrives when you do not have the answer and feel that you should — in front of a patient, on rounds, at two in the morning. The instinct is to end it, and the fastest way to end it is to name something. Any plausible name will do, because the relief is the same whether the label is right or wrong. That is what makes the relief dangerous: it arrives identically in both cases. The discomfort is not a failure of knowledge. It is your mind reporting, accurately, that the story does not yet close — and that report is data. Stay in it a moment longer than the room requires. Say I don't know yet, out loud, and mean the yet. Hold the finding you cannot name intact, without smoothing it into something you already have a word for, until you reach someone who can test it. Somewhere inside that discomfort is the question you have not asked.

The note is where you catch yourself.

Documentation is the part of medicine everyone agrees is a burden. Some of it is. But the note was never only a record of thinking — for most of us it was the place the thinking actually happened. You write the pain is atypical for cardiac and something in you objects, faintly, before you can say why. You reach for the sentence that will make the fever fit the pneumonia, and the sentence will not come out right. That resistance is not a writing problem. It is your reasoning refusing to be smoothed.

Reading a draft and composing one are different acts. Composing is a decision. Reading is a search — and a fluent paragraph is far harder to argue with than a blank page, because it arrives already sounding correct. Whatever writes the note, the discipline has to survive somewhere: before you sign, say the assessment out loud in your own words, without looking. If it comes easily, you probably understand it. If you find yourself working hard to make it hold together, that is the finding.

The sentence resists because the reasoning underneath it does not hold.