The IKEA Effect in Medicine: Why Typing Is Not Caring

It is half past eight in the evening and the ward is quiet. The last visitor left an hour ago. A registrar is still at the workstation, typing. She has been there since the round ended, reconstructing in full prose what she already knows by heart, because somewhere she absorbed the lesson that a note only counts as careful if she wrote every word of it herself. Her colleagues will see the late hour on the timestamp and read it as dedication. What it actually records is a system that mistakes effort for diligence.

Behavioural economists have a name for the underlying error. The IKEA effect: we place a disproportionately high value on things we assembled ourselves. The wobbly shelf we built feels better than the sturdier one we could have bought, precisely because we suffered through the assembly. European medicine has a severe case of it where documentation is concerned.

We have come to treat manual entry as a proxy for conscientiousness. A physician who works from a template or corrects an automatically generated draft is quietly suspected of cutting corners. One who stays two hours late typing by hand is admired. We have confused the pain of the process with the quality of the product — and in a department under pressure, the typing rarely proves that someone cares. More often it proves only that the system is making them inefficient.

Two ways to spend a clinician's attention

Consider what is actually happening in the mind of someone generating text from scratch at the end of a shift. A large share of their attention goes to spelling, to formatting, to recalling the standard phrasing. They are looking at the keyboard, not at the logic of the case. Composition crowds out scrutiny.

Now consider the same clinician reading a structured draft assembled from the consultation and asking the questions that matter: does this capture the character of the pain the patient described; is the plan consistent with the latest results; is anything here simply wrong? Catching an error in a draft demands more clinical judgement than reproducing a familiar sentence from memory. The author attends to syntax. The editor attends to truth. The second is the harder and the more valuable of the two.

What we ask of an architect

We accept this division of labour without difficulty in other demanding professions. An architect does not demonstrate thoroughness by laying every brick personally. They demonstrate it by inspecting the structure rigorously against the plan. An architect who insisted on mixing the cement by hand, to show how much they cared, would not be praised for dedication. They would be relieved of the project. The judgement is the work; the bricklaying is not.

Ambient documentation tools invert the medical version of this only in who lays the bricks. The machine produces the draft. The physician performs the inspection. The expertise migrates to where it belongs — to deciding whether the record is true — and away from the manual reproduction of text that any competent typist could manage.

The signature does not attest labour

The objection one hears most often is a legal one. If software wrote the note, who is liable for it? This misreads where liability sits. You are not held to account for who pressed the keys. You are held to account for what the note says. The signature at the foot of the document attests to its accuracy, not to the hours spent producing it. Whether the text came from a voice model, a template or ten fingers is, in this respect, beside the point. The only question that carries weight is whether the physician read it and stood behind it as true.

Which leaves hospital leadership with a quieter standard to set than the one we have inherited. The careful note is not the one that cost the most keystrokes. It is the one that has been verified — read by a clinician who confirmed it describes the patient in front of them. That is the diligence worth rewarding, and it is not visible in a timestamp.