Exercise · about twenty minutes
Reading a System-Drafted Note - Paediatrics
A twenty-minute exercise for clinicians
The second in a series. Each exercise puts one ambient-generated note in front of you, with four things wrong in it, and the same five questions to find them with.
Paediatrics is included early because it is where the failures are least forgiving. Almost nothing in a paediatric record means anything without its age, its weight, or its centile, and a system that carries the number without what qualifies it produces a note that reads correctly and dispenses incorrectly.
Nothing here requires paediatric practice. If you work with adults, read it anyway - the failures are the same ones, made visible by a setting that does not tolerate them.
Part 1
The note
Paediatric outpatient clinic. Follow-up.
Child attends with mother for review of recurrent wheeze. Two previous episodes in the past year, both associated with viral illness, both responding to inhaled salbutamol. No hospital admissions.
Mother reports the cough is worse at night. There is a history of eczema. Father has asthma.
Weight 14 kg. Temperature 36.8°C. Respiratory rate 28. Oxygen saturation 98% in air. Chest clear on auscultation, no wheeze audible today. No intercostal recession.
Impression: Asthma. Eczema, well controlled.
Plan: Continue salbutamol 100 micrograms as required. Commence inhaled corticosteroid. Amoxicillin 250 mg three times daily for five days. Provide spacer and demonstrate technique. Review in six weeks. Mother advised to return sooner if increased work of breathing.
Signed: ____________________
Read it once at the pace you would read it in clinic.
Then read it again, applying the five questions.
Part 2
The five questions
When reading any system-generated clinical output, do not ask is this good? Ask:
-
Is anything negated?
What is said to be absent, and could you tell absent from not assessed?
-
Is anything hedged?
Where is uncertainty expressed, and where has it quietly become fact?
-
Whose is it?
Does every finding belong to this patient?
-
When was it?
Is each statement about now, or about something in the past?
-
Which side?
Where laterality matters, is it there?
In paediatrics, one more, and it is not optional: what does this number mean at this age and this weight? A figure that is unremarkable in an adult may be seriously abnormal in a two-year-old, and a dose that is routine at 30 kg may be twice what a 14 kg child should receive.
Write down what you find before going further.
Part 3
The key
Do not read this until you have written your answers.
The age is missing, and everything depends on it
The note records a weight, a respiratory rate, a temperature and a saturation. It does not record the child's age anywhere.
Without it, not one of those numbers can be interpreted. A respiratory rate of 28 is normal in a two-year-old and abnormal in a ten-year-old. A weight of 14 kg is average at three and a serious concern at eight. Even the diagnosis shifts: recurrent viral wheeze in a toddler is a different clinical entity from asthma in a school-age child, with different treatment and a different prognosis.
What follows from it.
Age is the single most load-bearing variable in a paediatric record, and it is the one most often assumed rather than recorded - because everyone in the room could see the child. Nobody downstream can. A decision support rule reading this note has a weight and no denominator for it.
Certainty, converted - and the diagnosis changed with it
The history describes two episodes of viral-associated wheeze, both resolving with salbutamol, with no admissions. The impression reads asthma.
What the clinician said was that this pattern was suggestive and that a trial of inhaled corticosteroid would help establish it. That was a plan to find out. The note records it as a finding.
What follows from it.
In a young child this is not a subtle distinction. Recurrent viral wheeze frequently resolves; asthma is a register entry with recalls, repeat prescriptions, school documentation, insurance implications, and a label the child may carry into adulthood. It was created here by a system that heard a clinician thinking aloud and wrote down the conclusion they had not reached.
This is the most common failure in the series, and it does not become less common in paediatrics. It becomes more consequential.
Subject, transferred
The note reads there is a history of eczema. Father has asthma - and the impression lists eczema, well controlled.
The eczema is the child's. The asthma is the father's. Both statements are correct as written.
Now look at what the impression did with the first and not the second, and ask what would have happened had the system handled them the other way round. In the record this exercise is drawn from, it did: a parental diagnosis appeared in the child's problem list.
What follows from it.
Family history in paediatrics is clinically essential and structurally dangerous, because the people in the room are related and the conversation moves between them without announcing it. "There's asthma in the family" and "she has asthma" are one word apart in speech and a lifetime apart in a record.
A treatment with no indication in the note
Amoxicillin 250 mg three times daily for five days.
Nothing in this note supports an antibiotic. The temperature is normal, the saturations are normal, the chest is clear, and the impression names no infection.
Two possibilities, and both matter. Either the clinician prescribed it for something discussed and not documented, in which case the record cannot justify a prescription given to a child. Or the system generated a plausible paediatric prescription that was never issued, in which case the record now contains a treatment the child did not receive.
What follows from it.
Check the dose as well. At 14 kg, 250 mg three times daily is within the usual range for many indications and outside it for others - and no one downstream can tell, because the indication is absent and the age is absent. A prescription without an indication is not an incomplete entry. It is an unanswerable one.
Part 4
What to take away
Count what you found, and notice which you found last.
That is worth sitting with, because the antibiotic is the only one of the four a pharmacist would query. The other three pass every check the system runs, and they pass them because they are internally consistent.
Nothing in this note contradicts anything else in it. It is coherent, plausible, and missing the things that would let anyone else evaluate it.
Paediatrics makes this visible faster than adult practice does. A number without an age is obviously useless; a diagnosis without a certainty qualifier is not obviously anything, and it survives.
The five questions take seconds and require nobody's permission. In paediatrics, add the sixth, and ask it of every number you did not measure yourself.
One question for your institution
If a prescription appeared in a child's record with no documented indication, could your institution establish whether a clinician wrote it or a system drafted it?
The answer is diagnostic of a great deal else.
This exercise draws on Chapters 11 and 12 of Clinical Meaning in the Age of Health Data & AI, both published in full on this site. The clinical note is composite and illustrative, and is not drawn from any real patient record.
Clinical meaning is not generated by AI. It is captured, governed, and made computable.