AI at the practice, without breaking medical secrecy
Save hours on letters, certificates and reports, yes, but never at the cost of a patient’s data in an AI with no guarantees.
Medicine, dentistry, physiotherapy, psychology: every consultation produces health data, which Swiss law classes as sensitive. Dictated or pasted into consumer AI without a data processing agreement, it leaves your control; medical secrecy still holds you responsible for it.
Your team already uses AI, at the front desk and in consultation. The point is not to ban it, but to draw the line between what the tool may see and what it never will.
- Front-desk hours lost to reminders, certificates, reports and letters to colleagues
- Extracts from patient records dictated or pasted into public AI, beyond your control
- A reply to an online review that confirms, in one word too many, that someone is in your care
- A generated certificate or letter that states more than the record shows, under your signature
See for yourself
A prompt structured in 5 blocks: Instruction · Role and audience · Data · Constraints · Format. An unframed AI invents; the frame stops it.
In a workshop, we adapt these prompts to your practice, each with its safeguards: no patient identity, no health data ever enters a public AI.
Rappel de rendez-vous · le message qui n’en dit pas trop
A reminder that gives the time without revealing the reason or the specialty, even if read by a third party.
See it in Chaptal →Avis en ligne · répondre publiquement sans rien révéler
A courteous public reply that never confirms that someone is your patient.
See it in Chaptal →Facture impayée · relancer sans abîmer la relation de soins
A payment reminder firm on the amount, silent on the care, preserving the relationship.
See it in Chaptal →30 use cases cover your line of work in the Chaptal catalog, out of 750 in the corpus · each one states what it produces, what it expects from you, and what it refuses to do.
« Le Collaborateur, augmenté »
Your team learns to frame AI on its real tasks, reminders, letters, certificates, review replies, without exposing a patient.
See the training →A written frame for your health data
An inventory of the practice’s AI uses, a register, clear rules: who uses what, on which data, within which limits.
See audit & advisory →Foire aux questions
Can I dictate my consultations to a transcription AI?
Not into just any tool. A consultation transcript is health data under the nLPD; it requires a data processing agreement, hosting guarantees and informing the patient. The dedicated card in the catalog sets this frame before the first try.
What does the practice risk if patient data enters a public AI?
Whatever goes in leaves your control, and medical secrecy remains engaged even without intent. The rule is simple: anonymize first, or keep it out. Every card in the catalog repeats this safeguard.
Can AI write my medical certificates?
It can structure the document and its wording; the findings remain yours. A certificate that states more than the record shows becomes contestable, and your signature answers for it.
Can I answer online reviews with AI?
Yes, for tone and form. The limit lies elsewhere: a public reply never confirms that someone is your patient, even when the review says so. The framed prompt holds that line for you.
My assistant already uses a public AI. Should I ban it?
No, frame it. A ban pushes these uses out of your sight; written rules state what enters the tool, what never does, and how to handle the rest. That is what the training is for.