chatgpt prompts for dental practices: 20 prompts for patient comms, without ever pasting patient data

The short version: ChatGPT is really useful for a dental practice for recall reminders, treatment explanation leaflets, review request wording, and social posts, and a practice can save several hours a week once the prompts are built once and reused. The one rule that overrides everything else: never paste a real patient's name, date of birth, treatment history, or any identifying detail into ChatGPT, because that is a data protection breach under UK GDPR, not a shortcut.

Why patient data can never go into a prompt

This gets its own section first because it is the mistake that ends a practice's reputation, not just a bad piece of copy. ChatGPT is a third party outside your practice's data controls. The moment you type "write a letter to Mrs Johnson about her root canal follow-up on the 14th" you have moved identifiable patient information outside your clinical system into a tool with no signed data processing agreement covering that patient's record. Use placeholders instead: "[Patient Name]", "[Treatment Type]", "[Date]". Build the template with placeholders, then fill in the real details yourself, by hand, inside your own practice management software, never inside the chat window.

Prompts for recall and appointment reminders

• "Write a friendly six-monthly recall text message for a dental practice, under 160 characters, with a placeholder for [Patient First Name] and [Practice Phone Number]."
• "Write three versions of a 'we haven't seen you in a while' recall email for patients who have not booked a check-up in over a year, warm tone, no guilt-tripping language."
• "Write a short reminder text for a patient who has an appointment tomorrow, including a placeholder for [Time] and a line about what to bring if they have a new medical history."
• "Write a cancellation policy reminder to send 48 hours before an appointment, polite but clear about the practice's fee for late cancellation."

What to change each time: the tone (some practices want playful, most want reassuring), the channel (text has a character limit, email does not), and the placeholder fields. Never fill the placeholder in the chat itself.

Prompts for explaining treatments in plain English

• "Explain what a root canal treatment involves, in plain English, for a patient leaflet, avoiding jargon like 'endodontic' unless you also explain it in brackets."
• "Write a one-page explanation of the difference between a crown and a veneer, for a patient deciding between the two, in a reassuring, non-salesy tone."
• "Write an FAQ-style explanation of what to expect after a tooth extraction, covering the first 24 hours, what's normal, and what needs a call to the practice."
• "Explain the difference between NHS and private dental treatment in the UK in plain language, for a page on our practice website."

What to change: the exact procedure, the reading level (aim for a 12 to 14 year old reading age for anything patient-facing), and whether it is NHS or private context, since the two have different rules and costs.

Prompts for review requests and reputation

• "Write three short text messages asking a happy patient to leave a Google review, without offering any incentive, since incentivised reviews break Google's policy."
• "Write a polite response template to a negative online review about a long wait time, acknowledging the issue without admitting to anything clinical that could be used against the practice."
• "Write a thank-you message template to send after a patient leaves a five-star review."

Never use ChatGPT to write a response that discusses a specific patient's treatment, even if the reviewer named themselves. Keep every review response generic enough that it could apply to any patient, and check it with your indemnity provider's guidance if the review touches on a clinical complaint.

Prompts for practice marketing and social content

• "Write five social media captions promoting teeth whitening for a UK dental practice, without making any medical claim that isn't backed by evidence, and without mentioning a specific brand of product."
• "Write a short blog outline about the link between gum health and general health, for a general audience, referencing that this is general information and not personal advice."
• "Write a caption introducing a new hygienist joining the practice, warm and welcoming, under 80 words."
• "Suggest five FAQ topics that dental patients search for online, related to nervous patients and dental anxiety, that we could turn into short blog posts."

Prompts for internal practice communication

• "Write a short staff briefing note about a new hygiene protocol change, clear and to the point, for a team huddle."
• "Write a template email to send to a supplier chasing a late delivery of dental materials, firm but professional."
• "Draft a short script for reception staff to use when a patient calls asking about the cost of a private consultation, without quoting a specific price unless you fill that in yourself."

Prompts for training the wider team on safe use

• "Write a short internal policy note explaining to reception and nursing staff that placeholders must always replace real patient names and dates in any ChatGPT prompt."
• "Draft a one-page team briefing on the difference between a safe AI prompt and an unsafe one, with two clear examples of each for a dental practice."
• "Write a short reminder poster for the staff room summarising the practice's rule on never entering patient data into any AI tool."

Most practices already run some form of GDPR training, and this slots into that same structure rather than sitting apart from it. A short, clear internal rule, backed by two or three examples the team can picture, does far more than a long policy document nobody reads twice. Build it once, using ChatGPT to draft the wording, and revisit it whenever a new team member joins or a new AI tool gets introduced into the practice's workflow.

One weak prompt versus one strong prompt

Weak prompt: "Write a message to John Smith reminding him his crown fitting is on the 3rd and he owes 450 pounds." This puts a real name, a real date, and a financial detail about a specific patient straight into a third-party tool, which is exactly the data protection breach this article opened with.

Strong prompt: "Write a short, friendly appointment reminder template for a crown fitting appointment, with placeholders for [Patient First Name], [Appointment Date], and [Outstanding Balance if applicable], mentioning that the fee is due on the day." You then copy that template into your own practice software, fill in the real patient's details there, and send it from a system that has proper data protection in place. Same result, zero data risk.

Mistakes beyond the data risk

Do not let ChatGPT write anything that reads as a firm clinical promise, such as "this treatment will fix your pain" or "whitening will remove all staining." UK advertising rules for healthcare are strict about guarantees, and a generated line that overpromises can trigger a complaint to the Advertising Standards Authority.

Do not use ChatGPT-written clinical explanation copy without a dentist or hygienist checking it for accuracy. It writes fluently but can get a clinical detail slightly wrong, and a patient-facing leaflet with a factual error is worse than no leaflet at all.

Do not paste real complaint letters or safeguarding concerns into ChatGPT looking for a response, even with names removed, since the surrounding detail (a rare procedure, an unusual date, a specific staff member) can still make a patient identifiable.

I have written more around this on the site: How to Use AI for Email Marketing in Dental Practices, What Does AI Marketing Mean In Practice? A Working Definition With Real Examples, How to Use AI for Invoicing and Admin in Marketing Agencies.

This guide is part of my ChatGPT Prompts and Limits: 30 Guides to Fun Prompts, Photo Edits and Daily Caps.

For a quick check, run it through my tool that writes the prompt for you.

I publish contributed articles from dental professionals too. See how to write for this blog on dental.

Frequently asked questions

Is it ever safe to use a patient's first name in a ChatGPT prompt?

No. A first name combined with any other detail, such as a treatment or a date, can identify someone, and UK GDPR treats health data as a special category needing extra protection. Use "[Patient First Name]" as a placeholder every time and fill it in later inside your own system.

Can ChatGPT help write our practice's privacy notice or GDPR documentation?

It can draft a starting structure, but a practice's privacy notice needs review by someone who understands your actual data flows and your obligations under UK GDPR and the Data Protection Act. Treat any draft as a first pass for a professional to check, not a final document.

Can we use ChatGPT to summarise a patient's clinical notes?

No. Clinical notes are identifiable health records, and putting them into ChatGPT moves that data outside your practice's controlled systems. Any summarising of clinical notes should happen inside your practice management software, not a general AI chat tool.

How do we stop staff from accidentally pasting patient details into ChatGPT?

Build a short internal policy: templates only use placeholders, real details are filled in inside the practice system, and any AI-drafted patient communication gets a quick check before sending. A five-minute team briefing on this, backed by the placeholder habit, removes most of the risk.

Is ChatGPT reliable for answering clinical questions patients might ask on our website chatbot?

Treat it as a starting point for FAQ copy that a dentist has reviewed and approved, not as a live tool answering real patient questions unsupervised, since a wrong clinical answer given directly to a patient carries real risk.