Healthcare Diabetic Foot Assessment Records: Use case and AI Confidentiality

A podiatrist who sees a person with diabetes and a foot ulcer writes a structured record: the ulcer, the nerve and circulation findings, a classification, the offloading and the plan. Then come the letters to the GP and the diabetes educator, and the record is pasted into ChatGPT or Claude to summarise it or draft them. A diabetic foot assessment record also holds the patient’s name, medicines, a smoking habit, money worries and a missed appointment. This page marks each item in an invented record that the Privacy Act, NSW health records law and the Podiatry Board’s health records guidelines bear on, and shows how the same letters are drafted on Ermos inside the practice.

The diabetic foot assessment record

Written at the first appointment for a person with diabetes and a foot ulcer, and updated at each weekly review. Letters go to the referring doctor and the diabetes educator, and the record is kept under the health records law of the State where the patient was treated.

Fernbank Podiatry
Level 1, 9 Thrush Lane, Cremorne NSW 2090 · (02) 5550 0164 · Diabetic foot assessment record · first visit Wednesday 11 November 2026 · last updated Friday 20 November

Patient
Referral
Diabetes
Nerves
Circulation
Ulcer
Footwear and home
Plan
Consent
Practitioner

An invented record. The patient is the same invented patient as the GP documents, the podiatrist, his practice and the hospital are fictional, the Medicare and registration numbers are placeholders and every clinical finding is made up. Hover or tap any highlighted item for the analysis; the full list is below.

Diabetic foot record drafter

An Agent+ that sets out the clinician's findings in your diabetic foot assessment record: wound details, test results and scores as the clinician entered them. At each weekly review it drafts the update and the letter to the referring doctor, noting what the patient asked not to share.

What it does

  1. Reads the clinician's findings and earlier entries
  2. Drafts the record update in your template
  3. Prepares the referrer letter and notes sharing limits

What stays with your people. The treating clinician checks every score and finding, decides care and issues the letters. The agent drafts only.

An Agent+ is built once and anyone in the practice can run it. It works from your own documents and runs on Ermos Dominion, the AI unit in your office, or on Ermos Edge, your own instance in an Australian data centre.

What leaves the practice

  1. 1
    Name, date of birth, address, mobile and Medicare number

    Identifiers that tie every entry below to one person. A business that provides a health service and holds health information is not a small business operator under the Privacy Act (s 6D(4)(b)), and the NSW Act lists podiatry services as a health service (s 4, para (j)). The Medicare number is a government related identifier, and APP 9.2 limits when an organisation may use or disclose one. Clause 3.3(a) of the shared Code of conduct asks practitioners to seek informed consent before disclosing information. Privacy Act guide.

  2. 2
    The referrer, a missed visit and a lost job

    Which doctor referred him, a photographed ulcer, a missed appointment and a job loss. The ulcer is health information (s 6FA(a)(i)); the missed visit and the job are other personal information collected in providing a health service (s 6FA(b)). HPP 11(1) and APP 6.1 govern passing them on. The Podiatry Board’s Guidelines on health records ask for comments that are respectful and in objective, unemotional language (2(i)), so a line about embarrassment is worded with care. Patient data guide.

  3. 3
    Diabetes, HbA1c, medicines and smoking

    A diagnosis, a result, two medicines and a smoking habit, with a note that he lives alone. Health information (s 6FA(a)(i)), and sensitive information under s 6(1). The Board’s guidelines list current and relevant past health history among what a record holds. An AI tool can misstate a dose or drop a medicine when it summarises, and the Ahpra guidance says practitioners must apply human judgment to any output of AI. Patient data guide.

  4. 4
    Monofilament, vibration and reflex findings

    The podiatrist’s own examination findings and the patient’s symptoms. The Board’s guidelines ask that a record is an accurate and complete reflection of the consultation (2(c)) and say practitioners using AI are responsible for checking the accuracy and relevance of AI-generated records (1(e)(iii)). A letter that rounds, reorders or interprets these findings is the podiatrist’s clinical statement once it is sent. Patient data guide.

  5. 5
    Pulses, toe-brachial index and Doppler waveforms

    Circulation findings and two index values. Health information about a named person. An AI tool asked to summarise can attach a meaning to a number that the podiatrist has not assessed, or place a value on the wrong foot. The interpretation belongs to the podiatrist, and the same accuracy duty in 1(e)(iii) of the guidelines applies. Patient data guide.

  6. 6
    The ulcer and its SINBAD classification

    Site, size, depth, tissue and a six-part classification. The 2021 Australian guideline on wound classification of diabetes-related foot ulcers recommends, as a minimum, SINBAD for communication among health professionals about the characteristics of the ulcer, and says the individual components rather than the total score should be used. It also recommends caution in applying any classification to give an individual prognosis. A letter that gives only the total, or an outlook the podiatrist has not given, departs from both points. Patient data guide.

  7. 7
    Shoes, money, walking and doubt

    Worn-out shoes he cannot replace, how far he walks and whether he will follow the advice. Other personal information collected in providing a health service (s 6FA(b)). It is the part of the record that explains why an ulcer may not heal, and an AI-drafted letter can assume shoes have been bought or walking has stopped. APP 6.1 and HPP 11(1) apply to passing it on. Patient data guide.

  8. 8
    Review dates and the referral threshold

    When he will be seen next and the point at which the podiatrist would refer him on. A clinical decision with dates, part of the record’s management plan and expected review, which the Board’s guidelines list among what is recorded (section 3). An AI tool should not move the threshold or the dates. Patient data guide.

  9. 9
    Consent, a sister and the missing mention of AI

    He agreed to the treatment, the photographs and two letters, and asked that nothing go to his sister. The Board’s guidelines say practitioners using AI must obtain and document informed consent where relevant and should understand whether patient data is used to train AI models (1(e)(i) and (v)), and clause 3.3(g) of the shared Code of conduct asks for written and informed consent before transmitting or sharing a person’s information. Ahpra’s guidance asks practitioners to inform patients about AI use. A prompt that includes the sister’s name could undo his request. Patient data guide.

  10. 10
    The podiatrist and his registration number

    The practitioner’s name and registration number, which are on the public register. Under the Board’s guidelines a treating practitioner must not delegate responsibility for the accuracy of information in the health record (2(l)), so a letter drafted from the record remains his.

What happens in Frontier Models (ChatGPT, Claude)

  • The whole document is transmitted to the provider’s servers. On personal plans the user cannot choose where it is stored or processed, and for the major providers that has generally meant outside Australia, under that country’s legal process. Some business plans offer storage in Australia, but the provider still receives the content.
  • On consumer tiers, prompts can be retained and used to train future models.
  • The provider’s staff and systems can reach the content under the provider’s own policies.
  • Once sent, the practice no longer controls it. Retention and deletion happen on the provider’s terms.

The same document on Ermos

Ermos Dominion

The document is drafted on a unit in your office, from your own templates and the patient record. Every marked item stays on the unit, which keeps working with the internet switched off. Nothing in it is transmitted to an AI provider, so there is no disclosure to analyse: the confidentiality work is the same as for your file server.

Ermos Edge

Your documents, the index and your access boundaries stay on your own instance in an Australian data centre. When a staff member asks for this draft, only their question and the excerpts they are permitted to see go to a named Australian provider, who is contractually barred from keeping the content or training any model on it. Nothing is processed outside Australia.

Frontier Models (ChatGPT, Claude)Ermos EdgeErmos Dominion
Where the document is processedProvider servers, generally offshore on personal plansYour instance in an Australian data centre; answers generated in AustraliaThe unit in your office
Who can read itProvider staff and systems, under policyYour staff, within the boundaries you set. The named Australian provider processes only permitted excerpts, under contract, and keeps nothingYour staff only
Is it keptOn the provider’s termsKept on your instance; the provider is barred from retaining itStays on the unit
Can it train a modelOn consumer tiers, yesNo, by contractOnly your own unit learns from it, on your hardware

Part of Use cases in healthcare practices. See how this works in practice: Clinical documentation on Ermos.

Common questions

Do ChatGPT and Claude use the information in a diabetic foot assessment record?
Yes, a pasted record is sent to the AI company's servers and processed there to produce a reply. As at October 2026, whether a conversation on a personal plan (ChatGPT Free, Plus or Pro; Claude Free, Pro or Max) can be used to train the models depends on a setting each user controls. Business plans (ChatGPT Business and Enterprise, Claude Team and Enterprise) and the developer APIs sit under commercial terms, and OpenAI and Anthropic each say they do not train on that content by default (OpenAI, Anthropic). On either product, the thumbs up and thumbs down buttons can send a whole conversation back for training. This record carries a heel ulcer photographed for the record, an HbA1c of 9.1%, a lost job and a request that nothing go to the patient's sister.
Is it safe to paste a diabetic foot record into ChatGPT or Claude if we remove the patient's name?
No, because the rest of the record still points to one man. Removing names rarely makes a document anonymous. Information stays personal information while the person is reasonably identifiable, and the OAIC warns that information can be at risk of re-identification even when it has been de-identified or anonymised. An Artarmon address, a Medicare number, a heel ulcer first seen on 22 October and a lost depot job would let anyone who knows him pick him out. His diabetes and foot findings are health information, which is sensitive information under the Privacy Act. The Office of the Australian Information Commissioner (OAIC) recommends in its guidance on commercially available AI products that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools, because of the significant and complex privacy risks.
Can ChatGPT or Claude be trusted to draft a letter from a diabetic foot assessment record?
Not without the podiatrist checking every line. Language models can produce fluent text with details that are not in the source, wrong figures, wrong dates or references that look real but are not. In this record the most harmful errors would be a monofilament result given for the wrong number of sites, a toe-brachial index placed on the wrong foot, a changed ulcer size or SINBAD component, or an altered medicine dose. The podiatrist should check each one against the record before a letter goes to the GP or the diabetes educator.
How can a podiatry practice use AI for a diabetic foot assessment record without sending it to ChatGPT or Claude?
Ermos is built so the document stays under your control. On the Dominion AI unit it is drafted inside your office and none of it is sent to an AI provider. On Ermos Edge your documents and index stay on your own instance in an Australian data centre, and only the question and the excerpts the user is permitted to see go to a named Australian provider that is contractually barred from keeping them or training any model on them. The letters are drafted from the patient's record and the practice's own letter template. The podiatrist checks every finding against the record and signs each letter.
Do podiatrists have to follow the Privacy Act and NSW health records law?
Generally yes. A business that provides a health service and holds health information is not a small business operator under section 6D(4)(b) of the Privacy Act, so its turnover does not take it outside the Act. The NSW Health Records and Information Privacy Act 2002 lists podiatry services as a health service in section 4. Under section 25(1) a private sector health service provider must retain health information collected while the person was an adult for 7 years from the last occasion it provided a health service, and information collected while the person was under 18 until they turn 25.
What does the Podiatry Board say about AI in health records?
The Board's Guidelines on health records, listed on its website as dated 28 July 2026, say practitioners using AI must obtain and document informed consent where relevant, involve patients in decisions where their personal data is required, check the accuracy and relevance of AI-generated records and comply with privacy and health records law. They should also understand whether patient data is used to train AI models.
How should a foot ulcer be described in a letter to the GP?
The 2021 Australian guideline on wound classification of diabetes-related foot ulcers recommends that, as a minimum, health professionals use the SINBAD system to communicate the characteristics of an ulcer, and says the individual components rather than the total score should be used. It also recommends caution in using any classification to give an individual prognosis. The letter should rest on the podiatrist's own examination.

See Dominion answering from your own documents.

A demonstration takes 30 minutes, uses no client data, and comes with no obligation.

Book a demo