Aged Care Quality Indicators Returns: Use case and AI Confidentiality
Every quarter a home collects its quality indicators and submits them to the government, and the work is done in a spreadsheet: counts, eligible residents, percentages and a column of notes. It is exactly the kind of file a clinical manager pastes into ChatGPT or Claude to check a formula, compare quarters or explain a movement to the board. The submitted figures are aggregates, but the working sheet is not. It carries room numbers, a resident’s restrictive practice, a fracture, an incident register and survey comments. This page marks every item in a realistic indicators workbook, shows what the Aged Care Act 2024 and the Privacy Act bear on, and shows how the same sheet is checked on Ermos.
The quality indicators return
Built by the facility manager at the end of each quarter and submitted in the government portal by the 21st of the following month, with working papers kept in the same workbook.
Quarterly quality indicators return
Kurrajong Gardens Aged Care · Quarter 1 July to 30 September 2026 · 60 beds · Prepared by L. Keane
| A | B | C | D | E | F | G | H | |
|---|---|---|---|---|---|---|---|---|
| 1 | Indicator | Program measure8 | Residents counted | Eligible residents | Program value (%) | Events in incident register | Events per 1,000 OBD9 | Notes |
| 2 | Occupied bed days, 1 Jul to 30 Sep 2026 (92 days)1 | 5,152 | Average 56 residents. Quarterly Financial Report figure. | |||||
| 3 | Pressure injuries | One or more pressure injuries | 3 | 58 | 5.2 | 4 | 0.78 | 9A, 12B and 13A. Agnes Thornbury (14B) stage 2 sacrum noted 12 Oct: next quarter.2 |
| 4 | Restrictive practices | Residents with a restrictive practice | 5 | 58 | 8.6 | - | - | 14B risperidone (chemical), guardian consent on file. 9A bed rails. Three others on the Residents tab.3 |
| 5 | Unplanned weight loss | Significant | 4 | 49 | 8.2 | - | - | Nine residents excluded (palliative or unable to be weighed). |
| 6 | Unplanned weight loss | Consecutive | 2 | 49 | 4.1 | - | - | |
| 7 | Falls and major injury | Falls (total) | 12 | 58 | 20.7 | 19 | 3.69 | 19 falls by 12 residents. 14B twice. |
| 8 | Falls and major injury | Falls with major injury | 2 | 58 | 3.4 | 2 | 0.39 | 10B fall 14 Aug, wrist fracture, St Hilda’s. 6A fall 2 Sep, hip fracture, surgery.4 |
| 9 | Medication management | Polypharmacy | 21 | 58 | 36.2 | - | - | |
| 10 | Medication management | Antipsychotics | 6 | 58 | 10.3 | - | - | |
| 11 | Hospitalisation | ED presentations | 7 | 58 | 12.1 | 9 | 1.75 | 7 residents, 9 presentations. Includes 10B on 14 Aug. |
| 12 | Consumer experience | Survey completion (working count) | 31 | 40 | - | - | - | 9 declined. Free-text tab quotes two residents by name on night noise and a call bell.6 |
| 13 | Medication incidents (internal) | Not a program indicator | - | - | - | 4 | 0.78 | Includes the 8 July opioid strength error (resident in 8A). Reported to Dr Nair the same day.5 |
| 14 | Return | Due 21 Oct 2026. Submitted 19 Oct in the QI app by L. Keane. |
An invented workbook. Figures are illustrative and each rate was recomputed. Hover or tap any highlighted cell for the analysis; the full list is below.
Quality indicators workbook builder
An Agent+ that builds the quarter's indicators workbook from your own incident, weight and medication records. It works out the percentages and rates from the counts, keeps working papers in the same file, and flags any number that does not reconcile to its source.
What it does
- Reads the quarter's source records and last quarter's workbook
- Builds counts, eligible residents and rates with working papers
- Flags figures that do not reconcile to the source
What stays with your people. The facility manager checks every figure and decides what goes into the return. The agent builds the workbook; it does not submit anything.
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 service
- 1The home and its occupied bed days
The home’s own total for the quarter and the average number of residents. Not personal information, but unpublished operating data for one named home, sent before the return is final.
Frontier Models (ChatGPT, Claude) Unpublished home-level totals leave before the return is final Ermos Edge Stays on your instance; the submission to government is a separate step Ermos Dominion Stays on the unit; the submission to government is a separate step Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 2Pressure injury notes by room
Rooms and a resident named for the next quarter. A count of three is small enough that a room number identifies the person. Health information, which section 168 of the Aged Care Act 2024 ties to delivering her care. Patient data guide.
Frontier Models (ChatGPT, Claude) Rooms and a small count transmitted, generally offshore on personal plans Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 3A restrictive practice and a guardian’s consent
Which resident has a chemical restraint, which has bed rails and where the consent is held. Section 18(1)(f) requires the rules to require informed consent to a restrictive practice, and a use outside the rules is itself a reportable incident under section 16(1)(g). Health information about named people. Patient data guide.
Frontier Models (ChatGPT, Claude) A restrictive practice and a guardian’s consent transmitted; on consumer tiers it can train future models Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 4Falls with major injury
A named resident, a fracture, a date and a hospital, and a second resident by room. Health information from the home’s incident register. Patient data guide.
Frontier Models (ChatGPT, Claude) A named resident’s fracture and hospital transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 5The medication incident row
An internal count that is not a program indicator, with a note about a resident, identified by room, and an opioid error in July. The home’s own account of an earlier error, in a sheet that also feeds a complaint response. Patient data guide.
Frontier Models (ChatGPT, Claude) An earlier error transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 6Survey completion and comments
How many residents completed the consumer experience survey and a tab that quotes two residents by name. Residents who raise a complaint through a survey are entitled to do so without fear of reprisal (section 23(9)(b)), and their words are personal information. Privacy Act guide.
Frontier Models (ChatGPT, Claude) Residents quoted by name transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 7The Residents tab
The workbook’s other tabs list which resident was counted for which indicator. The AIHW technical notes say homes submit aggregated home-level data, so the names stay in the home. Uploading the whole file sends every tab, hidden column and cell comment, and properties such as the author, even when the screen shows one sheet. Privacy Act guide.
Frontier Models (ChatGPT, Claude) Every other tab, hidden column, comment and property travels with the upload Ermos Edge Stays on your instance in Australia; check what a workbook holds before it is indexed Ermos Dominion Stays on the unit in your office Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 8The indicator names and the program’s method
The program covers 14 indicators, including pressure injuries, restrictive practices, unplanned weight loss, falls and major injury, medication management, hospitalisation and consumer experience, and counts residents meeting criteria against eligible residents assessed. The method is published and names no one.
Frontier Models (ChatGPT, Claude) A published program method with no person in it Ermos Edge Nothing to protect in a published method Ermos Dominion Nothing to protect in a published method Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case - 9The computed rates
Each percentage is residents counted divided by eligible residents times 100, and each internal rate is events divided by occupied bed days times 1,000. The program’s own values are percentages, so the per 1,000 column is the home’s own measure. An AI tool can mis-add a column or misread a header, and a person has to recompute before the return is submitted.
Frontier Models (ChatGPT, Claude) Arithmetic a model can get wrong, with the counts attached Ermos Edge Stays on your instance, but a person still has to recompute each figure Ermos Dominion Stays on the unit, but a person still has to recompute each figure Use case Quality indicators workbook builder. Builds the quarterly quality indicators workbook from your own source records, keeping working papers beside each count, eligible total and rate. See the use case
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 service 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 resident’s 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 Edge | Ermos Dominion | |
|---|---|---|---|
| Where the document is processed | Provider servers, generally offshore on personal plans | Your instance in an Australian data centre; answers generated in Australia | The unit in your office |
| Who can read it | Provider staff and systems, under policy | Your staff, within the boundaries you set. The named Australian provider processes only permitted excerpts, under contract, and keeps nothing | Your staff only |
| Is it kept | On the provider’s terms | Kept on your instance; the provider is barred from retaining it | Stays on the unit |
| Can it train a model | On consumer tiers, yes | No, by contract | Only your own unit learns from it, on your hardware |
Part of Use cases in aged care providers.
Common questions
Do ChatGPT and Claude use the information in a quality indicators spreadsheet?
Is it safe to paste a quality indicators spreadsheet into ChatGPT or Claude?
Can I upload a quality indicators workbook to ChatGPT or Claude instead of pasting it?
How can an aged care service use AI for a quality indicators spreadsheet without sending it to ChatGPT or Claude?
Which quality indicators do aged care homes report?
How are the indicator values calculated?
Does a quality indicator return contain personal information?
See Dominion answering from your own documents.
A demonstration takes 30 minutes, uses no client data, and comes with no obligation.
Book a demo