Digital health with Aboriginal and Torres Strait Islander communities

A national login is not a clinic. That sentence should be obvious. It keeps being forgotten when digital health is presented to Aboriginal and Torres Strait Islander communities as a solution in a browser tab.

The working system, for a large part of the population this article is about, is already local. Aboriginal Community Controlled Health Services — also called Aboriginal Medical Services, or AMS, in many towns — are incorporated organisations governed by local Aboriginal and, where relevant, Torres Strait Islander boards. They run primary care, often allied health, often social support, and they hold the clinical record that staff actually open on a Monday. The national peak body is NACCHO. The local authority is the board, not a portal.

This report starts there. It does not start with a gap statistic, and it does not treat community-controlled care as a mobile outreach that digital tools will eventually replace. It is journalism about how digital products sit beside services that already exist. It is not clinical advice, and it does not speak for any particular community.

Community control is a governance fact

An ACCHS is not a mainstream practice with extra art in the waiting room. Community control means Aboriginal and Torres Strait Islander people elect or appoint the governing body, set priorities, and employ the service. Some organisations are urban and large. Some are small and very remote. Torres Strait communities have their own services and geography. Lumping those together as "remote Indigenous health" is how national programs buy the wrong software.

Digital tools arrive into that governance, or they try to go around it. A well-built tool is procured or agreed by the service, configured for the clinic's privacy rules, and staffed by people who already know the families. A poorly built tool is launched from a capital city, aimed at individuals, and then blamed on "uptake" when people do not enter a government password.

Staff in community-controlled services already do digital health. They use clinical information systems, pathology inboxes, telehealth, electronic prescribing, and, where it is useful, My Health Record. The question is not whether the sector is modern. The question is who the system is accountable to when something goes wrong.

Mainstream general practices and hospitals remain part of the picture. Aboriginal and Torres Strait Islander people live in every city and use those services. The difference, when an ACCHS is available and chosen, is that the relationship is collective as well as clinical.

Data sovereignty is not a synonym for privacy

Indigenous data sovereignty is a set of principles about the rights of Indigenous peoples to govern data about their communities, lands and members. In Australia, Aboriginal and Torres Strait Islander scholars and organisations — including the Maiam nayri Wingara collective — have set out what that means in practice: authority to decide why data are collected, how they are interpreted, and who they are given to.

That is broader than a privacy notice. A file can be stored securely and still be used to tell a story the community did not agree to. A de-identified extract can still identify a small community. A research dashboard can still rank a town without returning anything a clinic can use.

Where the data actually sit

Health data sit in several layers at once.

• The working record in the ACCHS or AMS clinical system.
• State hospital and pathology systems.
• Medicare and other payment trails.
• My Health Record, if the person has not cancelled it and if anyone is uploading.
• Consumer apps and wearables that never enter a community agreement.

Community-controlled services have spent years building data governance for the first layer: who may see a file, how research requests are handled, what is reported to funders. National infrastructure often arrives with its own secondary-use story already written. Those stories are not automatically illegitimate. They are not automatically agreed either.

If a government agency or a vendor wants to add a new extract, the adult process is a negotiation with the organisation that holds the relationship, not a poster in English about the public benefit. For the consumer-facing access rules on the national record, see who can see your My Health Record. Access controls and sovereignty are related. They are not the same job.

Kinship, nominated representatives and the next-of-kin field

Clinical software still thinks like a nuclear family. Next of kin. Emergency contact. Parent or guardian. Those fields matter in an emergency. They do not describe Aboriginal kinship systems, where responsibility for a child, an older person, or a decision may sit with a different aunt, uncle or grandparent than the form expects.

My Health Record tried to meet some of this with nominated representatives and with settings for children's records. The mechanism only works if someone sits down and sets it, and if the person whose record it is actually wants that representative. A family member who drives the person to clinic is not thereby authorised to open a portal. A family member who is the right person under local kinship rules may be missing from the form because nobody asked.

Staff who know the family are better at this than a national call centre. That is one reason the ACCHS or AMS is the right place to do identity and representative work, including for children's digital health consent. Rushing a parent through a myGov setup in a hospital corridor is how the wrong adult ends up with the password.

There is also a privacy risk in the other direction. Shared housing, shared phones and a busy clinic desk can disclose a diagnosis to people the patient did not choose. Digital systems that SMS results to "the mobile on file" inherit that risk. Ask which number, every time.

A nominated representative is a decision. It is not a courtesy title for whoever holds the car keys.

My Health Record, without an uptake lecture

Public commentary likes to treat My Health Record registration or use among Aboriginal and Torres Strait Islander people as a scoreboard. That habit says more about the commentators than about the record.

Some people use it because a hospital clerk set it up and the medicines list helped a locum. Some people cancelled it, or never wanted it, because a national government database is not a neutral object in families with long memories of surveillance, child-protection files, or police requests. Some people have a record and do not open it, because the working information lives at the AMS and the national summary is thin.

Those are reasons. They are not a literacy failure.

Clinicians in community-controlled services make a practical choice about what to upload. A shared medicines list can prevent a harmful duplicate. A mental-health letter can travel further than the person intended. The service's upload policy should be explainable in the room, in the languages the service actually uses, including Aboriginal languages and Torres Strait Creole where those are the languages of care.

If a patient asks whether to keep the record, a decent answer names the access rules, the cancellation path, and what the clinic will still hold either way. It does not sell the portal as a community.

Language, trust and design that does not perform

Digital products still ship with English first, legal English second, and a stock photograph of a sunset if anyone remembered "culture". That is not engagement.

Trust is built by staff who stay, by reception that does not make a person tell their story at the counter, and by systems that do not phone a relative the patient asked them not to phone. A login screen cannot do that work. It can undo it, quickly, if a password reset goes to the wrong house or a telehealth room is a corner of an open office.

Design that a clinic can actually use

Practical design that community services already ask for is unromantic:

• Rooms and devices at the clinic, so a person is not forced to use a shared home phone.
• Interpreters and local language speakers in the consult, including on video, booked ahead.
• Male and female staff options where that is required for the care.
• No surprise recordings. Say if a session is stored.
• Identifiers that can handle a skin name, a change of name, and more than one community address without creating a duplicate file.

None of this is special software. It is the same discipline a good ACCHS already applies to paper.

Scams that impersonate Medicare or myGov hit these communities too. Official-looking texts should be treated as suspect. The clinic, not the link, is the place to check.

Why a national portal is not a community clinic

Portals are good at transactions. Bookings. Scripts. A summary to take to a city hospital. Clinics are good at relationships, at finding the person who did not come in, at holding knowledge that should not sit in a federal index.

When funders lean on digital channels, they sometimes cut the unglamorous transport, the after-hours nurse, or the second Aboriginal health worker. The portal then looks busy and the service is thinner. That swap is a policy choice. It should be argued as one, not described as innovation.

Telehealth is the same argument in a different window. A visiting physician on a screen can be a genuine addition when the ACCHS books the room, the family is invited on the patient's terms, and the letter comes back into the local record. A marketplace app that never writes to the AMS fragments care and still bills the Commonwealth. Distance is not a reason to accept that fragmentation. It is a reason to refuse it.

Urban Aboriginal and Torres Strait Islander people are not a footnote. A community-controlled service in a capital city has different bandwidth and a different hospital network. It has the same right to govern its data and the same need not to be treated as a communications campaign.

What digital health can do if it stays in its lane

There are uses that earn their keep.

Electronic prescribing that actually reaches the pharmacy the person uses. Pathology inboxes that a GP in the ACCHS can see without waiting on a disc. Specialist video that the service hosts, so the patient is not alone with a link. Record systems that let a person ask for a copy, and get one. Agreements that say a university will not publish a community map without the board.

Those are implementations. They require procurement, training and the authority to say no. They do not require a new national slogan.

If you work in a mainstream practice or a hospital, the useful habit is to ask which community-controlled service the person already uses, send the letter there, and not invent a parallel digital relationship. If you work in a vendor team, the useful habit is to read the service's data policy before you demo a dashboard. If you are a patient or a family member, the useful habit is to set representatives on purpose and to treat a national login as an optional extra.

Digital health will keep arriving. Community-controlled services will keep being the place where care, data and accountability already live. The work is to make the arriving tools answer to that fact, not the other way around.