NDIS progress reports and
the notes behind them

The report goes to a delegate at the National Disability Insurance Agency, and in most cases that person has no clinical training. They decide against the reasonable and necessary criteria in section 34 of the NDIS Act: whether the support helps the participant pursue the goals in their plan and undertake activities that facilitate social and economic participation, whether it is value for money against both the benefits and the cost of alternatives, whether it is likely to be effective and beneficial having regard to current good practice, what families, carers and informal networks can reasonably be expected to provide, and whether another service system should fund it instead. Since the 2024 amendments the Act also works from a published list of what counts as an NDIS support and what does not, so a support sitting outside that list cannot be funded however strong the clinical case.

A support coordinator or local area coordinator often reads it first and uses it to frame the request. Where the request involves assistive technology, home modifications or supports the agency treats as complex, a technical adviser may read it as well and will look for the participant's environment and the options you ruled out. If the decision is contested it goes to an internal reviewer, and from there to the Administrative Review Tribunal, where your report becomes evidence you may be asked to explain under questioning.

None of those readers can ask you a follow-up question before the decision is made, so anything that requires inference is effectively absent. Four things have to be findable on the page.

• What the participant can do now, in which settings, and with what level and type of assistance.
• How the supports delivered so far changed that, measured the same way at the start and at the end of the period.
• What is being requested next, in hours or items, and what specifically it is expected to change.
• What happens if the support is not funded, written as the daily activities that stop or become unsafe, because an adjective like high risk carries nothing on its own.

A progress report is written inside the plan period, usually at the end of a funded block of therapy, and answers one question: did the supports that were funded do what they were funded to do. It reports against the goals in the words the plan uses, gives the measures you used and the results at both ends of the block, and says what is still outstanding.

A plan reassessment report carries the funding argument, and it does three things a progress report does not. It establishes the participant's current functional baseline from scratch, because the delegate cannot assume the last plan's picture still holds. It shows what changed across the plan period and attributes that change to something, including where the change was maintenance of a level that would otherwise have declined. It states what is requested next in the units the plan will use, hours against a named therapy support or a specific item, with the outcome each should produce. A report that shows excellent progress and says nothing about why the support is still needed invites the delegate to read the need as resolved.

Reports commissioned for a particular decision are different again. An assistive technology assessment, a home modification report and a specialist behaviour support plan each carry their own required content, and a general progress report will not stand in for any of them. A significant change in circumstances during the plan period has its own pathway and should not be held until the reassessment date.

A report that arrives after the delegate has decided goes into the file for next time. Ask the participant or their support coordinator when the reassessment falls due and send the report before the planning conversation.

Session notes written as a record of treatment delivered, along the lines of reviewed home program, practised sit to stand, tolerated well, are defensible as a clinical record and close to useless as evidence of function. Twelve months of those leaves the report writer reconstructing from memory and from whatever the participant says on the day.

The fields that decide whether a report can be written describe performance rather than activity: how much assistance was needed and of what kind, whether that changed within the session, whether the task happened in the clinic or the participant's own kitchen, how long it took, and how this day compared with their usual range. A note recording that a transfer was completed with supervision early in the session and with hands-on assistance once the participant fatigued is worth more at reassessment than a paragraph describing the intervention.

Blocks of ten or twenty sessions invite copy-forward notes, where each entry inherits the last and only the date changes. That destroys the report because no trajectory is visible, and it costs the participant therapy: report writing is non-face-to-face time claimed against their own plan, so hours spent rebuilding an evidence trail from thin notes are hours of treatment they do not receive. What prevents it is a small set of fields recorded the same way in every entry.

• Record the level and type of assistance for each task attempted, using the same words each time so change is visible across the block.
• Record the setting, because performance in a treatment room and performance at home are two different facts.
• Record variability, meaning what the participant managed on this day compared with their usual range.
• Attribute what a carer, support worker or teacher reported to the person who reported it, so it is not absorbed into your own observation.
• Name the plan goal you worked on the way the plan names it, so the report can trace session to goal without interpretation.

Delegates think in activity areas, and the six the Act uses when it tests functional capacity are communication, social interaction, learning, mobility, self-care and self-management. Evidence mapped onto those six is easier to act on than evidence organised by your assessment battery. Diagnosis sets the context and does not carry the argument, because two participants with the same diagnosis can need entirely different supports and the funding attaches to the functional consequence.

The functional version of any clinical finding answers six things: what the participant does, where, with what assistance, how often, how long it takes, and what happens when the assistance is not there. Reduced dynamic sitting balance is a clinical finding. Its functional statement names the daily task it affects, the person or equipment needed for that task to happen safely, the time it adds, and what the family has to do differently when nobody is available.

Standardised measures earn their place when they are used the same way twice. A measure introduced for the first time at reassessment establishes a level and says nothing about change, which is the thing the report exists to show. Choosing the instrument at the start of the plan period, administering it under comparable conditions, and reporting both results with the date of each is what converts a number into evidence. The sentence after the number has to say what that change means for something the participant does at home, at school or in the community.

Consent to record sits with the participant wherever they can give it, and capacity is decision-specific. Many participants who cannot weigh a complex treatment decision can understand and agree to a tool that helps their therapist write the notes. Explain what is captured, what happens to it afterwards, that they can decline, and that declining changes nothing about the therapy they receive. Record the outcome in their file, because a practice-wide consent policy is not evidence that this participant agreed.

For a child, consent comes from the parent or guardian, and the child's own agreement still matters in practice. If a child keeps watching the device or asks for it to be moved, turn it off and write the session up the ordinary way. Where an adult participant cannot consent, the request goes to the person with authority under the relevant state or territory law or to a nominee appointed under the scheme, and the file should record who agreed and in what capacity.

A substantial part of therapy happens with participants who do not communicate through speech, and ambient capture is built around voice. A session with a participant who uses a communication device, key word sign, gesture or behaviour produces a lopsided transcript: your words, the support worker's words, and almost nothing from the participant. The functional evidence then thins out in precisely the sessions that are hardest to reconstruct later. The working answer is to narrate what you observe as it happens, describing the participant's response aloud in the terms you would write it.

Everyone else in the room is captured as well, and the participant's consent does not extend to them, so what a parent or a support worker says about their own circumstances lands in the participant's health record alongside everything else. Tell them capture is running, and take third-party detail that has no bearing on the participant out of the draft before you sign it.

• State out loud at the start who and what is being recorded.
• Home visits capture the household, so be explicit about when capture starts and stops.
• Group sessions and school-based sessions capture other participants and other children, which is a reason to leave capture off unless everyone involved has agreed.
• Support workers, therapy assistants and interpreters are recorded too and should be told rather than assumed to have agreed.
• Stop without argument if anyone objects, and document the session the ordinary way.

Write on the assumption that the person the report describes will read it. A private allied health practice is covered by the Privacy Act 1988 whatever its turnover, because providing a health service takes it outside the small business exemption, and the Australian Privacy Principles give an individual a right of access to the personal information held about them, health information included. Records the agency itself holds can be sought under the Freedom of Information Act as well. Participants also receive reports through their support coordinator, and registered providers work under the NDIS Practice Standards, which expect a participant's records to be accessible to them.

The test that makes this workable is whether you could read each paragraph aloud to the participant and their family and stand behind it. That rules out clinical shorthand standing in for a person, unattributed characterisations of behaviour, and family circumstances recorded because they were interesting. Separate what you observed from what you concluded, and attribute anything a third party told you to the person who said it.

Reports softened to spare feelings do more damage, because they describe a participant coping better than they are and the funding follows the description. Naming what someone cannot do without support is the purpose of the document, and it can be done precisely without characterising the person. Where a participant disagrees with something you wrote, record their view alongside yours and leave yours in place.

Sending the report to the participant before it goes to the agency catches errors of fact about their home, their household and the supports they receive outside your service. It also stops them reading something about themselves for the first time in a document that has already decided their funding.

The clinical record is the contemporaneous account of each session, and the report is assembled from it for a reader deciding funding. Once the report exists the temptation runs backwards, into tidying a note so it matches the report or writing the next few notes in the shape the next report will need. That is where a defensible file becomes an indefensible one, because the divergence shows up to anyone reading the notes and the report side by side.

Pressure to write a stronger report usually comes from people acting in good faith, most often a support coordinator who has seen a plan cut or a parent frightened about the next twelve months. The response to a request for stronger wording is to go looking for evidence you left out. If it is not in your notes, the report states what you observed and says where you have no data.

The NDIS Code of Conduct applies to every provider delivering supports under the scheme, registered or not, and requires honesty and integrity in the delivery of those supports. Registered health practitioners carry their AHPRA obligations into every document they sign, and the self-regulating professions carry their association's code. A report is also a document you may have to explain to an internal reviewer or at the Administrative Review Tribunal years afterwards, and at that point the session notes underneath it are the only thing that helps.

Inflating a report harms the participant it is meant to protect, because a claimed level of need that a later assessment contradicts damages the credibility of everything else in the file, including the parts that were accurate. Keep the session notes for as long as your state or territory health records legislation requires. In New South Wales and Victoria, a private health provider holding information collected while the person was under 18 has to keep it until that person turns 25.