Writing NDIS Reports That Get Approved the First Time

A report that gets approved on first read is not a more elaborate report. It is a more targeted one. The NDIS reviewer is reading dozens of submissions a week. They want to see four things, in a predictable order, and they want to see them written in language the scheme already recognises. The clinicians who consistently get approvals on the first pass have internalised this audience and write to it deliberately.

This post unpacks what those four things are, where most rejected reports fall down, and how to shift your writing to read like a planner-ready submission. None of this is novel. It is the same pattern you would learn from a senior clinician sitting next to you reviewing your drafts. The goal is to make it explicit so you can apply it the next time you sit down to write.

What NDIS planners are actually looking for

• A clear link between the participant's disability and the functional impact in daily life
• Evidence that the requested supports are reasonable and necessary under section 34 of the NDIS Act
• Specific, measurable goals tied directly to the participant's plan
• A description of what happens without the support, sometimes called the 'do nothing' case

These four threads map almost exactly to the language used in the NDIS Operational Guidelines. When a report is rejected or sent back for further information, it is almost always because one of those four threads is missing, buried halfway through a paragraph, or implied rather than stated. Planners are not hostile readers. They are time-poor readers who are scanning for specific evidence. If they cannot find it in the first read, the report bounces.

The patterns we see in rejected reports

Pattern 1: clinical observation without functional translation

A clinician will write 'Client demonstrates reduced grip strength of 18 kg on the right and 22 kg on the left.' That is an observation. It is precise, it is measured, and it is useless to a planner unless the next sentence translates it into a functional consequence. What the planner needs is the immediate next line. 'This limits the client's ability to carry a four-litre shopping bag independently, prepare meals using standard cookware, or open jars without assistive devices.' Numbers belong in reports. Every number needs a functional consequence next to it.

The same rule applies to range of motion measurements, balance scores, fatigue ratings, and standardised assessment results. The score is the input. The functional translation is the output the planner is buying. If you find yourself listing scores back to back without a functional sentence between them, your report is in observation mode and needs to be lifted into translation mode.

Pattern 2: goals that do not pass the SMART test

'Improve independence in the community' is not a goal a planner can fund against. It is a direction of travel. 'Independently catch a bus from the participant's home to the local shopping centre within 12 months, supported by three practice sessions per fortnight in the first three months' is a goal. Specific, measurable, achievable, relevant, time-bound. SMART goals also force the rest of the report to align. If your goals are SMART, your recommendations almost write themselves, because each recommendation is now justified by a measurable target.

Pattern 3: the 'reasonable and necessary' test is not argued

The reasonable and necessary test has six criteria under section 34 of the NDIS Act. A strong report addresses each of them, even briefly. Most rejected reports address only effectiveness and benefit, and leave value-for-money, parental responsibility, and the 'most appropriately funded' criteria unaddressed. That is a fixable gap. A short paragraph that walks the planner through each of the six criteria, in order, removes most of the common reasons a report gets sent back for more information.

Pattern 4: the 'do nothing' case is absent

Planners need to understand what happens to the participant if the requested support is not funded. Without this, the request reads as a wish list rather than a clinical necessity. The 'do nothing' case does not need to be alarmist. It needs to be honest. 'Without weekly community access support, the participant is likely to remain reliant on family members for transport, which is unsustainable given the primary carer's own health concerns, and is likely to lead to social withdrawal and skill regression.' That is a complete and reviewable case.

Write to the reviewer, not to yourself

This is the single biggest shift we see in clinicians who consistently get reports approved on first submission. They write the report knowing exactly who is going to read it, what that person is scanning for, and where to put each piece of information so it can be found in 30 seconds. The mental model is not 'I am writing what happened in the assessment.' The mental model is 'I am answering the questions the planner is about to ask.'

“Every paragraph should answer a question the planner is already asking.”

Some clinicians find it helpful to literally write the planner's questions as section headings during their first draft and then delete the headings before submission. Try it once. It produces a noticeably more targeted report.

How this scales

An experienced clinician internalises all of this over years. A newer clinician, or a clinician carrying a caseload that means they are writing four reports a week, does not have the bandwidth to apply this discipline to every paragraph from a blank page. This is exactly where AI-generated drafts earn their place. The structural and rhetorical work that takes a senior clinician years to internalise can be encoded into a draft. The clinician then applies the clinical judgement that the model cannot, which is the part that matters and the part that justifies the signature on the report.

Used well, a generated draft accelerates the part that scales (structure, language, ordering) and protects the part that does not (clinical reasoning, professional judgement, contextual nuance). Used badly, a generated draft creates a polished report with the wrong clinical content. The difference is the clinician's review, which is why the workflow has to put review at the centre rather than treat it as a formality.

References

• NDIS Operational Guidelines, Reasonable and Necessary Supports.
• NDIS Act 2013 (Cth), s.34.
• Occupational Therapy Australia, NDIS resource hub.
• AHPRA Code of Conduct, documentation standards for registered health practitioners.