How to Write NDIS Evidence a Planner Can Use

Evidence has one job: show what happens without the support

The strongest NDIS evidence answers three questions in plain language. What can this person do, what happens when the support is not there, and how often does that situation arise. Everything else in a report is context around those three answers.

Funding decisions sit with the NDIA and turn on what the delegate can see in the file, so the practical goal is not persuasion. It is removing the need to guess. A file that leaves a planner inferring the answer produces a plan built on their inference rather than on your week.

Describe function, not diagnosis

A diagnosis explains why. It does not describe what a Tuesday costs. Two people with the same diagnosis can need completely different support, which is why a report that leads with the condition and stops there is weak evidence even when it is medically excellent.

Compare these two sentences. "Client has cerebral palsy affecting all four limbs." And: "Requires physical assistance of one person to transfer from bed to wheelchair, and cannot reposition independently overnight, so needs repositioning approximately every three hours." The first is a fact. The second is a support need, a frequency and an overnight implication, and only one of them can be turned into a roster.

The parts of a report that get used

When a report is doing its job, a planner can read it once and describe the support without asking anybody anything. These are the elements that make that possible.

• The task, broken into steps. Not "assistance with personal care" but which parts of showering, dressing and toileting need a person, and what that person does.
• Frequency and duration. How many times a day, how many minutes, how many nights a week. Ranges are fine and honest; "as required" is not, because it can mean twice a year or twice an hour.
• Level of assistance. Prompting, supervision, standby, hands-on with one person, hands-on with two. These are different costs and they should not be blurred into "support".
• Consequence without it. Stated flatly and without drama: skin breakdown, missed medication, a fall, no meal, no contact with anyone for four days.
• Who provides it now. Often an unpaid family member at an unsustainable level. Say so, with hours, because an informal support that is quietly holding everything together is central information rather than an aside.
• What has been tried. Equipment, therapy, a different routine, and what the result was. This is what separates a considered request from a first guess.

Write the numbers down, even rough ones

Support hours are built from counts, and a file with no counts gets counted by somebody else. If nobody has ever measured it, measure it for a fortnight. A phone note with date, time, what happened and how long it took is enough to turn an impression into a record.

Two weeks of ordinary days beats a memory of the worst week. If the worst week matters, and it often does, include it separately and label it as the exception rather than letting it stand in for normal. Overstating a typical day is the fastest way to lose credibility with somebody who reads these files all day.

What weakens a file

Reports get discounted for predictable reasons, and most of them are fixable. Undated documents, because a planner cannot tell whether a 2023 assessment still describes the person. Copy-and-paste sections that clearly came from a template, because they read as generic. Recommendations for a specific product or provider without the functional reasoning underneath.

Contradiction across documents does the most damage. If an occupational therapy report says one person assists with transfers and a support plan says two, somebody has to decide which is right, and they may resolve it the cheaper way. Read your own file end to end before it goes anywhere and reconcile the differences, or explain them.

Length is not strength either. A thirty page report with the key sentence on page nineteen is worse than four pages that say the same thing plainly.

Who writes which part

Split the file by who can actually observe the thing. An occupational therapist writes function, equipment and the home environment. A speech pathologist writes communication and swallowing. A physiotherapist writes mobility and transfers. A GP or specialist writes the medical picture and the risks. A behaviour support practitioner writes behaviour, triggers and strategies.

You write the week. Nobody else can, because nobody else is there for all of it. A family statement covering a normal seven days, hour by hour where it matters, is not a lesser document than a clinical report. It is the only one that shows the whole pattern, and it is where overnight support usually becomes visible for the first time. Our piece on evidence that moves a plan reassessment goes into what to gather when a plan is being redone.

Assemble it as one file, not a pile

Put it together the way you would want to receive it. A single PDF, newest first, with a one page cover summary listing what is enclosed, the date of each document and one line on what each one shows. Name the files sensibly. Number the pages.

This is not presentation for its own sake. A cover page is where you get to state the support need in your own words and have it read first, and a file that is easy to move through is a file that gets read properly. For SIL specifically, the SIL Handbook chapter on eligibility and evidence sets out what belongs in that particular file, including the draft roster of care.

Then keep it current. Add to it as reports arrive rather than assembling it in a fortnight of panic before a meeting, and keep a copy yourself. If you want a hand working out what is missing, that is the sort of thing our team helps with on the NDIS help page.