How I work: my method, published in full

Chronic pain is classified before anything else, and specifically whether it is chronic primary pain, where the pain is the condition, or chronic secondary pain, where it is a consequence of something else.

Why this step exists. It decides what everything downstream attaches to. Whether chronic pain is an impairment at all under s 24(1)⁠(a) is contested, and it turns on whether a current specialist characterises it as one, which makes this the earliest thing an audit can catch.

Pain is classified by mechanism (nociceptive, neuropathic, or nociplastic) against published clinical criteria, with a separate screen for inflammatory causes.

Why this step exists. It is the difference between listing the treatments that were tried and explaining why the ones that weren't tried would not have worked. A list can always be answered by naming one more thing. A mechanism can't.

Every diagnosis is checked against a published standard, and the report names which standard was used.

Why the order matters. AAPT is the stronger anchor because it is multidimensional and pain-specific, but it covers relatively few conditions. A method that stopped there would return “not applicable” for most of a chronic pain caseload. Naming the best available guideline instead gives the decision-maker something to check in almost every case.

Why this step exists at all. It anchors the diagnosis to a published standard rather than to a clinician's say-so. A report that names the standard it applied can be verified. One that does not, cannot.

1. The AAPT criteria, where the condition has a published set: five dimensions rather than a diagnostic label alone.
2. The best available clinical practice guideline, where AAPT has not published for the condition. The report says which one, and why it was chosen.
3. Where neither exists, that is recorded as a finding in itself, and the reasoning rests on steps 1 and 2.

Every treatment attempted is mapped against the nine elements of a modern pain management program. Duration, adherence and outcome are recorded for each, and every element that was never addressed is named as unaddressed.

Why this step exists. It is the single most common reason a chronic pain claim fails on permanence, and the difference between winning and losing it is documentation rather than severity.

Mapping rather than listing is the point. A list of treatments can always be answered by naming one more. A map across all nine shows which have been addressed, which haven't, and (read with step 2) why the unaddressed ones would not change the outcome.

And one element changes how a gap reads. Where a record shows treatment recommended and not sustained, the question is not whether the person complied. It is whether they were ever supported to. Education, pacing, confidence and barriers are a treatment domain in their own right, and an element never addressed is a clinical finding rather than a fault.

These are the recognised domains of multidisciplinary pain management. The mnemonic is mine; the content isn't. H–E–A are the foundations, the four P's are the intervention modalities, S–S are the context.

A structured interview, recorded as the person's own account, and reported as such, clearly separated from anything observed or measured.

Why this step exists. Self-report is evidence. It stops being useful when it is presented as though it were observation.

Each impairment is identified, measured where measurement is possible, and attributed to the condition classified at step 1.

Why this step exists. Attribution decides cases.

Frequency, duration, triggers, and the recovery cost of a flare, recorded as a pattern over time, not as a snapshot.

Why this step exists. This is the most commonly missed evidence in the entire published record.

The Pain, Sleep and Fatigue Diary is the simplest way to produce it. One line a day, about a minute: day type, sleep hours and night waking, pain and fatigue at their worst, whether the cost arrived a day or more later, and what was left undone. Four weeks to a form, and the scoring page returns a count by day type, medians against a stated denominator, and the longest consecutive run of hard or bed days.

If you don't already keep one, I provide it, and it costs nothing.

What is given up to get through a day. What is deferred, delegated, or abandoned. What the next day looks like.

Why this step exists. Capacity assessed one task at a time overstates what a person can actually sustain.

Ordinary daily tasks, performed in a realistic sequence rather than in isolation, with rests, stopping points and what was left undone all recorded. Then three follow-up contacts, at 24 hours, 72 hours and 7 days, to document what the sequence cost.

Three contacts rather than one, because the cost of a sequence usually arrives late. Post-exertional symptoms can begin a day or more after the activity that produced them, and can persist well beyond that. A single next-day call records the beginning of a flare and calls it the whole of it.

The 7-day contact is the evidentially valuable one. How long a collapse in capacity lasts is precisely what these files keep failing to establish. A follow-up that stops at 24 hours cannot answer that. One that runs to seven days can.

This is sequencing, not provocation. Tasks are performed in a realistic order with a stopping rule, not pushed to failure. The distinction matters clinically and it matters evidentially.

Supervision is the default, and the assessment is by video. Where a support person is available they are present throughout, and that is what makes video safe. Where nobody is available the sequence is modified rather than refused: tasks requiring someone in the room are removed, and the removal and its reason are recorded. A reachable third party during the session is the one condition that does not flex.

An absent support person is a finding, not a gap. Whether a person has informal support to call on, and what happens on the days they don't, is squarely within what these decisions have to determine. The report states what the absence establishes rather than leaving it unexplained.

Work, income, relationships, community, and what the condition has cost across each.

Why this step exists. The line that has to be held is between what a person cannot do and what they have stopped doing.

The expected trajectory of the condition. Whether ongoing management is required to prevent deterioration or adverse outcomes. What happens without support. And whether the need is stable, increasing, or fluctuating across the life course.

Why this step exists. It is a prognosis, and it is easy to leave implicit, which is how it gets missed. One Tribunal put the difficulty plainly: where an impairment is permanent and functional capacity substantially reduced, it is hard to imagine a circumstance in which the need for support would cease. But the requirement still has to be answered by evidence rather than by inference.

A warning that belongs here. Evidence that nothing will change the trajectory establishes permanence and lifetime support, and it is the same evidence that defeats an early intervention claim under s 25. There is no record that satisfies both. Which of the two is being argued has to be decided deliberately, and early.

An audit is not a lighter version of an assessment. It is the same eleven questions, asked of the documents rather than of the person, and read against whichever test your request faces: the support test or the access test.

For each step, the audit establishes three things:

• What your records already prove, and which document proves it
• Where they stop: the point at which the evidence runs out
• What would close the distance, or whether it can be closed at all

That last part is the whole value of the audit. Sometimes the evidence exists and nobody has stated it in the terms the decision uses. Sometimes it was never recorded. And sometimes the honest answer is that the request can't succeed: for a support, because it doesn't meet the test; for access, because a reasonable treatment hasn't been tried yet. Saying so is a real result, and it saves you a request that was never going to succeed.

Three of the eleven steps establish what the condition is. The other eight establish what it costs, what has been lost, and what happens next. In most files the first three are already well documented, because that is what treating records are for. The gap is almost always further along.

Four things are stated clearly in every report I write, without exception:

• My qualifications and the scope of my expertise, including what falls outside it
• Every source I relied on, listed, with dates
• A declaration that I have complied with the expert evidence guidelines applying to the Tribunal
• What I observed, separated from what I was told, labelled throughout the report

What the report does with what it couldn't establish. A fifth thing, rarer, and it matters as much. Where something could not be assessed (a task removed from the sequence, a session that stopped early, an assessment done without anyone else present), the report states what that establishes rather than leaving it unexplained. A silence in a report is filled in by whoever reads it, and it is not usually filled in favourably.

On the use of AI. Clinical observation, reasoning and opinion in my reports are mine. Where software assists with formatting or checking, it does not generate findings, and the words are mine. I state this because the Tribunal has already begun asking: in one decision an occupational therapist's report was discounted, and the competing assessor preferred, partly because she told the Tribunal she had used AI to improve the way it was written while he did not and recorded his own observations. Alsayed at [69].

Availability. I can be called on to answer questions about any report I write, and I will. Attendance at a hearing or a conference may incur an additional fee, quoted in advance. A report whose author cannot be asked about it carries less weight, and the record is consistent on that point.

The Tribunal gives less weight to practitioners who stand to be paid from the plan they are writing about. These four commitments are how I make sure that can never be said of my evidence, and all four can be checked:

• The fee is capped before the work starts and does not change with what the audit finds or what the decision-maker later decides. It can come down where the work is shorter. It never goes up.
• I do not deliver funded supports to anyone whose file I have audited or reported on, or whose future funding depends on a report I wrote. The evidence and the supports are separated.
• I am paid once for the evidence, the same either way, and I do not appear in the plan that follows.
• The audit is a single engagement, not an ongoing support.

The honest limit. I am paid for this work. Nobody who is paid can claim to have no interest at all, and I am not going to pretend otherwise. What I can say precisely is what the difference is: the Tribunal's criticism has been directed at practitioners with an ongoing interest in future funding: people whose continuing income depends on the plan being approved.

The second limit, stated plainly. The separation above is a policy, not an impossibility. I have always been able to deliver plan-funded supports. Self-managed and plan-managed participants have always been able to engage me for them. And I have always declined to do it for anyone whose evidence I wrote. Becoming a registered NDIS provider changed who can engage me, not what I will take. So the guarantee is a published commitment, checkable against what I actually do, and it should be read as that rather than as a structural bar.

That is a narrower claim than “I am independent.” It is also one that survives being checked.

This section exists because the rest of the page is only believable if the limits are stated.

• I can't tell you what the decision-maker will do. No assessment predicts an outcome. Members reach different conclusions on similar facts, and the published record makes that plain.
• I can't give you legal advice. Not which review pathway applies, not whether to appeal, not what a deadline means for you.
• I can't manufacture a history that didn't happen. The assessment produces new clinical evidence; that is what it is for. It cannot make an untried treatment have been tried, or close a two-year gap in a record. Where the honest finding is that something hasn't been done yet, that is the finding.
• I can't reach outside physiotherapy. Psychiatric diagnosis, neuropsychological testing, specialist medical opinion: where a file needs one, the finding will say which and why.
• An assessment is a point in time. It records what was true on the day and, through the follow-up, what that day cost. It does not describe every day, which is why I would rather you started the diary before we begin than after.
• And the method itself is not authority. It is a structure for making reasoning visible. Whether the reasoning is accepted is for someone else to decide, which is why it is written down where it can be argued with.

The eleven steps are not a clinical preference. Each exists because a decision-maker has to be satisfied of something specific, and because published decisions show what happens when they aren't.

The six activity domains at step 9 are the ones s 24(1)⁠(c) names: communication, social interaction, learning, mobility, self-care and self-management. A report that assesses four of them has answered two thirds of the question.

s 24(1)⁠(e) appears twice because it has two limbs, and they are answered by different evidence. The first is whether support will be needed for the person's lifetime: a clinical prognosis, and step 11 does it. The second is whether some other service system should be meeting the need instead.

That second limb is the one the assessment can't reach, and it would be dishonest to imply otherwise. Whether aged care, a community program or a mainstream health service should be meeting the need is settled by evidence about that system: what it will fund, what it won't, and whether an application has been made and tested. That evidence is documentary, it belongs to the records audit rather than the assessment, and it is the requirement that has moved most in the last year.

• Added how the eleven steps answer a support request under s 34(1), alongside the access request mapping. No step was added, removed or reordered. Wording updated throughoutV1.1 · 30 SEP 2026
• First publishedV1.0 · 18 SEP 2026