Cancer and Disability in Australia: Why Outcomes Are Worse

Cancer and disability in Australia: why the outcomes are worse, and what actually causes it

There is a number that should be the headline of every conversation about disability and health in this country, and it almost never is.

People with intellectual disability in Australia die at a median age of 54. The general population median is around 81. That is a gap of roughly 27 years. And 38 per cent of those deaths are classified as potentially avoidable — more than double the rate in the general population. Those figures come from the National Centre of Excellence in Intellectual Disability Health’s State of Play report, published in March 2025, drawing on the most recent available national data.

Potentially avoidable means something specific. It does not mean the person was frail and something was always going to happen. It means the death was from a cause that, with timely and effective healthcare, should not have killed them at that age. Cancer sits heavily inside that category, because most common cancers in Australia are now survivable if they are found early and treated properly.

So the question this guide is about is not whether people with disability get cancer at higher rates. Mostly they don’t — the incidence picture is mixed and varies by cancer type and by disability type. The question is why, when they do get cancer, they are more likely to die of it.

A systematic review published in 2024 pulled together 73 studies on cancer inequalities among people with disability, searching the literature up to November 2023. The findings were consistent across disability types and across countries: screening participation was lower, and cancer mortality was higher. The authors noted that the mortality gap was most pronounced among people with intellectual disability.

That combination — lower screening, higher death rate — is not a coincidence. It is a causal chain, and most of its links are things the health and disability systems do, not things the disease does.

We are a disability provider in New South Wales. We are not an oncology service and we do not treat cancer. What we see is the disability end of this: the participant whose supports fall over the week they start chemotherapy, the person whose weight loss was recorded in progress notes for eight months before anyone ordered a scan, the family who discover that nobody in the system considers it their job to book a cervical screening test. This library is about that end of it. For cancer information itself, Cancer Council’s 13 11 20 line is free, staffed by nurses, and far better than anything we could write.

What the numbers actually show

Let me set out the evidence properly, because this is a field where people reach for emotive language instead of data and it does the argument no favours.

Population size. Around 4.4 million Australians live with disability — roughly 18 per cent of the population. People with intellectual disability make up a much smaller group: approximately 1.8 per cent of Australians, or around 490,000 people as of 2024.

Life expectancy. Median age at death for people with intellectual disability: 54 years. Around 27 years earlier than the general population.

Avoidable deaths. 38 per cent of deaths among people with intellectual disability are classified as potentially avoidable, compared with a general population rate less than half that.

Health service use. People with intellectual disability use emergency departments at about 1.6 times the rate of the general population, and wait longer once they get there. They are admitted to hospital at roughly twice the rate. And they use preventive healthcare — the screening tests, the check-ups, the things that find cancer early — at lower rates.

Read those four together and a pattern emerges that has nothing to do with biology. High use of crisis care. Low use of preventive care. Earlier death from causes that were preventable. That is the signature of a system that responds to emergencies and misses everything before them.

Screening participation. This is where cancer specifically comes in. A scoping review published in the International Journal for Equity in Health in 2026 by Koko and colleagues, conducted as part of the EUCanScreen project, examined 78 studies on cancer screening among people with disability. Cervical screening came back lower in 24 of those studies, breast screening in 19 and colorectal in 10. The direction of the finding barely varied.

For context on what “lower” means, here are the Australian national participation rates for the general population, from the Australian Institute of Health and Welfare:

Those baselines are not impressive. Bowel screening — a free test that arrives in the post and involves a sample you collect at home — is taken up by two people in five. People with disability sit below each of these figures. When you start from 40.9 per cent and go down, you are describing a programme that is reaching a minority of a minority.

Discrimination. In 2019, 22 per cent of people with disability aged 15 and over reported experiencing discrimination, compared with 15 per cent of people without disability. Around a quarter of those who experienced discrimination from healthcare staff subsequently avoided health facilities. Roughly two in five people with disability reported difficulty accessing medical facilities.

That last cluster of figures is the one I would ask people to sit with, because it closes the loop. Someone is treated badly at a clinic. They avoid clinics. Their cancer is found late. They die of something that was survivable at stage one. Every step in that sequence is documented in Australian data, and none of it is about the tumour.

The chain, link by link

Link one: nobody is identified in the first place

Cancer screening in Australia works by invitation. The bowel programme posts a kit. BreastScreen sends a letter. The cervical register prompts a reminder. All three depend on a register that knows who you are and when you are due.

None of those registers record disability.

That sounds like an administrative footnote. It is actually the foundational problem. If the register does not know that the person receiving the invitation uses a wheelchair, or has an intellectual disability, or is non-speaking, then the invitation cannot be adapted. There is no accessible version triggered, no Easy Read letter, no flag to the clinic that a longer appointment will be needed, no note that this person may need two or three visits before a test can be completed. The letter goes out identical to everyone else’s, and for a substantial number of people it is functionally unreadable or describes a procedure they cannot access in the form offered.

The Koko review named this explicitly: registries that do not record disability mean people are not identified for accessible invitations. It is a design flaw at the top of the funnel, and everything downstream inherits it.

Link two: reasonable adjustments have to be requested in advance

Here is the second structural problem, and Cancer Council Australia has put it in its own policy material: reasonable adjustments in cancer care often depend on being requested ahead of time.

Think about what that requires. The person — or someone acting for them — has to know that adjustments exist, know which ones they need, know who to ask, ask early enough for it to be arranged, and be believed. If any one of those steps fails, the appointment happens without the adjustment, which usually means it happens badly or not at all.

For somebody with a well-organised family and a good GP, that chain holds. For somebody living in supported accommodation with high worker turnover, no family contact and a GP they see for fifteen minutes when something is already wrong, it does not. The adjustment system is built for the people least likely to need it.

Link three: the building and the equipment

This is the most concrete of the barriers and in some ways the most indefensible, because it is a purchasing decision.

Cancer Council’s policy work identifies the physical access problems as: absent ramps and automatic doors, inaccessible parking and transport, and — the one that matters most for cancer specifically — no accessible examination tables or imaging equipment.

A mammogram requires standing and positioning the chest against a plate at a fixed height. A person who cannot stand, or cannot hold a position, or cannot be positioned at that height, cannot have a standard mammogram. A cervical screening test requires getting onto an examination table and holding a position. A table that does not lower, in a room without a hoist, excludes anyone who cannot transfer independently.

These are not subtle access issues requiring cultural change. They are height-adjustable tables and a ceiling hoist in one screening room. The reason they are usually absent is that nobody bought them.

Link four: the symptom gets attributed to the disability

This one has a name — diagnostic overshadowing — and it has its own guide in this library because it is the single most lethal link in the chain. The short version: a new symptom in a person with a long-standing disability gets interpreted as part of the disability rather than investigated as something new.

The person with cerebral palsy who starts losing weight has swallowing difficulties, so the weight loss is the dysphagia. The person with Down syndrome who becomes withdrawn is in behavioural decline, or early-onset dementia. The person with an intellectual disability who hits themselves in the side of the head has behaviours of concern, and gets a behaviour support plan rather than an ear examination.

The NCEIDH report names diagnostic overshadowing directly, alongside unconscious bias, miscommunication, the absence of the person’s usual support people from consultations, and poor supported decision-making. Cancer Council’s policy position puts it bluntly: health professionals may dismiss or inadequately investigate symptoms reported by people with disability.

Late-stage diagnosis is the mechanism that converts all the preceding barriers into deaths. Bowel cancer caught at stage one has a five-year survival rate above 95 per cent in Australia. Caught at stage four, it is around 13 per cent. The cancer is the same cancer. The difference is when somebody looked.

Link five: the clinician has never done this before

A statistic from Cancer Council’s policy material that I found more revealing than any of the big population numbers: in one sample of general practice nurses, about 16.5 per cent reported having no professional or personal experience with people with disability at all.

Not “limited experience”. None.

The same material notes that oncology nurses report being less comfortable communicating with people with intellectual or cognitive disability. Which is predictable, and not a character flaw — it is a training gap. But its consequence in a consulting room is that the clinician talks to the support worker instead of the patient, rushes, misses information the person could have given, and does not notice they have missed it.

Link six: the appointment is too short and too rigid

Standard consultations are built around standard pace. The Koko review identified inflexible scheduling and short appointments as recurring barriers, and the NCEIDH report noted that funding structures do not support longer consultations or quiet environments.

A person who needs an hour gets fifteen minutes. A person who needs a dim, quiet room gets a waiting area with fluorescent lights and a television. A person who needs three visits to tolerate a procedure gets one attempt, which fails, and then a note in the file saying the patient was non-compliant.

Link seven: the diagnosis arrives and the disability supports fall over

This is the link that is most specifically our business, and it is the one I want to flag hardest, because Cancer Council has independently identified it as a gap.

Their policy material states that a cancer diagnosis and hospitalisation can disrupt existing disability supports. It also states that eligibility for additional support is hard to determine without a clear prognosis.

Both of those are true and both of them are unresolved. When an NDIS participant is admitted to hospital for cancer treatment, their plan does not automatically adapt. Their support workers may not be funded to attend. Their SIL arrangement may be charged for an empty bed or may be lost. Their plan may come up for review during treatment with nobody available to attend the meeting. And when they come home needing twice the support they needed before, the funding reflects the person they were at the last planning meeting.

Meanwhile the health system assumes the disability system is covering the support, and the disability system assumes the health system is covering the treatment, and the person falls into the space between them. Cancer Council’s policy priorities include clarifying roles between health and disability supports, and developing an Optimal Care Pathway for People with Disability — a document that does not currently exist. The fact that a peak cancer body has had to list “work out whose job this is” as a policy priority tells you how unsettled the boundary is.

What the evidence says actually works

The barrier literature is depressing and extensive. The intervention literature is smaller but more useful, because several things have been tested and some of them worked.

A community education intervention. Bowler and Nash, working in the United Kingdom in 2015, ran a twelve-week community-based intervention aimed at improving cancer screening uptake among people with intellectual disability. Participation rose from 47 per cent to 61 per cent. A fourteen-point improvement from twelve weeks of education is a substantial result for a cheap intervention.

Desensitisation and pre-visits. This appears repeatedly in the literature as a facilitator: the person visits the screening site before the appointment, meets the staff, sees the room and the equipment, and does not have anything done to them. Then they come back and have the test. For somebody who finds unfamiliar clinical environments frightening — which is a great many people — this converts an impossible appointment into a possible one. It costs one extra visit.

Longer appointments. Consistently identified as a facilitator. Not a complicated intervention. It requires the booking system to allow a double or triple slot and the funding structure to make that viable.

Easy Read materials. Named throughout the facilitator literature. Information about what the test is, why it matters and what will happen, written and illustrated so the person can understand it themselves rather than having it explained over their head.

Trained, empathetic staff, and female practitioners for cervical and breast screening. Both appear as facilitators. The second is not a preference to be negotiated away; for a substantial number of women it determines whether the test happens.

Telling the person their results directly. This appears in the facilitator literature and it is worth pausing on, because it means the default is often the opposite — results communicated to a carer, a family member or a group home, and not to the person whose body it is.

Alternative testing modalities. Two are documented. A seven-year Japanese programme used breast ultrasonography rather than mammography for women who could not be positioned for a standard mammogram, with low recall rates. And speculum-free Pap testing achieved 100 per cent specimen adequacy in reported use — meaning the sample was usable every time, without the instrument that makes the test intolerable or impossible for many people.

That second one deserves to be better known in Australia. The barrier to cervical screening for a great many people with disability is the speculum and the positioning. Self-collection, now available through the National Cervical Screening Programme, removes both.

Australian consensus work. Weise and colleagues, publishing in 2024, used a modified Delphi process to reach expert consensus on 52 strategies for improving breast screening access for people with intellectual disability in Australia. That is a developed, agreed, country-specific body of strategy. It exists. The question is implementation.

Why implementation is the actual problem

Australia has a National Roadmap for Improving the Health of People with Intellectual Disability. It is a real document with real commitments. In November 2024, the Roadmap Implementation Governance Group reported on progress: of 72 short-term actions, 56 had commenced and nine were complete.

Nine.

I am not going to pretend that policy implementation is easy or that the people working on it are not trying. But a roadmap with nine completed short-term actions after several years is not yet changing what happens in a screening clinic in Western Sydney. The barriers in this guide are, with very few exceptions, known. Documented. Reviewed. Agreed upon. The interventions that fix them are also known and mostly cheap.

What is missing is somebody whose job it is to make them happen for a specific person on a specific day.

The one lever that works at the individual level

If you take one practical thing from this guide, make it this.

For people with intellectual disability, Medicare funds an annual comprehensive health assessment. MBS item 707 is a prolonged health assessment, longer than 60 minutes, available annually, for people with intellectual disability. There is also MBS item 699, an annual Heart Health Assessment for the same group.

The tool designed to go with it is the CHAP — the Comprehensive Health Assessment Program. It is Australian, free, evidence-based, and updated annually, with an adult version (2023) and a youth version for ages 12 to 18 (2024). It is a structured checklist that walks a GP through the systems that routinely get missed, including the cancer screening that is due.

This is the mechanism that catches the things diagnostic overshadowing hides. It is funded. The tool exists. It is a priority under the National Roadmap, which allocated $6.7 million to this area. Easy Read material explaining yearly health checks exists.

And it is massively underused, because booking it requires somebody to decide it is their job.

The Roadmap’s own position is that the health and disability sectors should share responsibility for ensuring these assessments are completed. In practice, shared responsibility means nobody’s responsibility unless a named person — a support coordinator, a family member, a house manager, a key worker — puts it in a calendar and chases it.

If you support someone with intellectual disability and they have not had an annual health assessment using the CHAP, that is the single highest-value phone call available to you. It is worth more than any amount of reading about barriers.

What this means depending on who you are

If you are a person with disability: you are entitled to the screening programmes, you are entitled to adjustments that make them possible, and you are entitled to be told your own results. If a clinic tells you a test cannot be done, ask what would make it possible — a different time, a longer appointment, a different room, a self-collected sample, an ultrasound instead of a mammogram. The alternatives in this guide are real and some of them are available now.

If you are a family member: the two things that change outcomes most are the annual health assessment and refusing to accept that a new symptom is “just the disability”. Ask one question: what else could this be? Then ask for it to be ruled out, and ask for the answer in writing.

If you are a support worker or coordinator: you are often the only person who notices the change. The weight loss, the new reluctance to eat, the blood nobody else saw, the hand that keeps going to the same place. Write it down with dates, and escalate it as a change rather than as a behaviour. A dated log of a symptom over six weeks is the most useful document a GP can receive.

If you are a clinician: the person in front of you may be at the end of a long chain of avoided appointments, and the symptom you are being shown may have been attributed to their disability several times already. The longer consultation items exist. So does item 707.