A Healthy Country
Redesigning New Zealand’s approach to health from first principles
Executive Summary
New Zealand’s health system is good at fixing broken people. It is significantly less good at keeping people from breaking. The distinction carries a price tag that will, within two decades, be genuinely unpayable: diabetes diagnoses are projected to exceed 500,000 by 2044, up from around 228,000 today; dementia will cost the country $10.65 billion per year by 2050; cancer diagnoses are projected to rise from 30,000 per year to 45,000 by 2044; and the mental health burden already runs to roughly $21 billion annually, around 5% of GDP. All of this is happening while the system’s waiting lists grow, its workforce strains, and its prevention investment remains, despite 25 years of stated intention, marginal.
The crisis is not without a solution. It is without the right frame. The frame New Zealand needs is healthspan: not how long people live, but how many of those years are spent in genuine health. A system designed around that metric looks very different from what we have now. It invests earlier, intervenes upstream, treats the food environment as a public health question, and positions the doctor as a partner in maintaining vitality rather than a mechanic called when something finally breaks.
The tools to make that transformation are arriving fast. Artificial intelligence already outperforms clinicians in some diagnostic contexts. Personalised medicine is replacing one-size-fits-all pharmaceuticals calibrated to population averages. Longevity science is advancing toward what researchers call longevity escape velocity: the point at which life expectancy increases by more than one year per year of scientific progress. The people alive today will decide whether New Zealand, and New Zealanders, arrive at that threshold healthy.
This piece argues for a first-principles redesign of what New Zealand means by a health system, anchored in the evidence for what actually determines health outcomes and informed by what is now technologically possible. Inaction is not neutral. Every year the current frame persists, the cost of changing it compounds.
The Patient Who Became His Own Doctor
Eleven years ago, Peter Naylor received a diagnosis that most people would find devastating: a rare cancer, no known treatment protocol, and a statistical survival rate of 50% over ten years. The oncologist was honest and, in the way of the system, limited. There was nothing the health service could offer. Go home. Wait and see.
Peter did not go home and wait. He researched. He changed how he ate, how he moved, how he managed stress. He treated his body as a system to be understood and optimised rather than a patient to be monitored. Eleven years on, he is in better health than he was at diagnosis. His stated goal is to die with his cancer, not from it, at the age of 120, having spent most of that life in genuine health rather than managed decline.
That story is not an advertisement for alternative medicine, nor a dismissal of clinical expertise. It is something more uncomfortable: evidence that the system’s instinct, when it has nothing to offer, is to disengage rather than to support. Peter’s outcome was not despite the health system. It was largely independent of it. He became, of necessity, his own primary clinician.
The broader dynamic is not unusual. Every year, thousands of New Zealanders leave GP appointments with a prescription, a referral, or a waiting list number, but without a meaningful conversation about what they could change in their own lives to alter the trajectory of their condition. The system’s model is reactive: something goes wrong, the system responds. Prevention, lifestyle, environment, the upstream causes of disease, these sit outside the standard consultation’s frame.
What would a health system look like if it was designed not to manage disease but to support health? Not to fix people when they break, but to give them the tools, information, and structural support to not break in the first place? The question is not rhetorical. The technology and the evidence to answer it are both, for the first time, within reach.
Why This Question, Why Now
New Zealand’s health system performs, on many international measures, reasonably well. The Commonwealth Fund’s 2024 Mirror Mirror report, which benchmarks health systems across ten high-income countries, ranks New Zealand first in care process quality and among the top tier for health outcomes. Life expectancy at 82.3 years sits almost a year above the OECD average. On those numbers, you could argue the system is working.
The trend is less flattering. Government health expenditure fell from 7.4% of GDP in 2009 to 6.6% by 2018, recovering only partially since. Per capita, New Zealand spends the equivalent of US$3,929 on health, against Australia’s US$5,802 and Canada’s US$6,215. Cancer survival rates rank 22nd among OECD nations. In 2024, 73% of New Zealanders reported wait times for appointments were too long, up from 66% the year before, an accelerating deterioration. On equity, the Commonwealth Fund places New Zealand second-to-last among the ten countries surveyed: a signal that the system’s aggregate performance conceals serious disparities beneath.
The deeper pressure is not about the gap between New Zealand and its peers. It is about the gap between what the current system can sustain and what the next thirty years will demand.
Three forces are converging. The burden of chronic, largely preventable disease is accelerating: diabetes, dementia, cancer, obesity, and cardiovascular conditions all rising, all expensive, all driven substantially by lifestyle, environmental, and systemic factors the current model does not treat as within its remit. The mental health crisis, particularly among young New Zealanders, has reached a scale that defies incremental response. And a wave of technology, including AI diagnostics, personalised medicine, and longevity science, is changing what is possible while making the cost of delayed adoption increasingly legible.
Cancer deserves specific attention. New Zealand has the second-highest cancer incidence rate in the world, after Australia, at around 423 cases per 100,000 people. More than 30,000 New Zealanders are diagnosed each year; by 2044, that figure is projected to exceed 45,000. Yet five-year survival rates rank 22nd among OECD nations, a gap that reflects a system better at treating late-stage disease than catching it early. Screening coverage sits below target across all three national programmes: bowel cancer screening reaches 57% of those eligible, breast screening 70%, cervical screening 73%. The early detection failure is not primarily a clinical problem; it is a system design problem. A health system built around prevention and early intervention would treat those screening numbers as an emergency.
The reason none of this has been adequately addressed is not incompetence or indifference. It is a coordination failure at civilisational scale. Every actor in the system has behaved rationally within their own frame. Food manufacturers extended shelf life and improved palatability because it reduced waste and increased margin. Retailers optimised for what sold. Regulators approved substances one at a time, as their frameworks required, without accounting for cumulative exposure across an ultra-processed diet. Doctors treated the patient in front of them with the tools available. Governments funded the health system that existed rather than the one that might reduce demand in fifteen years. No single decision was wrong on its own terms. The harm is the aggregate of thousands of individually rational choices made without anyone tracking the systemic outcome, and with no mechanism requiring anyone to.
The most politically convenient response has been to assume that future medical intervention will fix what prevention has failed to address. There is enough partial truth in this to make it seductive: pharmaceutical innovation is genuinely impressive, and treatments like GLP-1 drugs are already changing diabetes management. But the assumption has a structural flaw. Medical intervention manages deterioration; it cannot restore the years of vitality already lost. And the cost of treating disease is socialised through the public health system while the profit from producing the conditions that cause it remains entirely private. No food manufacturer carries a line on their balance sheet for their contribution to New Zealand’s diabetes or dementia projections. Nobody is doing that accounting on behalf of the country.
New Zealand now has enough evidence to know where this leads. That changes the moral character of inaction. Not designing a harmful outcome is one thing. Knowing the outcome is coming, having the projections, and choosing to defer action to the next electoral cycle, or the one after, is something else. The coordination failure that produced the current crisis was nobody’s fault in particular. What happens from here is a choice.
Threading through all of this is a concept that most health policy discussion has not yet absorbed: longevity escape velocity, or LEV. Defined as the point at which advances in medical science extend life expectancy by more than one year per year of scientific progress, LEV is no longer theoretical. Aubrey de Grey, whose LEV Foundation ran the most ambitious combination anti-aging research yet attempted in late 2024, gives a 50% probability of reaching it within 12 to 15 years. Harvard professors David Sinclair and George Church project the mid-to-late 2030s. The implication for health policy is stark: people who maintain their health today are likely to live significantly longer than previous generations expected. Making healthspan the system’s primary design objective is no longer optional. It is overdue.
Current State: The Sickness System and Its Numbers
New Zealand spent approximately 7.1% to 7.2% of GDP on health in 2025, according to Budget projections. That figure is lower than comparable peer countries and skewed heavily toward treatment rather than prevention. Despite formal commitments to strengthening primary and preventive care dating back to 2001, there is little evidence that the prevention share of funding has grown as a proportion of total health spending. The gap is structural and persistent.
The system achieves solid results where it has invested. Life expectancy at 82.3 years is two years above the OECD average. Thirty-day mortality after acute myocardial infarction sits at 4.6%, below the OECD average of 6.5%. The Commonwealth Fund’s 2024 report places New Zealand first globally in care process quality. Cancer survival is a significant exception: five-year relative survival for all cancers ranks 22nd among OECD nations, shaped partly by late diagnosis and partly by inequitable access to treatment. The Commonwealth Fund equity ranking, second-to-last of ten, points to the same pattern: the system works better for some New Zealanders than for others, and the gap is large.
The key indicators and their international context are summarised below.
Sources: Commonwealth Fund Mirror Mirror 2024; OECD Health at a Glance 2025; World Bank health spending data; Te Aho o Te Kahu State of Cancer 2025; NZ Ministry of Health.
What the table does not capture is trajectory. The numbers that will define New Zealand’s health system over the next three decades are not in the current data; they are in the projections. And those projections are alarming.
The Burning Platform: Fiscal Projections and the Cost of Delay
The Trajectory If Nothing Changes
New Zealand currently has approximately 228,000 people living with Type 2 diabetes. By 2040, that number is projected to reach 400,000. By 2044, modelling published in the New Zealand Medical Journal projects more than 500,000, a near-doubling in just over two decades. The growth is not evenly distributed. Among Pacific peoples, prevalence is projected to reach 17% of the population by 2044, and a quarter of all Pacific New Zealanders are expected to carry a Type 2 diabetes diagnosis within twenty years. The cost implications, in direct treatment, complications management, and associated conditions, have not been fully modelled for the public record. The trajectory alone makes the scale plain.
Dementia is the sleeper crisis. Around 83,000 New Zealanders are living with dementia in 2025. By 2050, that number is expected to nearly double to 170,000, with new incident cases rising from around 22,000 per year now to 29,500 by 2040. The fiscal cost is projected at $10.65 billion annually by 2050 in inflation-adjusted terms. Every hour in 2025, three New Zealanders develop dementia. By 2050, that rate rises to four per hour. Budget 2026 was assessed by Alzheimers New Zealand as failing to meet the fast-growing challenge, despite these projections being well established.
Cancer sits alongside these conditions as a growing and inadequately addressed burden. New Zealand has the second-highest cancer incidence rate in the world, after Australia, at approximately 423 cases per 100,000 people. More than 30,000 people are diagnosed each year; by 2044, Te Aho o Te Kahu, the Cancer Control Agency, projects that figure will exceed 45,000. The 50% increase in annual diagnoses is driven partly by an ageing population, but also by the same lifestyle and environmental factors running through this section. Early-onset colorectal cancer has increased by 26% per decade among younger New Zealanders, a trend that points to causes beyond ageing. The early detection problem compounds everything: five-year survival ranks 22nd among OECD nations, and screening coverage sits below its own targets across all three national programmes. The gap between where screening is and where it needs to be is not primarily a funding problem; it is a system design problem. A health system oriented around early intervention would treat those numbers as unacceptable, and would be deploying every available tool, including AI-assisted detection, to close them.
Mental health is not a future problem; it is a current one. The annual economic cost of serious mental illness and addiction in New Zealand runs to around 5% of GDP, approximately $21 billion, by the estimates of the Mental Health and Wellbeing Commission. Workers with anxiety diagnoses lose more than 68 working days of productive output annually. The disease-related income loss for individuals diagnosed with mental health conditions averages around $5,000 per year. And behind the economic data sits a human crisis: New Zealand has the highest child and youth suicide rate among high-income countries, at nearly three times the OECD average for 15 to 19-year-olds. For child mental wellbeing overall, New Zealand ranks last of 36 countries with available data. Between 1 July 2023 and 30 June 2024, 617 New Zealanders died by suspected self-inflicted causes. Among Maori, the suicide rate is 16.3 per 100,000, against 9.0 for non-Maori: a disparity that reflects compounding disadvantage rather than any single cause.
The Hidden Cost of the Queue
Beyond the disease trajectory, there is a fiscal cost embedded in how the system manages time. As of September 2024, close to 188,000 New Zealanders were waiting for a first specialist assessment, with two-fifths waiting more than four months and over 8,000 waiting more than a year. A further 76,677 were waiting for elective procedures.
Those numbers represent more than clinical delay. For every person waiting six months for a procedure that restores mobility, addresses chronic pain, or resolves a condition preventing them from working, the calculation is not simply one of suffering: it is one of lost income, lost productivity, welfare costs to the state, carer burden, and the downstream clinical cost of complications that develop during the wait. A hip replacement deferred six months for a 55-year-old in a physical occupation is, in economic terms, a very different intervention from one deferred for a retiree. That difference should be quantified, publicly tracked, and treated as a cost, not an administrative statistic. Health officials estimated in 2024 that meeting planned care targets would cost $723 million over two years. The cost of not meeting those targets, in lost economic output alone, is not routinely calculated. It should be.
Productivity loss, however, is the least of it. The deeper cost is lives. For cancer, the relationship between wait time and outcome is not marginal; it is steep. A colorectal cancer detected at Stage I carries a five-year survival rate above 90%. The same cancer, detected at Stage IV because a referral was delayed or a screening appointment wasn’t available, drops below 10%. Every week on a waiting list is a week in which a cancer that could have been caught early is progressing toward a stage from which recovery becomes unlikely. In a country that already ranks 22nd in the OECD for cancer survival and runs all three national screening programmes below their own coverage targets, the waiting list is not just a system inefficiency. For some people, it is a death sentence written in administrative language.
In the year ending June 2024, 10,651 New Zealanders died from cancer, around 29 per day. Te Aho o Te Kahu has documented that the survival gap between New Zealand and peer countries is widest at the late-stage diagnosis end, where delays in detection are most consequential. Closing even a third of that gap through earlier detection would translate to hundreds of lives saved each year. When the system reports on waiting times as a performance metric, it is counting the people in the queue. It is not counting the people for whom the queue ran out.
Sources: NZMJ diabetes projections 2025; Alzheimers NZ 2025; Te Aho o Te Kahu State of Cancer 2025; Mental Health & Wellbeing Commission BIM 2025; Mental Health Foundation; UNICEF 2025; Health NZ September 2024.
The Engineered Environment: Food, Addiction and Regulatory Failure
Around 50 to 60% of New Zealand’s food energy comes from ultra-processed foods (UPF): products manufactured through industrial processes, combining cheap ingredients with additives designed to extend shelf life, enhance palatability, and encourage overconsumption. Among New Zealand adolescents, the figure is higher still; research published in 2026 found that UPF accounts for 64.8% of their daily energy intake.
The health consequences are well-documented. UPF consumption is strongly associated with rising obesity, Type 2 diabetes, cardiovascular disease, and depression. The mechanisms include not just poor nutritional profiles but the physiological effects of engineered palatability: these products are designed, with considerable scientific sophistication, to disrupt the normal satiety signalling that regulates food intake. Framing this as a matter of individual choice misrepresents the problem. A person eating in a food environment engineered for overconsumption is not exercising free choice; they are responding to incentives constructed by people with strong commercial interests in the outcome.
New Zealand has not had a national adult nutrition survey in 18 years, or a children’s survey in 23 years. The evidence base for food policy is operating on data from a fundamentally different food environment. That gap is itself a choice.
The regulatory picture mirrors tobacco in its structural dynamics: an industry producing documented harm, generating substantial revenue, and maintaining political influence sufficient to slow meaningful reform. Food labelling, advertising restrictions (particularly to children), and supply chain standards all fall well short of what the evidence on UPF harms would justify.
The Link to the Fiscal Projections
The reason this matters beyond the immediate diet-health association is its direct connection to the disease trajectory in the previous section. Type 2 diabetes is not an inevitable condition. Evidence from the DiRECT trial, replicated in multiple subsequent studies, shows that intensive dietary intervention achieves remission in roughly a third of patients at 24 months, without medication. The 2025 ICD-10 classification now includes a formal category for Type 2 diabetes remission (E11.A), recognising lifestyle-achieved remission as a clinical outcome.
The picture on dementia is comparably striking. The 2024 Lancet Commission on Dementia, the most comprehensive review of the evidence yet produced, identified 14 modifiable risk factors accounting for 45% of all future dementia cases, a 5-point increase on the Commission’s 2020 estimate. The factors include diet, physical inactivity, obesity, hypertension, excessive alcohol, diabetes, and social isolation. Every one of them is addressed, at least in part, by changing the food and lifestyle environment. If 45% of projected dementia cases can be delayed or prevented through population-level intervention, the $10.65 billion annual cost projected for 2050 is not a fixed number. It is a choice.
Effective food policy is the single highest-leverage upstream intervention available to New Zealand to alter the cost trajectory laid out in the previous section. Treating the food environment as a public health infrastructure question, rather than a consumer preference, is the precondition for changing those numbers.
The Invisible Burden: Environmental Health
Not all health burdens are chosen. Microplastics, pesticide residues, and food additives represent a category of exposure that is involuntary, cumulative, and only recently becoming legible through the research literature.
Microplastics are now detectable in human blood, breast milk, placental tissue, and brain tissue. Research published in 2025 documents their capacity to generate oxidative stress in neuronal cells, disrupt neurotransmitter systems critical for cognitive function, and trigger metabolic pathway disruption in liver tissue. Animal studies showing neurodevelopmental abnormalities in offspring exposed in utero are generating concern about long-term population-level cognitive effects. The research is not yet conclusive at the level of specific disease causation, but the precautionary case for regulatory action is strong, and the commercial and agricultural context of New Zealand, intensive farming, treated waterways, heavy reliance on plastics through the food supply chain, places its population in a specifically exposed position.
Pesticide residues present a related profile: widespread low-level exposure through food and environment, poorly understood cumulative effects, and a regulatory framework calibrated primarily to acute toxicity rather than chronic, low-dose outcomes. The connection to metabolic and neurological conditions is an active area of research, with the weight of emerging evidence supporting concern rather than reassurance.
Food additives form the third category. Many are approved under frameworks decades old, designed to assess individual substances at specified doses rather than the cumulative effect of dozens of additives consumed daily across an ultra-processed diet. The science has moved faster than the regulation, and the gap creates involuntary population-level exposure.
The argument for regulatory intervention here is not that individuals cannot reduce their personal exposure through careful choices. It is that population-level exposure is determined by industrial practices, and that the health cost of those practices falls on the public while the economic benefit accrues to producers. Environmental health is a collective problem requiring collective solutions, and it belongs in the health system’s frame of reference.
Mental Health: The Epidemic We Keep Managing
New Zealand’s mental health crisis has been described, acknowledged, inquired into, and underfunded for long enough that the cycle of description and inquiry has become a substitute for structural response. The data has not moved in the direction of improvement.
The annual economic cost of serious mental illness and addiction sits at around 5% of GDP, approximately $21 billion, by the estimates of the Mental Health and Wellbeing Commission: higher than comparable countries and rising. Workers with anxiety diagnoses lose more than 68 working days of productive output annually. The income loss associated with mental health conditions averages around $5,000 per year for those diagnosed, a number that compounds over careers.
Behind the economic figures is a more urgent human one. New Zealand has the highest child and youth suicide rate among high-income countries, at nearly three times the OECD average for 15 to 19-year-olds. For overall child mental wellbeing, New Zealand ranks last of 36 countries with available data. Between mid-2023 and mid-2024, 617 New Zealanders died by suspected self-inflicted causes. Among Maori, suicide rates run at 16.3 per 100,000 against 9.0 for non-Maori: a disparity that reflects compounding disadvantage rather than any single cause, and that should make plain the inadequacy of treating this as a clinical problem alone.
The system’s response to mental health has been, structurally, identical to its response to physical health: wait for crisis, then treat. Mild to moderate conditions are frequently unaddressed until they deteriorate. Access to psychological services is gated by cost and wait time in ways that primarily affect those with fewer resources. The upstream conditions producing anxiety, disconnection, and despair at scale, including economic precarity, social isolation, and the documented mental health effects of heavy social media use among adolescents, are not currently treated as health system problems. They are.
A first-principles health system would treat mental wellbeing as a product of social conditions and individual resilience, not simply of medication compliance. The investment needed is upstream: in communities, in economic security, in schools, and in reversing the social conditions that produce crisis. That is a health system commitment, not a soft policy add-on.
A First Principles Framework: Designing a Health System
If New Zealand were designing a health system from scratch, unburdened by existing institutions, funding arrangements, and professional incentives, what would a coherent and principled framework look like? Seven pillars suggest themselves, each representing a design choice the current system either fails to make or makes only partially.
Pillar One: Prevention as the Primary Investment
A genuine health system spends its largest resources on keeping people well, not on treating them once they are sick. Incentive structures should reward health outcomes rather than treatment volume: GPs measured on their patients’ metabolic health over time, not on consultation throughput; community health investments evaluated against population wellbeing data; and a prevention budget that represents a meaningful share of total health spending rather than a rounding error. The evidence that prevention costs less than treatment is not in dispute. What is missing is the political will to restructure a system that has organised its workforce and institutions around the current model.
Pillar Two: The Doctor as Health Partner
The GP’s role needs redesigning around wellbeing, not ailment management. Longer consultations with space for lifestyle conversation, digital tools giving patients access to their own health data between appointments, and a professional culture that treats health coaching as core medical practice are the structural changes required. The system should also extend this model to support people who, like Peter Naylor, want to take ownership of their own health trajectory: providing education, tools, and clinical partnership rather than disengaging when no standard treatment is available.
Pillar Three: Food System Regulation as Public Health Infrastructure
Regulating the food environment is not an intrusion on consumer freedom. It is the recognition that the food environment is already regulated, by commercial decisions made in the interest of shareholders rather than consumers. Minimum baseline measures include mandatory front-of-pack labelling reflecting nutritional reality, restrictions on UPF marketing to children, supply chain standards reducing the most harmful additives, and a national nutrition survey updated on a regular cycle. These are not radical interventions; they are the calibration of regulation to current evidence.
Pillar Four: Health Literacy and Education
From school age through life, New Zealanders should have access to practical education in nutrition, cooking, movement, sleep, and stress management. The current curriculum gives these topics minimal space. A health system designed around prevention would treat health education as foundational rather than supplementary, and would fund community health literacy programmes reaching people outside formal education settings.
Pillar Five: Environmental Health Standards
New Zealand’s regulatory framework for microplastics, pesticide residues, and food additives needs updating to close the gap between the current evidence on chronic, cumulative health effects and standards calibrated to older science. Precautionary regulation in these areas is the appropriate response when the evidence base is moving faster than the frameworks designed to assess it.
Pillar Six: Mental Health Upstream
Treating mental health as a downstream clinical problem is, at population scale, an expensive failure. The upstream investments that produce mental wellbeing, including economic security, community connection, social cohesion, and reduced digital-environment harms for adolescents, are health system concerns. A health system budget should include explicit investment in the social conditions that determine mental health at scale.
Pillar Seven: Individual Agency, Structurally Supported
The goal is not to mandate behaviour but to make the healthy choice the easy choice. Walkable communities, affordable whole foods, accessible health information, and a clinical system that supports rather than ignores patient-led health management are the features of a system designed around agency rather than compliance. The structural work is to remove the barriers that make the unhealthy choice the path of least resistance for the people with the fewest resources.
Healthspan, Not Lifespan: The Right Metric
Every major health system currently reports its performance primarily in terms of life expectancy. The metric is intuitive, comparable across countries, and politically legible. It is also the wrong metric, and choosing it shapes everything the system tries to optimise for.
Life expectancy measures survival. Healthspan measures vitality: the number of years spent in good health, free from significant disease or disability. The gap between the two is where the real story lives. New Zealand’s life expectancy of 82.3 years does not tell you how many of those years are spent managing chronic conditions, in pain, cognitively declining, or unable to live independently. Across comparable populations, the answer is substantial, and it is growing as populations age into multimorbidity without the upstream investment that might have prevented it.
A system designed around healthspan asks different questions. Not: how long will people live? But: how many years will people be genuinely well? Not: did we avoid this death? But: did we support this person to maintain vitality through the decades of their life when that was still achievable? The shift in metric changes what gets funded, what gets measured, what clinical success looks like, and how the system’s performance is held to account.
Peter Naylor’s framing captures it precisely: the goal is to die with his cancer at 120, having had most of those years as healthspan. That is not an eccentric ambition. It is the logical target for a health system designed around the right objective.
The connection to longevity escape velocity makes this urgent rather than merely desirable. If LEV is reached within the next 15 to 20 years, the people who arrive at that threshold in good metabolic health will benefit from subsequent advances in ways that those managing multiple chronic conditions will not. Healthspan investment now is, in this framing, a civilisational-scale decision: it determines who gets to take full advantage of one of the most significant transitions in human history. New Zealand should be designing its health system with that possibility in view.
Technology: The Convergence and the Cost of Being Late
AI in Diagnosis and Early Detection
Artificial intelligence is already changing diagnostic medicine, and the case for early adoption is straightforward: later diagnosis produces worse outcomes and higher treatment cost across virtually every disease category. A 2025 systematic review found that AI assistance achieved pooled sensitivity of 0.79 and specificity of 0.87 in cancer diagnosis from imaging, against 0.67 and 0.82 for unaided clinicians. In breast cancer screening and lung cancer detection, AI tools are performing at or above expert radiologist level for specific detection tasks, catching lesions that might otherwise go undetected until a later, more expensive, and less survivable stage.
The detection advances are specific and significant. Google’s LYNA (Lymph Node Assistant) system detects metastatic breast cancer in lymph node biopsies with an accuracy rate of 99%, outperforming pathologists working without AI support. NHS trials in England found AI-assisted breast screening caught 13% more cancers than standard double-reading by radiologists alone. Grail’s Galleri multi-cancer early detection blood test, now in clinical use in several countries, can detect over 50 types of cancer from a single blood draw before symptoms appear, with a false positive rate below 1%. For New Zealand, where bowel screening reaches only 57% of eligible people and where the cancer survival gap with peer countries is concentrated at late-stage diagnosis, AI-powered screening and detection tools represent the most direct available path to closing that gap without waiting a decade to rebuild clinical capacity.
Peter Naylor’s experience is directly relevant here. A rare cancer with no standard treatment protocol is precisely the context in which AI-assisted pattern recognition, trained on datasets vastly larger than any individual clinician could consult, changes what is possible. The value of AI is not only in catching common cancers earlier; it is in the long tail of rare and unusual presentations where the information available to a single practitioner is inherently limited. Every year New Zealand delays adopting AI-assisted diagnostics at scale is a year of avoidable late-stage diagnoses, in both common and rare presentations. Given that New Zealand already ranks 22nd in OECD for cancer survival, the opportunity cost of delay is measured in lives.
The economic case reinforces the clinical one. Treating a Stage I colorectal cancer costs approximately NZ$15,000 to $25,000. Treating Stage IV colorectal cancer costs $150,000 to $250,000 or more, with outcomes far worse. An AI-assisted screening programme that shifted even 10% of late-stage diagnoses to early-stage would generate savings that dwarf its implementation cost, while delivering survival gains the system cannot currently achieve any other way. Early adoption is not a speculative bet; it is the highest-return investment available to New Zealand’s health system right now.
The case extends equally to chronic disease management. AI tools that monitor patterns in patient data over time, flagging early indicators of metabolic deterioration, cardiovascular risk, or cognitive decline, offer the possibility of intervention well before the clinical threshold of disease is reached. A system that catches Type 2 diabetes at the pre-diabetic stage, or identifies cardiovascular risk a decade before a cardiac event, is doing prevention in the most economically rational way possible. These tools exist. The question is how quickly New Zealand puts them to work.
Robotics and Automation in Care Delivery
Surgical robotics is already in clinical use and expanding. Precision advantages in minimally invasive procedures reduce complications, shorten recovery, and improve outcomes, particularly in surgical domains where the margin for error is small. In the context of New Zealand’s workforce constraints, robotic systems change the equation on surgical capacity: a skilled surgeon operating with robotic assistance can achieve outcomes that would otherwise require teams or that carry higher risk in under-resourced settings. In aged care and rehabilitation, automation offers a different kind of value: enabling more people to maintain independence for longer, reducing the transition to residential care that is both expensive for the health system and unwanted by most people who face it.
The Connected Transport System
The autonomous vehicles piece in this series covered accident and injury reduction in detail. The relevant connections here are dual: reduced road trauma lowers acute care demand and disability burden, and the urban form changes enabled by autonomous transport create environments more conducive to the active transport that is a documented health determinant. Health outcomes and transport infrastructure are not separate domains.
Personalised Medicine: The End of the Average Patient
Current pharmaceutical development is calibrated to populations, not individuals. A drug approved against a particular cancer or condition is assessed for its average effect across the trial population. That average conceals wide variation: the patient whose biology makes them a strong responder and the patient for whom the drug does nothing, or worse, are both represented in the mean. Some chemotherapy regimens achieve response rates of around 40% in clinical practice, meaning a substantial majority of patients are receiving treatments with significant side effects and limited personal benefit.
Pharmacogenomics, the study of how individual genetic variants affect drug response, is beginning to change this. Research published in 2025 shows that over 90% of the general population carries at least one actionable pharmacogenomic variant that meaningfully alters their response to one or more common medications. Matching treatment to individual genetic profile is no longer theoretical; it is a clinical reality in an expanding range of therapeutic areas, improving both efficacy and the reduction of unnecessary side effects.
CAR-T therapy represents the headline example of personalised medicine in practice. Rather than administering a generic pharmaceutical, CAR-T therapy engineers T cells from the patient’s own immune system to target their specific cancer. In blood cancers, response rates in clinical trials have reached 72 to 87%, with substantially reduced systemic toxicity compared to conventional chemotherapy, because the treatment is matched to the person rather than calibrated to a population average. Expanding access to personalised medicine is both a clinical opportunity and an argument for investing in the genomic and diagnostic infrastructure that makes personalisation achievable at scale.
Longevity Escape Velocity: The Approaching Threshold
Longevity escape velocity is the subject of serious scientific research, funded at scale, generating results in animal models, and projected by credible researchers at major institutions to be achievable within the lifetimes of people alive today. Aubrey de Grey, whose LEV Foundation ran the most ambitious combination anti-aging therapy trial yet attempted in late 2024, gives a 50% probability of reaching LEV within 12 to 15 years. Harvard’s David Sinclair and George Church project the mid-to-late 2030s.
The health system implication is not that everyone should plan to live to 150. It is that people who maintain their metabolic health now will be better positioned to benefit from advances in longevity medicine as they arrive. A person who reaches their late sixties without diabetes, cardiovascular disease, or significant cognitive decline is a very different candidate for emerging longevity therapies than someone managing five chronic conditions. Healthspan investment now is, in this framing, also longevity access investment. The two objectives are the same.
Scale-Up Scenarios
The table below models three tiers of health system reform. Each tier builds on the previous. The downstream cost impacts reflect the general direction of the evidence rather than audited fiscal projections; the purpose is to illustrate the relationship between ambition and outcome rather than to make precise budget claims.
Political Sustainability
Health system reform in New Zealand has a history of structural disruption without cultural change: reorganisations that rearrange the institutions while the underlying model of sickness management persists. Making the shift to a genuine health system durable across election cycles requires more than policy; it requires a political case that spans the usual dividing lines.
The cross-party value alignment for prevention-focused health investment is stronger than it might appear. Fiscal conservatives gain from a model that demonstrably costs less over time than treating preventable disease at the acute end. Social investment advocates gain from the evidence that healthier populations are more productive, more connected, and carry lower long-term welfare costs. Individual freedom advocates gain from a system that supports people in taking ownership of their health rather than mandating behaviour.
Structural durability requires institutional design as much as political agreement. Publishing healthspan as a national metric alongside life expectancy creates accountability that persists across governments: a future administration that allows healthspan to deteriorate is arguing against a visible, tracked number. Early investment in AI diagnostic infrastructure, genomic capacity, and food system regulation creates sunk-cost momentum that is genuinely difficult to reverse. Community health investment, once made, creates local stakeholders with a direct interest in maintaining it.
The tobacco analogy is instructive. Regulation of the tobacco industry was resisted for decades on the grounds of personal freedom, economic disruption, and regulatory overreach. The eventual consensus, reached through the accumulation of evidence and the rising political cost of denial, produced durable policy change. The food system is two to three decades behind that arc, but travelling it. The governments that move early will occupy the advantageous political ground when consensus catches up.
Conclusion: What Dying Well Looks Like
Peter Naylor’s goal is not unusual in its ambition. It is unusual in its clarity: to die with his cancer, not from it, at 120, having spent most of those years in genuine health. He is not waiting for the system to make this possible. He is building the conditions for it himself, with the knowledge and the agency that most people are not currently supported to develop.
The gap between the life that is possible for someone with the information, the motivation, and the resources to take ownership of their health, and the life that is probable for someone navigating an engineered food environment, a reactive health system, and a social environment that produces anxiety and disconnection at scale, is the problem this piece has set out to name.
The evidence is clear that a large proportion of the disease burden projected to overwhelm New Zealand’s health system over the next three decades is not inevitable. It is the product of choices: in food regulation, in prevention investment, in the metrics by which the system measures its own success, and in the model of medicine that treats patients as problems to be solved rather than people to be supported.
The technology arriving in the next decade makes the case for changing those choices more urgent, not less. AI diagnostics, personalised medicine, and longevity science are tools that amplify the benefit of being healthy. They do not rescue the unhealthy from the consequences of a system that waited too long to act. The people who arrive at the longevity frontier in good health will benefit from what science delivers there. Those who arrive managing five chronic conditions, having spent their health in a system that waited for them to break, will not.
The distinction between what produced this situation and what perpetuates it matters enormously. A society that stumbled into a health crisis through uncoordinated decisions made without full information is a society that made mistakes. A society that looks at the projections, understands the trajectory, and defers action anyway, hoping medicine will manage the consequences, or that a future government will take the harder road, is making a different kind of decision. The odds being stacked against a healthy life did not require anyone’s malice. But knowing they are stacked, and choosing not to change them, is something we cannot claim innocence about.
Longevity escape velocity is approaching. Healthspan is the investment that determines who benefits from it. Changing the system now, from one designed around sickness to one designed around health, is not expensive relative to the alternative. The alternative is a fiscal and human crisis that compounds every year action is deferred, and a generation that arrives at one of history’s most significant thresholds in no condition to take advantage of it.
New Zealand has the evidence, the technology access, and the governance capacity to make this shift. What has been missing is the frame. A healthy country does not manage disease. It cultivates health. The difference between those two sentences is the difference between the system we have and the system we need.
References
1. Commonwealth Fund. Mirror, Mirror 2024: A Portrait of the Failing U.S. Health System. September 2024. https://www.commonwealthfund.org/publications/fund-reports/2024/sep/mirror-mirror-2024
2. OECD. Health at a Glance 2025: New Zealand Country Profile. 2025. https://www.oecd.org/en/publications/health-at-a-glance-2025_15a55280-en/new-zealand_2154f240-en.html
3. Hobbs M et al. Projected increases in the prevalence of diabetes mellitus in Aotearoa New Zealand, 2020-2044. New Zealand Medical Journal, 2025. https://nzmj.org.nz/journal/vol-138-no-1608/projected-increases-in-the-prevalence-of-diabetes-mellitus-in-aotearoa-new-zealand-2020-2044
4. Alzheimers New Zealand. Facts and Figures 2025. https://alzheimers.org.nz/explore/advocacy/facts-and-figures/
5. Alzheimers New Zealand. Budget 2026 fails to meet fast-growing dementia challenge. 2026. https://alzheimers.org.nz/news/budget-2026-fails-to-meet-fast-growing-dementia-challenge/
6. Livingston G et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet, 2024. https://www.thelancet.com/commissions-do/dementia-prevention-intervention-and-care
7. Te Aho o Te Kahu. The State of Cancer in New Zealand 2025. Cancer Control Agency, December 2025. https://teaho.govt.nz/index.php/CanStats/reports/state-cancer-new-zealand-2025
8. Te Hiringa Mahara Mental Health and Wellbeing Commission. Briefing to the Incoming Minister of Health, February 2025. https://www.mhwc.govt.nz/assets/Accountability-documents/BIM/2025/Briefing-to-the-Incoming-Minister-of-Health_Feb-2025.pdf
9. Mental Health Foundation of New Zealand. Statistics on suicide in New Zealand. https://mentalhealth.org.nz/suicide-prevention/statistics-on-suicide-in-new-zealand
10. UNICEF New Zealand. New global data: New Zealand ranks alarmingly low for child wellbeing. 2025. https://www.unicef.org.nz/media-releases/new-global-data-new-zealand-ranks-alarmingly-low-for-child-wellbeing
11. Turley M et al. Ultra-processed foods contribute two thirds of energy intake in New Zealand adolescents. ScienceDirect, 2026. https://www.sciencedirect.com/science/article/pii/S0271531726000618
12. Thow A et al. Ultra-processed foods and their inputs increasingly dominate New Zealand’s food and beverage imports. PMC, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC13097854/
13. Stanford Medicine. Microplastics and our health: what the science says. January 2025. https://med.stanford.edu/news/insights/2025/01/microplastics-in-body-polluted-tiny-plastic-fragments.html
14. Lean MEJ et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT). The Lancet, 2018. American College of Lifestyle Medicine review 2025. https://lifestylemedicine.org/type-2-diabetes-remission/
15. Khodotskaya A et al. The role of pharmacogenomics in personalized medicine. PMC, 2025. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12836009/
16. Rodriguez JA et al. CAR-T cell therapy for cancer: current challenges and future directions. Nature Signal Transduction and Targeted Therapy, 2025. https://www.nature.com/articles/s41392-025-02269-w
17. LEV Foundation. Longevity Escape Velocity research programme. https://www.levf.org/
18. CEO Today. Dr Aubrey de Grey’s Longevity Escape Velocity: When Will Humanity Outrun Aging? 2025. https://www.ceotodaymagazine.com/2025/08/dr-aubrey-de-greys-longevity-escape-velocity-when-will-humanity-outrun-aging/
19. Health New Zealand. Waiting list and elective procedure data, September 2024. https://info.health.nz/about-us/what-we-do/planning-and-performance/health-targets
20. Easton B. How well is the New Zealand healthcare system doing? An international comparison. March 2025. https://www.eastonbh.ac.nz/2025/03/how-well-in-the-new-zealand-healthcare-system-doing-an-international-comparison/
21. Doran C et al. The economic cost of suicide and non-fatal suicide behaviour. MATES in Construction NZ, 2024. https://mates.net.nz/wp-content/uploads/2024/09/Doran-2024-The-economic-cost-of-suicide-and-non-fatal-suicide-behaviour_web.pdf
22. Google Health. LYNA: AI-assisted lymph node tumour detection. https://health.google/caregivers/ehr/lyna-ai-metastatic-breast-cancer/
23. McKinney SM et al. International evaluation of an AI system for breast cancer screening. Nature, 2020. https://www.nature.com/articles/s41586-019-1799-6
24. Grail. Galleri multi-cancer early detection test. https://www.galleri.com/