The map
Thirteen challenges.One chain.
Every primary health failure in the world breaks at one of five links. We map all of them the same way, so they can be compared.
Below: the chain we map against, the cascade for each major condition with the published numbers behind it, the field we use to decide what to work on first, and the ten challenges themselves.
The loss chain
A person with a treatable condition is lost at exactly one of five points. Naming which one turns a complaint into a budget line.
The person is never screened. No system has seen them. Nothing downstream exists yet.
A positive screen that never becomes a diagnosis. The commonest loss in camp-based programmes.
A diagnosis that never begins treatment. Distance, cost and a referral nobody owns.
Treatment begins and the outcome is never reached. Titration, supply and follow up.
The outcome is reached and not sustained. The step almost nobody measures at all.
A programme that reports how many people it screened has reported link one and nothing else. We do not accept a headcount as an outcome, in our own work or anyone else's.
Where each condition actually loses people.
Published figures only. Where a step has no credible global number, the bar is left undrawn and the gap is named. Select a condition.
Hypertension
Size of burden is not the same as size of opportunity.
We plot each challenge on two axes: how many people it affects, and how much of it can be closed at the primary level. We work in the top right. Select a point to read it.
Hover, tap or tab a point · scroll sideways on a narrow screen
Two challenges are deliberately not plotted. Epidemic and pandemic risk carries an episodic burden, so an annual figure on this axis would misrepresent it: a shock year and a quiet year are not the same number. Antimicrobial resistance is a multiplier on the treatment step of every other challenge here rather than a burden that stands alone. Both are covered in the list below. Cancer is plotted on annual incidence rather than prevalence, and the two are not the same measure.
Vertical axis: people affected, logarithmic. Sources as listed against each challenge below. Horizontal axis: Forum assessment of the share of the burden addressable at the primary level, August 2026. It is a judgement, and it is labelled as one.
The challenges, and the numbers behind them.
Three are system conditions that cap everything else. Nine are disease burdens. One, antimicrobial resistance, is a multiplier on all of them. Every figure names its publisher and its year.
Half the world is not fully covered by essential health services. Not underserved. Not partially served. Not covered.
The service coverage index rose from 54 to 71 points between 2000 and 2023, and the annual rate of progress fell from 1.5 percent before 2015 to 0.5 percent after it. On the current line the index reaches 74 by 2030, against a target of full coverage. Coverage is the ceiling on every other challenge in this list, because a cascade cannot start for a person no health system has ever seen.
- People not fully covered, 20234.6 bn
- Service coverage index, 202371 / 100
- Projected index by 203074 / 100
- Countries improving on both coverage and cost1 in 3
Care that exists and cannot be afforded is not care. Out of pocket payment is the quiet failure mode of every system on this list.
Two point one billion people face financial hardship from health payments, and one point six billion of them are pushed into poverty or deeper into it. The share fell from 34 percent to 26 percent since 2000, and the poorest populations still carry the heaviest load. Any redesign that improves detection without touching payment simply moves the loss one step later in the chain.
- Facing financial hardship, 20222.1 bn
- Pushed into or deeper into poverty1.6 bn
- Share of the world population26%
- Share in 200034%
The projected health worker shortfall is the hard constraint on everything else, and it is concentrated where the burden already is.
WHO projects a shortfall of 11.1 million health workers by 2030, with close to 70 percent of the gap in the African and Eastern Mediterranean regions. This is the reason a system design that assumes a doctor at every step will not survive contact with a district. Task shifting, screening at the community level and referral triage are not efficiency preferences. They are the only arithmetic that closes.
- Projected shortfall by 203011.1 m
- Share in two WHO regions~70%
- Africa: stock covering needs, 202243%
- Africa: projected coverage by 203049%
The largest treatable burden in the world, running on a medicine that costs cents, controlled in fewer than one in four people.
Cardiovascular disease accounts for at least 19 million deaths a year, the largest share of all noncommunicable disease deaths. Hypertension is its principal modifiable driver: 1.4 billion adults aged 30 to 79 live with it, 44 percent do not know, and 23 percent have it controlled. Nothing else on this list has so wide a gap between what treatment costs and what treatment delivers.
- Adults with hypertension, 20241.4 bn
- Unaware they have it44%
- Controlled23%
- Cardiovascular deaths a year19 m+
A condition detected with a finger prick and a strip, undiagnosed in four of every ten people who have it.
The number of people living with diabetes has more than quadrupled since 1990. Two hundred and fifty two million adults are undiagnosed. More than half of everyone with diabetes was taking no medication in 2022, and treatment coverage is lowest in exactly the countries where prevalence is rising fastest. Diabetes is also the reason eye and kidney screening cannot be treated as separate programmes.
- Adults with diabetes, 2024589 m
- Undiagnosed43%
- Deaths in 20243.4 m
- Global health expenditure, 2024US$1.0 tn
One billion of those cases were preventable or are still unaddressed, and the fix in most of them is a pair of glasses or half an hour of surgery.
Cataract accounts for 94 million cases of distance vision impairment, uncorrected refractive error for 88 million, and unaddressed presbyopia for 826 million. Prevalence of distance vision impairment in low and middle income regions runs about four times that of high income regions. Annual productivity loss is put at US$411 billion against an estimated US$24.8 billion to meet the unmet need.
- People with vision impairment2.2 bn
- Preventable or unaddressed1.0 bn
- Needing cataract surgery who get it1 in 2
- Cost of meeting the unmet needUS$24.8 bn
The clearest case of a burden that is counted and a cascade that is not.
Hearing loss reaches a projected 2.5 billion people by 2050, with more than 700 million needing rehabilitation. Ninety five million children aged 5 to 19 live with it now. The annual cost of leaving it unaddressed is put at almost one trillion dollars, and the cost of scaling ear and hearing care at under US$1.40 per person per year. What does not exist is a published global measure of identification, fitting and continued use.
- Living with hearing loss1.5 bn
- Projected by 20502.5 bn
- Children aged 5 to 1995.1 m
- Cost to scale servicesUS$1.40 pp/yr
Roughly one in seven people, treated adequately in fewer than one in ten cases of depression.
Anxiety and depressive disorders are the most common conditions and the second largest cause of long term disability. Access splits sharply by income: over 50 percent receive care in higher income countries against under 10 percent in low income countries. Median government spending on mental health has held at 2 percent of health budgets since 2017. Depression and anxiety alone are estimated to cost the world economy a trillion dollars a year in lost productivity.
- Living with a mental disorder1 bn+
- Adequately treated for depression9%
- Median share of health budgets2%
- Annual productivity costUS$1 tn
Curable, and still the world's deadliest single infectious agent, with a detection gap of 2.4 million people.
TB killed 1.23 million people in 2024. Incidence has fallen 12 percent since 2015 against an End TB milestone of 50 percent. A WHO recommended rapid molecular test was the initial diagnostic for only 54 percent of the 8.3 million people newly diagnosed. Of the 390,000 people who developed multidrug or rifampicin resistant TB, 42 percent accessed treatment.
- Fell ill in 202410.7 m
- Died in 20241.23 m
- Detection gap2.4 m
- Drug resistant cases treated42%
The second leading cause of death in the world, and the clearest case of survival depending on where a person was born.
Cancer claims more than 26,000 lives a day and cases are projected to reach nearly 35 million a year by 2050. Nearly four in ten cases are linked to preventable risk factors, particularly HPV, hepatitis B and C, helicobacter pylori, alcohol, tobacco, high body mass index and inactivity. That share is squarely primary care work: vaccination, screening and early detection. The treatment gap is starker still. Fewer than one in three countries include cancer care in their universal health coverage package, and 87 percent of women with breast cancer survive five years in high income countries against about 42 percent in low income countries.
- New cases a year, 202420.6 m
- Deaths a year~10 m
- Projected new cases by 2050~35 m
- Countries with cancer in UHC packagesunder 1 in 3
COVID showed what a shock does to a system with no slack. The slack has not been rebuilt.
Global life expectancy fell 1.8 years between 2019 and 2021, the largest drop in recent history, and the damage was not confined to the virus. Routine immunisation, screening and chronic disease follow up all stopped, and some of it has not restarted. In 2025 first measles dose coverage stood at 84 percent and second dose at 77 percent, against the 95 percent needed to stop outbreaks, and 57 countries reported large or disruptive measles outbreaks. Seven point three million infants received a first DTP dose and never received a measles dose, which is a drop out inside the system rather than a failure to reach them. Preparedness is not a parallel programme. It is routine immunisation, surveillance and a working first point of contact.
- Life expectancy lost, 2019 to 20211.8 yrs
- First measles dose coverage, 202584%
- Coverage needed to stop outbreaks95%
- Countries with large outbreaks, 202557
Not a burden of its own so much as a multiplier on every other burden in this list.
AMR was directly attributable to 1.14 million deaths in 2021 and associated with 4.71 million. On current trends that reaches 1.91 million attributable deaths a year by 2050, with 39.1 million cumulative deaths between 2025 and 2050 and South Asia carrying the largest share at 11.8 million. The modelling is unusual in how clearly it points at primary care: better care of severe infection and improved access to the right antibiotics could avert 92 million deaths over the same period, roughly eight times what a new gram negative drug pipeline would achieve. The intervention is diagnosis, stewardship and access, not discovery.
- Deaths directly attributable, 20211.14 m
- Deaths associated, 20214.71 m
- Projected attributable deaths, 20501.91 m
- Avertable by better infection care92 m
The oldest measured cascade in global health, and the one where progress has most clearly slowed.
Under five mortality fell 52 percent since 2000 and the rate of decline is slowing. Of the 4.8 million children who died before their fifth birthday in 2023, 2.3 million were newborns. Two hundred and sixty thousand women died in pregnancy and childbirth. WHO puts the cost of not correcting course at 700,000 maternal and 8 million child deaths between 2024 and 2030.
- Under five deaths, 20234.8 m
- Of which newborns2.3 m
- Maternal deaths a year260,000
- Maternal ratio per 100,000197
Every report we publish carries this section. It is not a disclaimer.
Three things on this page could be wrong, and we would rather say so here than be corrected later.
The cascades draw on different publishers with different base populations and different survey years, so the steps within one condition are not always strictly comparable. We have named the base for each and we will not silently blend them.
The horizontal axis of the priority field is our assessment, not a measured index. Reasonable people would place mental health further right than we have.
Global figures conceal district level variation that is often larger than the gap between countries. That is the reason our first published work is a district study and not a global one.
How we work →