Set as preferred source
Insights from WHO'so Wrld Health Statistics 2026.
Global health is still moving forward. That is easy to miss.
New HIV infections fell 40% between 2010 and 2024. Tuberculosis incidence is down. More people have access to safe water and sanitation. Maternal and child mortality continue to improve.
But the gains are becoming harder to produce.
Malaria incidence has moved in the wrong direction. Routine childhood immunization remains below global targets. Progress on maternal mortality has slowed. And of the health-related Sustainable Development Goal indicators with numeric targets reviewed by WHO, not one is currently on track at the global level.
Universal health coverage tells the same story. Coverage is still expanding, but between 2015 and 2023 the pace of improvement fell to roughly one-third of what it had been between 2000 and 2015.
The important signal is therefore not that healthcare has stopped working.
It is that progress is becoming harder to buy.
For much of the past two decades, expanding proven interventions produced relatively predictable gains: vaccinate more children, extend sanitation, improve maternal care, expand treatment access.
That model is not disappearing. But another constraint is becoming much harder to ignore.
Healthcare increasingly depends on whether the system around the treatment has enough money, people, data and operating capacity to deliver it consistently.
The easy expansion phase is ending
The economics have changed first.
During the pandemic, healthcare moved rapidly up government spending priorities. In several regions, that momentum has already reversed. Government health spending as a share of total public spending peaked in 2021 in Africa, Europe and the Americas, then declined for two consecutive years. WHO also estimates that official development assistance for health was 30%–40% lower in 2025 than in 2023.
Those numbers matter because healthcare capacity is unusually difficult to switch off and rebuild.
Hospitals cannot replace experienced clinicians overnight. Public-health systems cannot restore lost surveillance capacity with a procurement order. Training pipelines take years. Infrastructure has fixed costs. And many health services cannot simply be postponed without creating a larger problem later.
WHO is already seeing the workforce effect. In its assessment of 108 low- and lower-middle-income countries, 63% reported employment impacts after reductions in development assistance, including job losses, salary suspensions and temporary leave.
At the same time, the demand side is not becoming easier.
Populations are aging. Chronic diseases require continuous management rather than one-time intervention. Reaching the remaining unvaccinated population is more difficult than reaching the first 80% or 90%. Antimicrobial resistance makes familiar treatment pathways less reliable.
Healthcare is being asked to do more complicated work while the resources supporting that work become less predictable.
That is a very different operating environment from simple expansion.
The workforce problem is not only about how many people exist
Healthcare workforce shortages are often discussed as a global headcount problem.
The WHO data suggests something more complicated.
Europe has around 80 nurses and midwives per 10,000 people. The Americas have around 71. The global average is about 40.5. Africa has less than one-fifth of Europe’s density. Physician distribution is even more uneven: around 40 doctors per 10,000 people in Europe compared with 3.1 in Africa.
Yet the global nursing workforce is relatively young. WHO reports roughly 174 nurses under 35 for every 100 nurses aged 55 or older.
That sounds reassuring until geography enters the picture.
People do not move automatically to where healthcare demand exists. Training capacity, compensation, migration policy, working conditions and public budgets determine where clinicians actually practice.
The real constraint is therefore not simply talent scarcity.
It is talent location.
That distinction matters for health-system leaders because adding a service, opening a facility or deploying new technology does not automatically create delivery capacity.
A hospital may be able to finance a new platform and still lack the clinicians required to use it effectively. A country may increase nominal service coverage while rural or underserved regions remain difficult to staff.
In the next phase of healthcare, workforce availability becomes part of almost every strategic decision.
COVID exposed a difference between coverage and resilience
Before the pandemic, it was easy to assume that countries with strong healthcare infrastructure and broad coverage were inherently resilient.
COVID made that assumption harder to defend.
WHO estimates 22.1 million excess deaths associated with the pandemic between 2020 and 2023, compared with roughly 7 million officially reported COVID-19 deaths. By 2021, almost a decade of global progress in life expectancy and healthy life expectancy had been erased.
The damage was not neatly divided between “developed” and “developing” systems.
The Americas experienced the largest initial decline in life expectancy. But once mortality is adjusted for population age, South-East Asia shows the highest regional excess mortality in 2021, while Africa and the Eastern Mediterranean also look harder hit than crude death rates suggest.
That matters because headline health indicators can hide very different kinds of weakness.
A system can have high baseline coverage but little room to absorb a sudden surge. Another can appear healthier in aggregate simply because its population is younger.
Resilience is therefore not the same thing as normal-period performance.
For healthcare leaders, the relevant question becomes less comfortable:
How much pressure can the system absorb before the operating model begins to fail?
COVID was an extreme test. The same question now applies at smaller scale to staffing shocks, supply shortages, cyber incidents, disease outbreaks and sudden demand spikes.
Healthcare can be highly digital and still operate with poor visibility
One of the most surprising findings in the WHO report has almost nothing to do with new medicine.
It is about how little the system sometimes knows about itself.
At the end of 2025, only 18% of WHO Member States had reported mortality data no more than one year old. Another 32% had never submitted cause-of-death data to WHO.
The gap becomes even clearer when the data is followed through the system.
WHO estimates that roughly 61 million people died globally in 2023. Around 21 million of those deaths reached WHO with cause-of-death information. About 15 million had detailed ICD coding. Once vague and ill-defined classifications were removed, only around 12 million deaths—roughly one-fifth of the global total—contained meaningful detailed cause-of-death information.
That is more than a statistical problem.
It is a visibility problem.
If changes in disease burden become visible years late, resources respond years late. If causes are classified inconsistently, prevention priorities become harder to set. If clinical, civil-registration and population data cannot move reliably between systems, leaders are forced to fill the gaps with surveys, estimates and models.
Healthcare can therefore become more digital without becoming proportionally more informed.
The app is only one part of the problem.
The harder work increasingly sits underneath it.
Some countries are already rebuilding that layer
The examples WHO highlights are interesting precisely because they are not primarily stories about new patient apps.
They are stories about connecting the system.
Chile has integrated civil registration into a broader digital-government and health-data framework, including daily information exchange between the Ministry of Health and the civil registry.
Malaysia has combined legal and institutional reform with stronger interagency data exchange, digital tools and improvements in mortality-data quality.
Morocco moved from a largely centralized model toward a national web-based reporting platform with decentralized data entry, automated coding and real-time quality controls. Coverage increased while the share of poorly defined causes of death declined.
The specific systems differ, but the direction is similar.
Reliable healthcare data increasingly depends on interoperability, common classifications, digital certification, strong identifiers, automated quality controls and systems that can exchange information across institutional boundaries.
WHO explicitly describes stronger digital-health infrastructure in terms of interoperable platforms, secure data architecture and scalable systems capable of producing reliable information continuously—even during instability.
That is a very different definition of digital health from simply putting another interface in front of the patient.
The next pressure may look less like a crisis
The pandemic was impossible to ignore because the shock arrived quickly.
Some of the pressures now building inside healthcare are harder to see because they arrive gradually.
Chronic disease is one. Aging is another. Antimicrobial resistance makes diagnostics and surveillance more important as some familiar therapies become less reliable. Reaching the final gaps in vaccination and essential care tends to require more effort than reaching the early majority.
None of these needs to produce a dramatic global event to change healthcare economics.
They can do it quietly.
More complex patients require more clinician time. Harder-to-reach populations increase delivery costs. Poor data creates additional coordination work. Drug resistance makes treatment and diagnosis more demanding. Workforce shortages turn routine bottlenecks into persistent queues.
The danger is that chronic pressure can begin to look normal.
And once it does, health systems risk optimizing around the bottleneck rather than removing it.
What health leaders should watch now
The WHO report does not provide a management playbook, and it should not be read as one. But the direction of travel points to several signals worth watching:
- Capacity per worker, not just workforce size. Headcount matters, but so does how much clinical time is lost to administrative work, fragmented systems and avoidable handoffs.
- The gap between digital adoption and usable data. More software does not guarantee better visibility. Timeliness, interoperability, classification quality and data ownership increasingly matter.
- The durability of healthcare financing. Systems built during periods of fiscal expansion may face very different economics when public budgets or external funding tighten.
- Access to products as well as services. Medicines, diagnostics, medical devices and assistive products are becoming a more explicit part of how WHO measures health-system capacity.
- Performance under stress. Coverage and utilization numbers show how a system works normally. Resilience shows what happens when staffing, demand, supply or infrastructure moves outside normal conditions.
The common thread is capacity.
Not capacity as one metric, but as the ability of the surrounding system to turn medical knowledge, people, products, data and funding into care—reliably and repeatedly.
Healthcare's next phase is less forgiving
The World Health Statistics 2026 report is not a story of global healthcare failure.
Many of its long-term trends are positive.
But positive direction is no longer enough.
Progress toward universal coverage is slowing. Public financing is under pressure. Workforce capacity remains dramatically uneven. The pandemic showed how quickly apparently strong systems can lose resilience. And large parts of the world still lack timely, usable information about something as fundamental as why people are dying.
These problems have different causes, but they increasingly collide inside the same operating system.
That is what makes the current period different.
The next decade of healthcare progress may depend less on discovering whether an intervention works and more on whether health systems can finance it, staff it, supply it, measure it and keep delivering it when conditions become difficult.
The treatment may already exist. The harder problem is keeping the system around it working.
