World Bank Health Nutrition and Population Indicators

The health section of the World Bank's World Development Indicators catalog, resolved by Datadory into typed, join-ready rows. Of the 1,498 series in the WDI catalog, 240 carry the Health topic: service capacity (hospital beds, physicians, nurses and midwives, community health workers, specialist surgical workforce), financing (current expenditure per capita, government spend, out-of-pocket share), universal health coverage (service coverage index, financial-hardship measures), mortality (under-5, neonatal, maternal, causes of death), immunization, disease burden (HIV, TB, malaria), nutrition, and water and sanitation. Two companion databases extend it - Health Nutrition and Population Statistics with 475 indicators including the Human Capital Index, and a wealth-quintile family of 420 indicators splitting outcomes into population fifths.

Most series originate with the WHO and UN agencies and are harmonized by the Bank into one catalog, so an indicator code means the same thing everywhere - which is why this is the source behind the majority of cross-country health comparisons in policy papers, grant reports and journalism. As published it is a browsing experience; as delivered it is a schema.

One verified observation, laid out exactly as it arrives:

indicator      : SH.XPD.OOPC.CH.ZS
indicator_name : Out-of-pocket expenditure (% of current health expenditure)
country        : NG   country_name: Nigeria
iso3           : NGA
date           : 2021
value          : 76.24

Read the shape, not just the number. One row per indicator x economy x year, with the code, the readable name and the display precision travelling beside the value - so 76.24 stays attributable to Nigerian households in 2021, in percent of health spending, forever. Scale the anatomy up: roughly 1.9 million such populated rows across the health topic.

Start with a sample of this dataset cut to the indicators and economies you track.

Ten documented fields cover every observation, all ten verified during the August 2026 research pass. The locating four - indicator code, indicator name, country identifiers and year - pin each row to a concept, an economy and a moment. VALUE carries the measurement, typed numeric on arrival.

The interpretive set is what separates an analyzable panel from a spreadsheet-shaped liability: UNIT documents that WDI leaves the unit column blank on purpose (units ride inside the indicator label), OBS_STATUS flags special values, and DECIMAL records recommended display precision so 12.62 hospital beds per 1,000 never masquerade as 12.6231. Hand-collected extracts almost always drop these first, which is why so many cross-country comparisons disagree with each other.

217 economies recognized by the World Bank, plus its regional and income aggregates - up to 265 entities on a single series, so composite views like High income or Africa Eastern and Southern arrive in the same columns as countries. History varies by family: hospital beds, physicians and under-5 mortality run from 1960, immunization from 1980, maternal mortality modeling from 1985, and the modern expenditure and UHC families from 2000 through 2023 and 2024.

For panel work the useful property is comparability: a country-year row from 1975 joins cleanly to the same country-year row from 2024 because the indicator codes hold still. Gaps arrive as nulls rather than silence, and the reporting-lag pattern differs by indicator - capacity series trail mortality series by years, visibly.

Health economists build the cross-country panels that papers and preprints run on. Global health teams report immunization, mortality and WASH indicators for every economy they operate in without re-collecting any of them. Pharma and medtech market-access teams size territories from per-capita spending and capacity density. Actuaries lean on six decades of mortality trend, policy analysts benchmark against structural peers, and journalists resolve the famous charts early enough to check them.

See the persona pages for data scientists and journalists and academics working in health care services.

Because the published form spans three databases, 700-plus health-relevant indicators and a catalog of codes whose meanings hide in their labels - and the analysis lives in the joins between them. Datadory resolves the family into one normalized schema - indicator, economy, year, value, integrity fields - so assembling a multi-theme extract stops being a data-engineering project. Field definitions are verified rather than guessed from headers, uncertain corners fold under additional-fields-on-request, and everything ships delivered daily, weekly, or hourly over API, files, or your warehouse.

Start with a sample cut to the indicators, economies and years you actually need, then let the panel grow from there.