How to Scrape WHO Global Health Data

The WHO Global Health Observatory contains a large catalog of country indicators, but its OData structure exposes technical dimensions that can be easy to drop during a quick export. The WHO Global Health Observatory Data Scraper turns selected indicator codes into rows while preserving country, region, year, sex, value, bounds, and update metadata.

Query explicit indicator codes

Before scaling a query, inspect several rows for unexpected dimensions, formatted values, and null bounds. WHO indicators do not all share one practical grain. A small exploratory run helps define the right key and prevents a broad export from mixing populations that only look comparable.

{"indicatorCodes":["WHOSIS_000001"],"countryCodes":["USA","IND"],"sexCodes":["SEX_BTSX"],"yearFrom":2000,"yearTo":2026,"maxResultsPerIndicator":100}

The example collects life expectancy at birth for both sexes. Every row carries the numeric value and WHO's formatted display value, which may include uncertainty bounds. Keep low and high beside estimates instead of presenting only the center value. Country and region names are resolved from WHO dimensions, while the official code remains available for stable joins.

Indicator coverage differs by country and year. A missing observation does not mean zero disease, zero mortality, or zero service coverage. Validate indicator definitions and methodology before comparisons, especially when measures combine modeled and reported data.

For recurring research, upsert on indicator code, country code, year, and dimension values, then retain the WHO update date and your collection timestamp. Separate raw observations from derived rankings or trends. That structure lets analysts refresh official data without silently rewriting earlier conclusions and keeps every dashboard number traceable to its source record.