WHO Medical Devices – Definitions and Policy Repository

The dictionary the global device industry argues by. The World Health Organization's Health Product Policy and Standards team maintains this repository as the reference entry point for what a medical device legally is, which devices countries prioritize, and how national device policy actually stands today. It publishes WHO's working definitions — health technology, medical devices (adopted from the Global Harmonization Task Force definition finalized in May 2012), and medical equipment: devices requiring calibration, maintenance, repair, user training and decommissioning, explicitly excluding implants and single-use devices. Companion pages extend the vocabulary to priority medical devices (including the COVID-19 priority list), in vitro diagnostics, personal protective equipment, oxygen therapy and local production.

Two layers sit under the definitions. The indicator layer runs through twelve DEVICES-series codes verified live during cataloging: eight categorical policy flags — national standards or recommended lists (DEVICES06), approved procurement or reimbursement lists (DEVICES18), procurement guidelines (DEVICES19), technical specifications availability (DEVICES20), a Ministry of Health device unit (DEVICES16), national-level procurement (DEVICES17), a regulating authority (DEVICES24) and an HTA unit (DEVICES25) — plus equipment density per million population for MRI (DEVICES08), CT (DEVICES09), PET, mammography, linear accelerators, telecobalt, radiotherapy and gamma camera/nuclear medicine units. Verified pulls returned roughly 290–540 country-year rows per indicator, about 4,700 rows in total. The publication layer is the Global Atlas of Medical Devices: seven survey topics — health technology policy, device regulation, HTA units, health technology management, nomenclature adoption, national priority device lists and high-cost equipment — folded into editions from a 2010 baseline through the 2017 and 2022 atlases, the newest carrying country profiles for 177 countries.

Within Datadory's catalog of 1,744 datasets across 159 viable industries, this record scores 7/10 for quality and is the only non-US source among Health Care Equipment's three primaries — the global-policy counterweight to American regulator feeds that stop at the border. Get a sample of this dataset cut to your indicators and countries before anything else.

Three real observations exactly as they ship — one density row, one OECD-sourced row, one policy row:

IndicatorCode : DEVICES08     # MRI units per million population
SpatialDim    : TTO           # Trinidad and Tobago (ISO-3)
ParentLocation: Americas
TimeDim       : 2010
Value         : 2.98          NumericValue: 2.98
Comments      : Only public sector data

IndicatorCode : DEVICES08
SpatialDim    : USA           ParentLocation: Americas
TimeDim       : 2019
Value         : 40.11         NumericValue: 40.11
Comments      : OECD 2017-19

IndicatorCode : DEVICES06     # national standards / recommended lists
SpatialDim    : VUT           # Vanuatu
TimeDim       : 2010
Value         : For different healthcare facilities and specific procedures

Read them side by side and both halves of the corpus are visible. The Trinidad and USA rows are the quantitative layer: same indicator code, same typed numeric value, twenty-nine MRI units per million apart, with the Comments column carrying the provenance split — one survey wave, one OECD compilation. The Vanuatu row is the policy layer: identical schema, but the value is descriptive text describing which facilities the national standards list covers. One table shape carries both, so a density benchmark and a regulatory-status screen pull from the same delivery instead of two reconciled sources. Name your indicators and countries when you request the sample and it arrives cut to exactly that shape.

Nine documented fields carry every row, each verified down to type and example value during the research pass — nothing inferred from column names alone. Five do the identifying: the indicator pair (IndicatorCode, IndicatorName), the geographic pair (SpatialDim, ParentLocation) and the year. Three carry the measurement itself, and the last two are bookkeeping: a scope footnote and the warehouse load timestamp.

The design consequence worth planning around: Value is text and NumericValue is numeric where applicable. Density indicators populate both; policy indicators populate only the text form, because 'For different healthcare facilities and specific procedures' has no float representation. Filter on NumericValue IS NOT NULL and you have the quantitative corpus; keep the text rows and you have the policy census.

Four extensions sit outside the verified nine under additional fields on request: a value-mapping table for the eight categorical policy indicators (their descriptive strings repeat across countries but were never published as a coded enumeration), any age or sex disaggregations attached to specific indicators, the Atlas country-profile tables behind the 177-country 2022 edition, and per-indicator notes distinguishing nationally reported observations from OECD-derived ones beyond the Comments footnote. They are named at sampling, confirmed with your extract, rather than guessed here.

• Geography: up to roughly 143 WHO Member States per density indicator, tagged with ISO-3 codes and rolled up through WHO regions (Africa, Americas, Europe, Eastern Mediterranean, Western Pacific, South-East Asia); the 2022 Global Atlas adds structured country profiles for 177 countries.
• Temporal: Atlas survey waves span 2009–2022 across three editions (a 2010 baseline survey plus the 2017 and 2022 atlases). The indicator layer holds observations from 2010 to 2021, some sourced from OECD 2017–19 compilations, with no DEVICES-series code refreshed since the 2010–2013 wave except those OECD values.
• Granularity: one country-year observation per indicator; policy indicators are categorical statements about a country's standing arrangements rather than measured quantities.

That footprint is the niche. Set against the wider catalog — average quality score 7.81 across all 1,744 datasets — this slice scores 7/10, held up by field documentation that maps fully and pulled down by freshness: this is the record you cite for how device policy stands rather than the one you watch for what changed last month. For current-state imaging capacity, imaging equipment density is the term of art these indicators define.

API, files, or your warehouse. Daily, weekly, or hourly.

Name the indicators, the countries and the year range when you request the sample — all twelve DEVICES-series codes across their full Member State span, or a narrow cut like MRI and CT density for a competitor cohort plus the procurement-list flags beside them. The sample ships first either way; the ongoing feed lands shaped identically, so joins prototyped on the sample survive into production unchanged.

A global policy-and-capacity ledger earns its keep in five specific jobs:

• Market sizing and market-entry screening — MRI, CT and radiotherapy density per million population is the installed-base denominator medtech strategy teams size against; compare forty candidate markets on equipment intensity before choosing where a launch lands.
• Regulatory-affairs mapping — the eight categorical flags say which countries run a regulating authority, publish procurement guidelines or maintain approved reimbursement lists, turning 'what does this market require' from a fifty-tab research project into a filtered query.
• Health-systems and NGO planning — density benchmarks against WHO-region peers give procurement programs a defensible target: how many imaging units per million a system should plan for, not just how many it has.
• Citation-grade research and journalism — WHO's definitions are the ones other publishers quote. When a device-policy claim has to survive review, the organization that standardized the vocabulary is the provenance reviewers stop arguing about.
• Investment screens — equipment density trends paired with policy-readiness flags separate markets where demand exists but procurement machinery does not from markets ready to absorb capital.

Personas: journalists academics use cases rate this record highest in the industry at 3/3 relevance — citable definitions and country profiles for device-policy scholarship. Data scientists and developers rate it 2/3: a compact, fully-documented country-year panel for global-health dashboards and models, best joined against the event-level FDA feeds rather than substituted for them.

Three things worth knowing upfront.

First, freshness has a shape: no DEVICES-series indicator has been updated since the 2010–2013 survey wave except OECD-sourced values, and each row's load timestamp says exactly when it entered the warehouse. Treat this as the structural baseline it is — benchmark structure, regulatory posture, relative capacity — and pin another feed for anything that must move weekly.

Second, the policy half is words, not numbers: DEVICES06/16/17/18/19/20/24/25 return descriptive text such as 'For different healthcare facilities and specific procedures'. The repeating phrases map cleanly once codified, which is why the value-mapping table waits under additional fields on request rather than shipping as a guess.

Third, density is a ratio, not a count: values state units per million population, so a small country can post a striking figure off a handful of machines. Join population or absolute unit counts alongside before ranking markets on raw density alone. None of these bite unexpectedly — they are flagged in your sample against the analysis you plan to run.

Provenance note — compiled by the World Health Organization's Health Product Policy and Standards team, the same program behind the Global Harmonization Task Force device definition finalized in May 2012. One team owns the vocabulary, the priority lists and the Atlas surveys, which is why the definitions and the indicators never disagree with each other.

Methodology note — the seven Atlas survey topics (policy, regulation, HTA units, technology management, nomenclature adoption, priority lists, high-cost equipment) are self-reported by Member States and reconciled by WHO; density indicators compile from national sources including OECD compilations. Read policy rows as statements of standing arrangements, density rows as measured ratios.

Completeness note — Datadory scores this record 7/10 with all nine fields documented to type and example. What sits outside the verified core is listed above under additional fields on request rather than approximated here.

Where to go next — inside Health Care Equipment, the openFDA Device Adverse Events (MAUDE) API supplies the event-level US record this repository's policy layer contextualizes, and the openFDA Device Registration & Listing API maps who manufactures what. Across industries, the WHO Global Health Observatory Data Repository widens the aperture to 3,000+ health indicators over the same Member State geography. Start from the best health-care-equipment datasets ranking or the health care equipment data guide to see where this record lands in the slice.