Pharmaceuticals Data Provider: 22 Cataloged Datasets · Head-to-head
EMA Medicines Catalogue vs EMA Download Medicine Data Tables
Which pharmaceuticals data provider: 22 cataloged datasets data fits your job: EMA Medicines Catalogue, or EMA Download Medicine Data Tables. API, files, or your warehouse. Daily, weekly, or hourly.
EMA Medicines Catalogue
EMA Download Medicine Data Tables
Where the fields line up
15 shared fields — join on these.
| Field | EMA Medicines Catalogue | EMA Download Medicine Data Tables |
|---|---|---|
category | documented | documented |
name_of_medicine | documented | documented |
ema_product_number | documented | documented |
medicine_status | documented | documented |
international_non_proprietary_name_common_name | documented | documented |
active_substance | documented | documented |
atc_code_human | documented | documented |
pharmacotherapeutic_group_human | documented | documented |
orphan_medicine | documented | documented |
conditional_approval | documented | documented |
additional_monitoring | documented | documented |
marketing_authorisation_developer_applicant_holder | documented | documented |
european_commission_decision_date | documented | documented |
marketing_authorisation_date | documented | documented |
last_updated_date | documented | documented |
Coverage, side by side
| EMA Medicines Catalogue | EMA Download Medicine Data Tables | |
|---|---|---|
| Geographic | European Union / European Economic Area - centrally authorised medicines only | European Union / EEA centrally authorised medicines, plus opinions for use outside the EU |
| Temporal | Current register snapshot; individual medicine pages retain history since the mid-1990s centralised procedure era | Current regulatory state per record with full procedural date history; table generation log begins April 2026 |
| Granularity | One row per authorised medicine (product level), with procedure-level tables for post-authorisation actions | One row per medicine product (EMA product number), with separate tables per procedure type: referrals, PIPs, orphan designations, PSUSAs, DHPCs, shortages, MRLs |
What each contains
They tie on 1 attribute. Pick by fit, not by loyalty.
| EMA Medicines Catalogue | EMA Download Medicine Data Tables | |
|---|---|---|
| Publisher | European Medicines Agency (EMA) | European Medicines Agency (EMA) |
| Subject lens | Register of centrally authorised human and veterinary medicines plus herbal products, searchable by medicine name or active substance, with advanced search across refused and withdrawn applications | Flat-table estate of the same regulatory data: a medicines core plus ten procedure-level tables from post-authorisation actions to supply shortages and veterinary maximum residue limits |
| Documented fields | 34 fields, definitions verified during research | 31 fields on the medicines table alone (39 populated columns in the source file), definitions verified during research |
| Geographic coverage | European Union / European Economic Area - centrally authorised medicines only | European Union / EEA centrally authorised medicines, plus opinions for use outside the EU |
| Temporal coverage | Current register snapshot; individual medicine pages retain history since the mid-1990s centralised procedure era | Current regulatory state per record with full procedural date history; table generation log begins April 2026 |
| Granularity | One row per authorised medicine (product level), with procedure-level tables for post-authorisation actions | One row per medicine product (EMA product number), with separate tables per procedure type: referrals, PIPs, orphan designations, PSUSAs, DHPCs, shortages, MRLs |
| Scale | ~2,700 authorised medicines per export: 2,337 human + 393 veterinary in the August 2026 file | Medicines table: 2,730 records across 39 populated columns, plus ten additional procedure tables |
| Secondary industries tagged | Biotechnology | None |
| Best for | Register-grade annotation per medicine: nine regulatory flags, MeSH and ATC classification, veterinary species, and the canonical medicine page link | Recurring pulls of the whole medicines panel plus the adjacent procedure tables - referrals, shortages, orphan designations, MRLs - in one consistent estate |
Where they're equivalent
More than their different shapes suggest. Same publisher throughout: the European Medicines Agency, with the register compiled through the EU centralised procedure. Same subject population: centrally authorised human and veterinary medicines in the EU and EEA - about 2,700 products, 2,337 human plus 393 veterinary in the August 2026 exports - with nationally authorised products outside both scopes. Same grain: one row per medicine, keyed on the EMA product number, which joins the two sides without translation. Same temporal posture: both hold the current register rather than an archive, each row carrying its full procedural date history, and both lean on the underlying medicine pages for history reaching into the mid-1990s. Same classification spine, same nine regulatory flags, same holder field, same document pointer. On paper this pair disagrees about almost nothing - it disagrees about how much hangs off the register.
The verdict
Verdict: sample both, pick by fit - one register, one estate, and your question decides.
Take EMA Medicines Catalogue if the question lives inside a single medicine. Competitive profiling of an authorised product, ATC-based market maps, orphan or biosimilar landscape scans, veterinary portfolio views with target species, or any deliverable where the refusal date and the label-revision procedure must be stated precisely - the 34-field register answers these directly, and it is the highest-scoring pharma record in the catalog at 10 of 10.
Sample both, pick by fit. See EMA Medicines Catalogue · See EMA Download Medicine Data Tables
Or take both in one feed
They stack naturally, because they are two readings of one register joined on a stable key. A defensible workflow: land the Catalogue as the annotation layer - 34 fields per product, flags and classification included - then attach the tables' procedure-level files on ema_product_number, so each medicine carries its referral events, PIPs, DHPCs and shortage entries as child rows. Reconcile counts monthly against the register total of roughly 2,700 authorised products; when the sum moves, new authorisations, withdrawals or suspensions have landed and both layers are due for a re-pull.
Three seams decide whether the merge holds. First, the end-of-life asymmetry: the tables fold withdrawal, expiry, revocation and lapse into one dated field while the Catalogue keeps refusals separate, so normalise lifecycle logic before joining statuses. Second, veterinary attributes: target species exists only on the Catalogue side, so animal-health cuts must pull from there. Third, scope: neither side knows about nationally authorised medicines, so any EU-wide claim needs Member State sources beyond this pair. Or take both in one feed. Datadory normalizes each record to its documented dictionary, attaches sample rows for validation, reconciles the two on the product number, and ships them beside the rest of the thirteen-record Pharmaceuticals slice - delivered daily, weekly, or hourly, your call.
API, files, or your warehouse. Daily, weekly, or hourly.
Fair questions
Do both datasets cover the same EMA medicines?
The medicines layer is the same corpus. Both describe centrally authorised EU/EEA medicines at roughly 2,700 rows per export - 2,337 human plus 393 veterinary in the August 2026 files - keyed on the EMA product number, such as EMEA/H/C/005243, which matches across both sides. Neither carries nationally authorised products; those live in Member State registers.
Which fields appear in both datasets?
Thirty of the Catalogue's 34 documented fields reappear in the tables dictionary: category, medicine name, EMA product number, statuses, INN and active substance, MeSH therapeutic area, human ATC and ATCvet codes, pharmacotherapeutic group, therapeutic indication text, all nine regulatory Yes/No flags, the holder company, the EC decision and authorisation dates, revision number and the medicine page URL. The Catalogue adds four fields the tables record does not document; the tables add one consolidated withdrawal-date field.
What can I get from the tables that the Catalogue does not include?
Ten procedure-level tables beyond the medicines core: post-authorisation procedures, referrals, paediatric investigation plans, orphan designations, PSUSA assessments, DHPC safety communications, medicine supply shortages, herbal medicines, opinions for use outside the EU, and maximum residue limits for veterinary products. The Catalogue documents the register itself; only the tables estate carries these adjacent regulatory workloads.
How current is the data in each dataset?
Both describe the current register rather than an archive: each export holds today's authorisation state alongside full procedural date history, and underlying medicine pages retain history into the mid-1990s centralised-procedure era. The tables estate's generation log begins April 2026, so observed drift between the two sides has been measurable only since then. Delivery cadence is your call: daily, weekly, or hourly.
Can Datadory deliver both EMA datasets together?
Yes - sample both and pick by fit, or take both in one feed. Datadory normalizes each to its documented field dictionary, attaches sample rows for validation, reconciles the two on the shared EMA product number, and ships them beside the rest of the thirteen-record Pharmaceuticals slice, delivered daily, weekly, or hourly.