OHDSI WebAPI Extractor

Point at any OHDSI WebAPI — the open-standard REST backend behind ATLAS, the observational-health / real-world-evidence platform built on the OMOP Common Data Model — and pull its server info, data sources, saved cohort definitions, concept sets, and OMOP vocabulary concept search into clean, flat, tabular rows.

One generic runner for every OHDSI network host — no per-site scraper. Give a base URL and a mode; every conformant WebAPI answers the same uniform REST grammar, so the same normalized row shape comes back regardless of which OMOP network operates it.

Why this actor

The OHDSI (Observational Health Data Sciences and Informatics) community runs a global federated network of OMOP-CDM databases queried through a common WebAPI. Researchers, pharma real-world-evidence teams, and health-data engineers routinely need to inventory what a given ATLAS/WebAPI instance exposes — which data sources are configured, what saved phenotype cohorts and concept sets exist, and how OMOP standard concepts map for a search term. This actor does exactly that against any WebAPI host, in one call.

Modes

Mode	What you get	Endpoint
sources	One row per configured OMOP data source, with its daimons (CDM / Vocabulary / Results / …) flattened + booleans for the common ones	/source/sources
info	The server's version, build & configuration (security enabled, Solr enabled, …)	/info
cohortdefinitions	One row per saved cohort/phenotype definition (id, name, description, timestamps, tags)	/cohortdefinition
conceptsets	One row per saved concept set (curated OMOP concept collections)	/conceptset
vocabulary-search	One row per matching OMOP concept for a search term (id, name, code, domain, vocabulary, standard flag, validity dates)	/vocabulary/{sourceKey}/search

Input

Field	Required	Description
baseUrl	✅	One or more WebAPI root URLs (e.g. https://atlas-demo.ohdsi.org/WebAPI). The root path varies per host — some mount at /WebAPI, some behind a proxy at /atlas/WebAPI — so give the full base; the actor appends the endpoints.
mode		One of the five modes above. Default sources.
query		Search term for vocabulary-search (e.g. aspirin, type 2 diabetes).
sourceKey		For vocabulary-search, the source whose vocabulary to search. Auto-discovered from /source/sources (first source with a Vocabulary daimon) if omitted.
maxRecords		Total row cap across all hosts. Default 1000.
bearerToken		Optional OAuth2 bearer for a security-enabled host. Not needed for public reference servers. Never logged.
extraHeaders		Optional extra request headers (JSON object).

Example

{
  "baseUrl": ["https://atlas-demo.ohdsi.org/WebAPI"],
  "mode": "vocabulary-search",
  "query": "aspirin",
  "maxRecords": 500
}

You can pass several hosts in baseUrl to build a cross-network census in one run — a host that is unreachable is skipped with a warning and the run continues; a run where no host produced any row fails fast so a broken configuration never looks successful.

Output

Each result row is flat and tabular, with _source / _mode / _webApiVersion provenance columns plus mode-specific fields, and a _raw column preserving the full original object (capped for size). The dataset ships an overview table view so results render cleanly in the Apify console and export to CSV/JSON/Excel.

Pricing

Pay per result — you are charged one event per emitted row. No subscription; you pay only for what you extract.

Notes

• Works against public OHDSI reference servers out of the box; no credentials required.
• Vocabulary search can return tens of thousands of concepts — use maxRecords to bound a run.
• Timestamps from the WebAPI (epoch-millis) are normalized to ISO-8601 UTC strings.
• OHDSI, OMOP, ATLAS and WebAPI are projects/trademarks of the OHDSI community; this actor is an independent client of the public open-standard API and is not affiliated with OHDSI.