You don't need to be a data scientist to get value from the OMOP common data model. This is a place to get oriented — short videos, a starter path, the foundational reading, an example study protocol, and a shortcut into the OHDSI forums. It sits beside the toolkits on this site: the same standards (vocabularies, the CDM, phenotypes) underpin the registry work here.
Watch first
A few minutes each. Start with the gentle introduction, then the ATLAS demo, then the shorts when you want quick hits. From @DrDanielleTheDataDiva.
Full playlist on @DrDanielleTheDataDiva; more talks on OHDSI's channel.
Free resources
Canonical, free places to build a foundation — read, reference, take a course, and try the tools.
The community's free, comprehensive textbook — the single best starting point for the CDM, vocabularies, cohorts, and network studies.
The table-by-table specification for the current CDM version.
A plain-language, field-level reference — 258 fields across 24 tables. Search a term or filter to one table. The friendliest way to learn what the model actually holds.
Download and browse the OMOP standardized vocabularies — where clinical concepts live.
Free self-paced courses on the CDM, vocabularies, ETL, and the OHDSI tool stack.
Working groups, community calls, and the annual symposium.
The browser-based cohort and analysis tool — try it on the public demo, no install needed.
The full open-source stack for building and running observational studies.
Then try it
Once the concepts click, use the Analytic Use-Case Generator to shape a clinical idea into a well-formed OHDSI research question and check its feasibility for a network study.
Launch the Analytic Use-Case GeneratorGo deeper
A deeper, load-bearing set once you're ready, grouped by theme. You don't need all of these on day one — the Book of OHDSI plus the founding vision paper is plenty to start.
Vision & community
The founding statement of the collaborative and its goals.
Reliability & calibration — the methodological core
Introduces empirical calibration against negative controls.
Extends calibration from p-values to confidence intervals using positive and negative controls.
The OHDSI methods benchmark and reference-set thinking behind method evaluation.
Estimation, prediction & phenotypes
Flagship demonstration of high-throughput, systematically calibrated network estimation.
The PLP framework for building and evaluating prediction models on observational data.
Estimating phenotype sensitivity, specificity, and PPV without chart review.
Data quality & privacy
Conformance / Completeness / Plausibility × Verification / Validation — the DQ backbone.
The Kahn framework as executable checks — the write-up behind the DataQualityDashboard.
The Shift-and-Truncate method for obscuring dates while preserving within-patient intervals.
Search the whole corpus
Articles, videos, repositories, discussions, and documentation with relevance scoring.
Protocols & study packages
An example OHDSI study package to verify your site can run a network study using HADES — the cleanest way to see how a study package is structured before building your own.
Join the conversation
The forums are where the community works things out. These run a live search on forums.ohdsi.org.
Popular searches