Session 5: From Cohorts to End-to-End OHDSI Pipelines
Then: guided module practice.
A typical OHDSI study combines multiple packages:
Running each package manually is hard to standardize across sites.
Strategus orchestrates HADES modules as one reproducible pipeline.
Today is foundation, not method depth.
Next sessions dive into characterization/incidence and CohortMethod details.
Strategus separates:
This separation is what makes network execution practical.
Design study spec (JSON)
↓
Site-specific execution settings
↓
Strategus execute()
↓
Module outputs + upload-ready results
A module is a standardized analytic task.
Examples:
CohortGeneratorModuleCohortDiagnosticsModuleCharacterizationModuleCohortMethodModuleSome inputs are reused across modules:
These go in sharedResources in the analysis specification.
Before any module settings, get cohort assets cleanly organized:
Cohorts.csv, JSON, and SQL folders structuredThis helps keep your study comprehensible to others and easy to modify/work with.
Each module contributes its own settings block.
Examples:
The analysis specification JSON captures:
No credentials or site-specific database details.
Execution settings are site-local and typically include:
cdmDatabaseSchemaworkDatabaseSchemaStrategus instantiates modules and runs them in sequence.
This supports transparent, auditable network studies.
JSON-based settings make it easier to:
We will walk through one concrete script:
modules/05_strategus-intro/strategus-practice.R
It covers:
Semaglutide NAION study repository:
https://github.com/ohdsi-studies/SemaglutideNaion
Use it as a realistic reference for study structure and settings.
renvWe’ve moved from:
to orchestrating full study pipelines.
Hands-on:
modules/05_strategus-intro/strategus-intro.qmd