Build a cohort revision, load its deidentified facts and accept or reject each one.
Outcomes Analyst confirms the source snapshot covering the cohort window is Fresh and its record count matches the extract. A Stale or Failed snapshot goes back to the Data Engineer.
Outcomes Analyst creates the cohort with its number, name, cohort date, organization, owner, outcome definition and source record ID. It starts at Configured.
Outcomes Analyst starts the construction workflow from the Healthcare Analytics Overview page or by hand. It deduplicates by operation key, records an operation run with numbered steps, writes the state history entry Status: Configured → Collecting and opens a pending review decision when approval is required.
Deidentified outcome facts are loaded against the cohort with their source reference, observation date, counts and consent basis. Each starts at Outcome = 'Pending'.
Outcomes Analyst moves each fact Outcome: Pending → Accepted or Pending → Rejected and attaches supporting evidence documents.
When every fact has an outcome and the counts reconcile with the snapshot, the cohort moves Status: Collecting → Validated and the change is written to state history.
And how to keep it from happening.
Start free, then use Proto to change fields, steps and approvals before your team goes live.