Psychiatric & Pharmacovigilance
Some psychiatric drugs come with a mandatory safety net attached — clozapine, the most effective antipsychotic for treatment-resistant schizophrenia, can also cause severe neutropenia, so its use is legally tied to an FDA Risk Evaluation and Mitigation Strategy: regular blood counts, on a fixed schedule, for as long as the patient is on the drug. That monitoring window is a real mathematical object, not just an administrative rule, and the dataset in this category models it directly rather than treating it as a footnote.
It ships in OMOP Common Data Model format, so it opens in the same tooling a hospital's own pharmacovigilance pipeline would use, and it is built around a synthetic cohort large enough — tens of thousands of patients — to give a stable picture of how often the monitoring actually catches something in time.
What this category is for
- REMS-anchored datasets — built around a real regulatory monitoring requirement, with the schedule and thresholds taken from the actual programme, not invented.
- OMOP-formatted psychiatric cohorts — ready for pharmacovigilance tooling without a custom ingestion pipeline.
- Toxicity-prediction labels — a lead-time model for when a monitoring signal would actually have preceded an adverse event, not just whether one occurred.
As with everything else in the catalog: no real patient data, and every modelling choice is disclosed rather than left for you to discover later.