Summary
A data catalog describes what a database contains — which concepts, over which periods, in which units, for how many patients — without ever publishing an individual record. You pick the variables and the crossings to count; Linkr computes, anonymizes, then produces a standalone HTML page with Health-DCAT-AP metadata, to download or to deploy on GitLab or GitHub Pages.
Two 'catalogs' not to confuse
The data catalog, described here, answers: “what is in this warehouse?” It describes your data so other teams can tell whether it is worth asking for.
The community catalog answers an entirely different question: “which schemas, plugins or projects can I install?” It indexes published Linkr entities, not data.
The problem: announcing without exposing
A hospital warehouse is invisible from the outside. A team looking for a cohort of septic patients cannot tell whether you have any, how many, or over which years — short of asking, waiting, and often finding out the answer was no.
Publishing the data is out of the question. But publishing its description — which concepts are present, how many records, how they spread over time and across units — does not carry the same risk, provided the counts are too coarse to re-identify anyone.
That is what a catalog produces: a counted, anonymized, publishable inventory. It gives orders of magnitude; for exact counts on a precise question, you run a feasibility study, with cohorts.
In this guide
A catalog is created from the warehouse’s Data catalog page, with New catalog, then set up tab by tab: you configure and compute, anonymize, describe, then publish. The first page follows that path from end to end; anonymization, set between computing and publishing, has its own page because it deserves a closer look:
Building and publishing a catalog
The database described, the variables and crossings to count, the computation, the Health-DCAT-AP metadata, then the export or the deployment on GitLab / GitHub Pages.
Anonymizing a catalog
The threshold, secondary suppression, count perturbation and the disclosure audit — to set after computing, before publishing.
The published page
The page has four tabs:
- Explore — pick one, two or three variables in the sidebar and filter each of them: the charts, key figures and table follow.
- Metadata — the dataset’s Health-DCAT-AP description, and the JSON-LD.
- Schema — the structure of the source warehouse, as tables and as a diagram.
- Info — how the counts were made and protected.
Intensive care warehouse
Concept catalog
Adult intensive care stays since 2021: patients, stays, laboratory, vital signs and drugs.
| Period ╲ Gender | Male | Female | Unknown |
|---|---|---|---|
| 2021 | 2,310 | 2,015 | < 10 |
| 2022 | 2,402 | 2,120 | 14 |
| 2023 | 2,451 | 2,188 | masked |
| 2024 | 2,498 | 2,231 | 11 |
Under filters, the key figures add up the cells concerned where they add up. When that is only an approximation — a patient present in two periods counts in each, masked cells are left out, counts are perturbed — the figure is preceded by ≤, ≥ or ≈, and the information icon next to it says why.
What never leaves
A catalog holds only aggregate counts. No patient row, no identifier, no individual event date — and, for a masked cell, not even a count, in any file.
It is the same logic as for databases: what leaves the institution describes the data without carrying it away.
Going further
- Building and publishing a catalog — from configuration to the published page.
- Anonymizing a catalog — threshold, secondary suppression, perturbation and audit, before any publication.
- Schemas — required for the computation to know what to count.
- Cohorts — for exact counts on a precise question.
- Community catalog — the other catalog, the one of published entities.