The rules of this glossary
- Definition first. An entry begins with what the term means, independently of any product. The corresponding feature, where one exists, appears only at the end.
- One term, one page. A concept treated in depth under topics has no separate entry here: it is cross-referenced instead. Two pages on the same term would compete without serving anyone.
- A real example. An abstract definition is hard to retain; every entry carries a concrete case.
- Standards cited and linked. Where a standard defines the term, it is named and its official address given.
Terms treated as topics rather than glossary entries
Several terms that would belong in a glossary are treated at length under topics instead, because a short definition would not do them justice. They are listed here so that nothing appears to be missing.
| Term | Where it is treated | Why there |
|---|---|---|
| Audit trail | Audit trail | The mechanism matters more than the definition: a modifiable log proves nothing |
| Electronic case report form | eCRF | Six distinguishing properties, none of which fits in a definition |
| Clinical registry | Clinical registry | The governance questions are inseparable from the concept |
| Clinical dataset | Clinical dataset | It is six deliverables, not one object |
| Data quality | Clinical data quality | Six dimensions, two of which cannot be recovered afterwards |
The rule behind this split is the one stated above: a term is treated in exactly one place. A short definition and a long article on the same concept would compete with each other, and a reader arriving on the shorter one would leave less well informed than if it did not exist.
Glossary
- Clinical data dictionaryDefinition of the clinical data dictionary: what it declares for each variable, why it precedes data capture, a concrete example, and the related standards.
- Data provenanceWhat data provenance means in clinical research: the four possible origins of a value, how it differs from the audit trail, and how it is verified.
- PseudonymisationWhat pseudonymisation is, how it differs from anonymisation, why pseudonymised data is still personal data, and how it applies in research.