SUPP-- dataset declares no value-level metadata for its QNAM values

Rule

NS-007

Category

Define-XML

Basis

Confirmed Traceable to the published specification text cited below.

Detectability

Define-XML conclusive The Define-XML settles this one way or the other. No further evidence is needed to know whether it applies.

Finding type

Metadata gap

Applies to

SDTMIG 3.2, 3.3, 3.4

What changes in v4.0

A SUPP-- dataset's ten columns are the same in every study; what is specific to a study is its QNAM values, and those are declared in Define-XML value-level metadata. This Define-XML declares none for this dataset, so it says nothing about which non-standard variables the study carries, what they are called, how long they are, or where they came from. The v4.0 NS-- structure needs exactly that information: each non-standard variable becomes its own column, with a label, a data type, a length, an origin and — where applicable — a codelist. None of it can be recovered from this Define-XML, which does not make the conversion harder in principle but does mean the metadata for it has to be rebuilt from the source conversion code or the mapping specification.

What to do about it

Treat the missing value-level metadata as work in its own right, before the conversion rather than during it. Enumerate the QNAM values from the data or the mapping specification, decide the type, length, codelist and origin each one should have as an NS-- column, and add the value-level metadata to the Define-XML. Doing it first means the conversion has a specification to be checked against; doing it afterwards means the converted output becomes its own specification.

What the scan can and cannot see

The Define-XML declares no value list for this SUPP-- dataset, which is the finding. How many QNAM values the dataset actually contains is record-level and is not claimed — including the possibility that it contains none.