The DataSubset class is used in the reporting event's dataSubsets attribute to specify any additional data subsets that are used to restrict the subjects or records that are included in any given analysis, but are not considered to be defined analysis sets (e.g., subjects with a non-missing baseline value, or inclusion of only treatment-emergent adverse events in an AE summary analysis). Once a data subset has been defined in the reporting event's dataSubsets attribute, it can be used for any analysis by specifying its identifier value in the dataSubsetId attribute of the instance of the Analysis class that represents the analysis, as described in the Analysis section.

Each data subset is defined as an instance of the DataSubset class, in which:

• The id attribute contains the assigned identifier value for the data subset.
• The name attribute contains the assigned name for the data subset (e.g., "Treatment-Emergent Adverse Events").
• The description attribute may be used to record a detailed description of the data subject.
• The label attribute may be used to record a short description of the data subset (e.g., "TEAE").
• The level and order attributes are both assigned a value of "1".
• Either:
  • The condition attribute contains the definition of a simple data selection criterion represented as an instance of the WhereClauseCondition class, as described in the WhereClauseCondition section.

Or:

• • The compoundExpression attribute contains the definition of a compound expression represented as an instance of the CompoundSubsetExpression class, in which:
    • The logicalOperator attribute contains a value from the ExpressionLogicalOperatorEnum enumeration.
    • The whereClauses attribute contains 1 or more subclauses being combined or negated by the specified logical operator. Each subclause is represented as one of the following:
      • A where clause represented as an instance of the WhereClause class, with the subclause expressed using either the condition or compoundExpression attribute.
      • A reference to another defined data subset represented as an instance of the ReferencedDataSubset class, in which the subClauseId attribute contains the identifier value of the referenced data subset.

Refer to the WhereClauseCompoundExpression section for more information about compound expressions, including examples of data subsets defined as a compound expressions.