For a given parameter within a BDS dataset, it is important (1) to be able to distinguish analysis values that are special cases of AVAL/AVALC (those for which the value is determined differently than the other analysis values within the parameter), and (2) to understand what method or algorithm was used to populate each special case. The variable DTYPE is to be used to identify records within a given parameter that contain these special-case analysis values. The value of DTYPE indicates the method used for populating the analysis value; a null value of DTYPE indicates the analysis value was not a special case. The metadata for AVAL (or AVALC) will give further information about the details of any algorithm or statistical method used to derive or impute these values. When the analysis value on a record within a parameter has been imputed or modified, DTYPE will indicate the method used to populate the analysis value. DTYPE would be used if there are special cases within the new parameter that should be identified. If a parameter is wholly derived (e.g., a time-to-event parameter), then it is a misapplication to populate DTYPE for all records in that parameter because, by definition, all records are derived using the same method.
If analysis timepoints are defined by relative day or hour windows, then the variables in the following table may be used along with ADY or ARELTM to clarify how the record representing each analysis timepoint was chosen from among the possible candidates. The record chosen is indicated by the analyzed record flag ANLzzFL (see Section 2.9.6.10, Indicator Variables for BDS Datasets). Note that the variables in this table may not be applicable in all situations and are presented as an option.
Overview
Content Tools