RELREC Examples for Record-To-Record Relationships

This example shows how to use the RELREC dataset to relate records stored in separate domains for subject 123456 who had a clinical sign (rows 1 and 4) that was related to 2 palpable masses (rows 2 and 3) and two microscopic findings (rows 5 and 6).

This example shows the relationship between a grouped clinical observation and 2 laboratory results.

This example shows the relationship between a clinical observation for a cage (pool) and 2 laboratory results on individual subjects.

This example shows how data collection applications should use --RECID to enable downstream applications to create RELREC records that must persist between an initial creation of SEND datasets and subsequent SEND datasets.The variable --RECID is defined in the SDTM, which indicates that it may be added to SEND domains unless the SENDIG restricts it. So, --RECID may be used in most SEND domains, for example in situations where Clinical Observations (CL), Microscopic Findings (MI), Macroscopic Findings (MA), and Palpable Masses (PM) domain records need to be correlated and the following situations exist:

• A facility uses one vendor to supply software for their in-life data collection and another vendor to supply software for necropsy and histopathology data collection.
• A part of a study (e.g., histopathology) is conducted by a different organization than conducted the in-life and necropsy portions.

In both situations, the correlation between the findings in the different domains must be collected; --RELREC is not supposed to describe correlations made later in the report.

In these situations, software that periodically receives SEND transmissions (e.g., following interim necropsies) cannot rely on the --SEQ variable values on the related records to remain consistent because the standard does not require this. With this use of --RECID, the receiving system is enabled to identify changes between the current transmission and previous ones (if any) and to determine if and how previously established record relationships need to be adjusted.

In this example a gross brain finding of focus/foci, red, is correlated to the microscopic finding of thrombus.

MA Example