Senior Director, Standards Development & Education
CDISC United States
Using both CDASH and SDTM together is critical for producing high quality data, and understanding their similarities and differences is critical to using them correctly.
ORAL PRESENTATION: 1:00PM
There is some confusion among CDISC standards implementers as to the difference between the CDASH (data capture) and SDTM (data tabulation) standards, the roles each plays, how they work together, and why both are necessary to achieve high quality clinical research. This results in some organizations choosing to bypass CDASH and map their collected data directly to SDTM. This poster will illustrate the relationship between CDASH and SDTM, what each brings to the process, and SDTM relies strongly on CDASH to produce reliable, poolable and analyzable data.
Do you feel that CDISC’s CDASH standard is an unnecessary tag-along to SDTM? If so, you are not alone. There is confusion among CDISC standards implementers about the difference between the CDASH (data capture) and SDTM (data tabulation) standards. This results in some organizations choosing to bypass CDASH and map their collected data directly to SDTM. While very similar to SDTM, CDASH fulfills a different role, and this has implications for data capture, quality, usability, repurposing, and traceability, among other considerations. For example,
• They define different metadata: CDASH specifies the wording for the question on the CRF, whereas SDTM provides the variable label for display.
• CDASH ensures that, if no data were received, that no data were generated. In contrast, SDTM assumes that if no data are present, then no data were generated.
• CDASH allows variables from different domains to appear on the same CRF, and variables from the same domain to be presented on different CRFs. SDTM requires that all data for a given domain must appear in that domain.
This poster will define the similarities and differences between these two standards, and explain why they are both essential for high quality clinical research.