SUPPQUAL – Where’s My Mommy? Sandra VanPelt Nguyen Midwest CDISC Users Group May 2012 Abstract When using CDISC standards, there may be situations in which a field has been collected on a CRF or included in a vendor data transfer which seems to be clinically relevant but is not “topic” data belonging in a standard or custom SDTM domain (any of the three general observation classes). Since non-standard variables cannot be added to SDTM domains, this data typically gets mapped to a SUPPQUAL (Supplemental Qualifiers) dataset, but what do we do when there is not an obvious “parent” record corresponding to this data within one of the SDTM domains? This presentation will provide a set of guidelines to use to determine how best to handle these situations. Discussion Topics ► Introduction/Background ► Guidelines ► Examples © 2011. All rights reserved. 3 Intro to Supplemental Qualifiers (SUPPQUAL) ► CDISC SDTM organizes data in domains (datasets) based on topic ► Topics are divided amongst 3 general domain classes - Interventions, Events, Findings ► Topics which do not fall into a reserved SDTM domain but fit into a general domain class may go into custom domains ► Variables cannot be added to domains with exception of certain timing, identifier and qualifier variables allowable within the general domain class ► SDTM allows for the inclusion of additional variables using the Supplemental Qualifiers special-purpose domain Intro to Supplemental Qualifiers (SUPPQUAL) ► SUPPQUAL is used to capture non-standard variables and their association to parent records in general observation class datasets or DM (demographics) ► QNAM (qualifier variable name) and QLABEL (qualifier variable label) are used to describe the particular qualifier ► Variables IDVAR (identifying variable) and IDVARVAL (identifying variable value) provide the linkage between records (qualifier child to observation(s) parent) by identifying the variable to join on and the variable value of the specific record(s) which are related ► SUPPQUAL is typically split as SUPP--, with the suffix based on the related domain’s code, e.g. SUPPAE for qualifiers related to the AE domain SUPPQUAL Example CMTRT CMTRT Reported Name of Drug, Med, or Therapy CMCAT Category for Medication CMINDC Indication CMDOSE Dose per Administration CMDOSTXT Dose Description CMDOSU Dose Units CMSTDTC Start Date/Time of Medication CMENDTC End Date/Time of Medication CMLOC Location of Dose Administration CMSTDTC CMENDTC CMLOC CMDOSE/CMDOSTXT CMDOSU © 2011. All rights reserved. 6 SUPPQUAL Example: Parent-Child Relationship CM: USUBJID CMSEQ CMTRT CMDOSE CMDOSU CMLOC CMSTDTC CMENDTC ABC-001-0001 1 PALLIATIVE RADIATION 68 Gy PROSTATE 2010-01-20 2010-03-01 SUPPCM: USUBJID RDOMAIN IDVAR IDVARVAL QNAM QLABEL QVAL ABC-001-0001 CM CMSEQ 1 REASDISC Reason for Discontinuation Completed therapy SUPPQUAL Example: No Parent CM: USUBJID CMSEQ CMTRT CMDOSE CMDOSU CMLOC CMSTDTC CMENDTC ABC-001-0001 1 PALLIATIVE RADIATION 68 Gy PROSTATE 2010-01-20 2010-03-01 ABC-001-0002 1 PALLIATIVE RADIATION 72 Gy PROSTATE 2010-11-01 2010-11-25 NO MATCH ON USUBJID, CMSEQ SUPPCM: USUBJID RDOMAIN IDVAR IDVARVAL QNAM QLABEL QVAL ABC-001-0001 CM CMSEQ 2 REASDISC Reason for Discontinuation Completed therapy SD0078: Referenced record not found SUPPQUAL Example: No Parent SUPPCM: What to do when you do not know what to use for IDVAR/IDVARVAL? USUBJID RDOMAIN ABC-001-0001 CM IDVAR IDVARVAL QNAM QLABEL QVAL REASDISC Reason for Discontinuation Completed therapy Sample CRF Mapping Guidelines 1. Does the data belong in a general observation class domain? 2. Is there a not-so-obvious parent? 3. Is the data actually needed? 4. Time to get creative Applying the Guidelines – Example 1 (1) Does the data belong (3) (2)IsIsthe there data a not-soactually in Time a general observation to getparent? creative! obvious needed? class domain? Applying the Guidelines – Example 1 Applying the Guidelines – Example 2 (1) Does the data belong in a general observation class domain? Applying the Guidelines – Example 2 Applying the Guidelines – Example 3 (1)(2) Does the data belong Is there a not-soin a obvious general observation parent? class domain? Applying the Guidelines – Example 3 Applying the Guidelines – Example 4 (1)(2) Does the data belong Is there a not-soin a obvious general observation parent? class domain? Applying the Guidelines – Example 5 (1) (3)Does (2) Is the Isthe there data data actually a not-sobelong in a general obvious needed? observation parent? class domain? Applying the Guidelines – Example 5 Applying the Guidelines – Example 6 getdata creative! (1)Time Doestothe belong (3) (2) Is the Is there data actually a not-soin a general observation obvious needed? parent? class domain? Applying the Guidelines – Example 6 Conclusion ► Be on the lookout for possible “lost children” when annotating and creating mapping specifications. ► Be mindful of the intent and relevance of the data. ► Consider how/if the information will be used for review or analysis. ► Determine the best fit within the SDTM standard. Thank you Sandra VanPelt Nguyen svanpeltnguyen@pharmanet-i3.com © 2011. All rights reserved. 24