Menu Back to Poster-Presentations-Details

W-18: SDTM/ADaM Compliant or Sufficient Submission for Legacy Studies

Poster Presenter

Nancy Wang

Associate Director, Biostatistics
Celerion
United States

Objectives

To propose content of the data submission package from the different approaches for preparing SDTM/ADaM compliant or sufficient data submission for legacy studies since retrospectively created SDTM data alone do not provide direct links to the study summary results programmed from legacy data.

Method

Identified the different stages of study and format of data/programming of legacy studies and proposed the content for a data submission package that is compliant or sufficient as a SDTM/ADaM submission. The goal is to provide data and supporting information compatible with SDTM/ADaM requirements.

Results

Implement the full SDTM and ADaM standards for a study that is at the beginning stage of data analysis. It is also the best approach for a legacy study that is still at clinical study report (CSR) preparation stage in order to avoid complication that might arise from retrospectively creating datasets, tables/figures/listings (TFL) and CSR since the time stamps should be in sequential order of collection datasets, analysis datasets, TFL and CSR. If a legacy study CSR is finalized, it may not be necessary to re-create the CSR from scratch starting with SDTM/ADaM. Adequate information can be prepared to support a successful regulatory submission. If TFL were programmed directly off the collection datasets without analysis datasets, the submission should include collection datasets in SAS transport format, annotated Case Report Form (aCRF), corresponding data definition and reviewer’s guide along with the SAS/ program code that created the TFL. If a file with subject level information to identify treatment and analysis population (similar to ADSL in ADaM) was used, it should be included. Also include explanation on queries that would be raised as if OPEN CDISC (now Pinnacle 21 Community) equivalent checks were run against the data. If analysis datasets were created, then a SDTM like packet (with collection data, aCRF, data definition and reviewer’s guide) should be submitted as well as a ADaM like packet that includes the analysis datasets in SAS transport format, data definition which details analysis data derivations, reviewer’s guide to link analysis datasets to collection datasets and TFL, and SAS/program code used to create the TFL. An alternate approach is to prepare SDTM/ADaM datasets from legacy data following guidance which requires a re-program of all TFL using SDTM/ADaM with a compare of the previously generated final TFL to show no change in the TFL. Submit the SDTM/ADaM packets and retain the proof of no change in the TFL. The final CSR remains as is.

Conclusion

Certain submissions of clinical trials started after December 2016 are required to have data submitted in SDTM/ADaM format by the US Food & Drug Administration (FDA). A SDTM packet includes SAS transport datasets prepared according to required specifications, annotated Case Report Form , a data definition file (define.xml) and a study data reviewer’s guide (SDRG) on data description and conformance. An ADaM packet includes analysis datasets derived from SDTM datasets, define.xml (describes format and derivation of analysis variables), analysis data reviewer’s guide (ADRG) that provides the links between analysis datasets and SDTM datasets and the TFL in the CSR, and SAS/program code used to generate TFL using ADaM datasets for the CSR. Open CDISC (now Pinnacle 21 Community) conformance check report can also be included. To fully convert a finalized legacy study to follow SDTM/ADaM specifications for data submission is time consuming and costly while retrospectively created SDTM data do not provide adequate links to the TFL in the final CSR for reviewer verification. It is possible to prepare a successful data submission for legacy studies if complete data and necessary information are provided to regulatory reviewers. Appropriate approach to successful data submission to FDA can be determined for legacy studies based on the status of the study, structure of the data and how TFL are programmed. The basic concepts and specifications of FDA required SDTM/ADaM submission packets provide valuable guidance on what information a reviewer would need to effectively perform regulatory responsibility in a timely fashion. The key is to ensure that transparency and traceability are achieved in any data submission for a smooth regulatory review.