Data Traceability in Clinical Submissions

Every number appearing in a clinical study table submitted to the FDA must be 100% traceable back through ADaM analysis datasets, SDTM tabulation domains, down to the original raw EDC electronic case report form (eCRF).

VISUAL FRAMEWORK: 4-STAGE CDISC DATA FLOW 1. Raw EDC eCRF Data (Unstandardized) 2. SDTM DM, AE, LB, VS (Tabulations) 3. ADaM ADSL, ADAE, ADLB (Analysis Ready) 4. TLFs Tables & Figures (Regulatory CSR)

The 4 Critical Stages of CDISC Flow #

  1. Raw Electronic Data Capture (EDC): Patient observations entered by clinical research coordinators at investigative hospital sites.
  2. Study Data Tabulation Model (SDTM): Standardizing variables into CDISC general classes (Events, Findings, Interventions). SDTM never imputes missing data or performs statistical derivations.
  3. Analysis Data Model (ADaM): Deriving endpoints, baseline flags (ABLFL), treatment windows, and subject demographics (ADSL) so statistical procedures (PROC FREQ, PROC GLM) can run with 1 step.
  4. Tables, Listings, and Figures (TLFs): Formatting analytical outputs into clinical study report appendices using PROC REPORT and strict regulatory pagination.
🔬 Authoritative Knowledge Hub Blueprint

Deep Dive: ADSL Subject-Level Analysis Dataset & SDTM vs. ADaM Architecture

Master the exact SAS DATA step derivations for ADSL (TRT01P, TRT01A, SAFFL, ITTFL) and compare SDTM vs. ADaM datasets in our Clinical SAS Knowledge Hub.