Direct comparison
SDTM vs ADaM: Which CDISC Standard Applies When
SDTM tabulates collected trial data for FDA submission; ADaM builds analysis-ready datasets from it. See which CDISC standard applies at each stage.
Written and maintained by CASRAI Editorial Board
Last updated
Ask CASRAI · included with Regulatory Radar
Ask about SDTM vs ADaM: Which CDISC Standard Applies When
Ask CASRAI answers research-administration questions and cites the passages behind every claim — and says so when the corpus does not cover something, instead of guessing. It comes with a Regulatory Radar subscription at $29 a month, alongside the daily digest of regulatory changes and the dashboard of what changed.
150 questions a day, on this site, over the API, or inside your own tools through the CASRAI MCP server.
Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays free to read, with no account and no card.
How do SDTM, ADaM compare side by side?
The table below compares SDTM, ADaM across 12 procurement-relevant dimensions, from full name through nonclinical equivalent.
Side-by-side comparison
| Dimension | SDTM | ADaM |
|---|---|---|
| Full name | Study Data Tabulation Model | Analysis Data Model |
| What it structures | Collected/observed trial data, reorganized into standard domains | Analysis-ready datasets derived from SDTM |
| Stage of use | Post-collection, pre-analysis — tabulates what happened | Pre-submission — supports statistical analysis and the CSR |
| Contains derived variables? | Minimal — close to as-collected/as-reported | Yes — baseline flags, population flags, change-from-baseline, and similar |
| Governing implementation guide | SDTM Implementation Guide (SDTMIG) | ADaM Implementation Guide (ADaMIG) |
| Anchor dataset | DM (Demographics) — required top-level domain | ADSL (Subject-Level Analysis Dataset) |
| Structure classes | General observation classes: Interventions, Events, Findings/Findings About | ADSL, BDS (Basic Data Structure), OCCDS (Occurrence Data Structure) |
| Traceability requirement | Traces back to the case report form / collected data | Must trace back to SDTM (ADaM’s core traceability principle) |
| Typically built by | Clinical data management / SDTM programming staff | Biostatisticians and statistical programmers |
| Companion metadata + reviewer guide | define.xml + SDTM Reviewer’s Guide (SDRG) | define.xml + ADaM Reviewer’s Guide (ADRG) |
| FDA requirement | Required for CDER/CBER submissions per the Study Data Technical Conformance Guide | Required alongside SDTM per the same guide |
| Nonclinical equivalent | SEND (nonclinical/toxicology data) | No direct nonclinical equivalent |
Common questions
Common questions about SDTM vs ADaM
Do I need both SDTM and ADaM for an FDA submission?
+
Yes. FDA's Study Data Technical Conformance Guide requires SDTM (or SEND for nonclinical studies) and ADaM datasets, each with its own define.xml, for standardized study data submitted to CDER and CBER. Japan's PMDA has adopted an equivalent requirement.
Can ADaM be built without SDTM?
+
Not in a way that would pass regulatory review. The ADaM Implementation Guide's core traceability principle requires that every analysis dataset and derived value be traceable back to the SDTM data it came from, so a reviewer can reconstruct how a result was produced.
What is the difference between SDTM and CDASH?
+
CDASH standardizes the data collection stage — case report form fields — while SDTM standardizes how already-collected data is tabulated afterward for submission. CDASH comes first in the workflow; SDTM comes after data collection is complete.
Where does SEND fit in?
+
SEND (Standard for Exchange of Nonclinical Data) is CDISC's SDTM-equivalent standard for nonclinical/toxicology study data rather than human clinical trial data. It is required alongside SDTM and ADaM for the nonclinical portion of an FDA submission.
Who typically builds SDTM datasets versus ADaM datasets?
+
SDTM datasets are typically built by clinical data management or SDTM programming staff, mapping collected data into standard domains. ADaM datasets are typically built by biostatisticians and statistical programmers, since they encode the analysis logic behind the statistical analysis plan.
