The CDISC Analysis Data Model (ADaM)

The Analysis Data Model (ADaM) describes how to create analysis datasets and associated metadata, which in turn facilitates statistical programmers to generate tables, listings, and graphs(TLG’s) super easy. ADaM has an unique feature that enable us traceability from analysis results to analysis datasets, and from analysis datasets to SDTM datasets. A fundamental principle of ADaM datasets is clear communication.

Currently ADaM has three standard data structures: ADSL, BDS, and OCCDS. These analysis datasets follow the ADaM fundamental principles and other ADaM conventions.

Fundamental Principles of ADaM
Communicate “clearly” and “unambiguously”
Provide a level of “traceability” between the analysis data and its source data (SDTM)
“Readily Usable” by currently available software tools
Associate with machine-readable “Metadata”
Analysis Ready Datasets

The Subject-Level Analysis Dataset (ADSL) : Contains subject level information “one record per subject” like subject-level population flags, planned and actual treatment variables, demographic information, stratification and sub-grouping variables, and important dates. ADSL and its related metadata information is required in a CDISC-based submission of data as per FDA TCG, even if no other ADaM datasets are submitted. ADSL is a source for all other ADaM datasets for subject-level variables, such as population flags, treatment variables, subgrouping variables and statistical model covariates).
Required Variables for ADSL:
Identifier Variables: STUDYID, USUBJID, SUBJID, SITEDID,
Subject Demographics Variables: AGE, AGEU, SEX, RACE
Treatment Variables: ARM, TRTxxP

The Basic Data Structure (BDS) : Contains PARAM, AVAL and AVALC and related variables information (Mostly “findings” data) “one or more records per subject, per analysis parameter, per analysis timepoint”.
Required Variables for BDS:
Identifier Variables: STUDYID, USUBJID
Analysis Parameter Variables: PARAM, PARAMCD

The Occurrence Data Structure (OCCDS) : Contains Adverse Events, Concomitant Medications, and Medical History (Mostly “events” and “interventions”).
Required Variables for OCCDS:
Identifier Variables: STUDYID, USUBJID, — SEQ
MedDRA Dictionary Coding Variables: — TERM
Adverse Event Descriptive Variables: AESER(Serious Event)
WHO Drug Dictionary Coding Variables: CMTRT

ADaM OTHER : Analysis datasets which do not follow one of the three defined structures (ADSL, BDS, OCCDS) considered as the ADaM OTHER.

As per the FDA TCG, “core” subject-level variables must present in all analysis datasets. If any variable is present in both ADSL and any other ADaM dataset then it should have same values, type, and
label.

Required documents to create a ADaM datasets

The Study Protocol
The Statistical Analysis Plan (SAP)
Mockup Shells
Case Report Form (CRF)
Mapping Specifications
SDTM Specification and datasets
The Analysis Data Model Implementation Guide (ADaMIG)

Important terms used by Statistical Programmers

CDISC: Clinical Data Interchange Standards Consortium
SDTM: Study Data Tabulation Model(standard for interchange of collected data)
ADaM: Analysis Data Model (standard for interchange of analysis data)
SAP: Statistical Analysis Plan
CRF : Case Report Form
FDA TCG : The Study Data Technical Conformance Guide

Protocol : It describes the study objectives, design, methods, assessments, study timings, endpoints, and statistical considerations for analyzing the data.

Statistical Analysis Plan (SAP) : It describes the planned analysis of the endpoints in the study and more detailed information to programmers how to create the ADaM datasets and the final outputs(TLG’s).

Case Report Form (CRF) : It is the tool used by the investigators to collect study data into the database from each participating patient (paper or electronic from).

Traceability : It allows the understanding of the relationship between the analysis results, the ADaM datasets, the SDTM datasets, and the data collection tool(Metadata traceability and Datapoint traceability).

REFERENCES:
Introduction to Clinical Trials
An Introduction to the Standard Data Tabulation Model (SDTM)
Link between Clinical Research and SDTM
Legacy clinical data for CDISC SDTM compliance and Data Unification
AMALGAMATION OF BIG DATA ANALYTICS, SDTM, LEGACY CLINICAL DATA
Analysis DataModel Implementation Guide(ADaMIG)