DIA EUROPE – ADaM Specifications for Statistical Reporting: Making Documentation Almost Fun
On March 17, IDDI is participating at DIA Europe. Jean-Christophe Chiem, PhD, Senior Statistical Scientist at IDDI, will be sharing his insights on ADaM Specifications for Statistical Reporting.
The rise of electronic data capture in clinical trials has opened remarkable opportunities ranging from real-time monitoring to precision-medicine based on a massive amount of data. However, this large amount of data also requires intense cleaning and documentation activities to enforce the traceability of the data and the reproducibility of the results.
To this end, CDISC has set metadata standards to describe the ADaM datasets, further used to write statistical reports. The corresponding sets of metadata are ultimately processed with Pinnacle21 to produce Define.xml file, which is required for consecutive filing activities such as approval by regulatory authorities. The nature of this documentation makes it highly repetitive. It also represents a considerable workload for the statisticians and statistical programmers and the interpretation of some required information may remain speculative.
In this talk, we will showcase how a user-friendly graphical interface can guide statistical programmers filling in the accurate information.
We also detail how the documentation of ADaM specifications can be optimally automated and limit manual input to the strict minimum.
The integration of this automated process has multiple advantages:
- it avoids error-prone copy-pasting of redundant information among sources,
- it fosters timely documentation and avoids backlog at the end of the study,
- it allows tracking the modification of the datasets to ensure the alignment with reported specifications,
- it clarifies and deploys standards for documentation across the company by limiting individual interpretation of the guidance documents,
- the evolution of standards can be implemented directly in the system.
The optimal automation of the specifications can secure compliance with the Pinnacle21 input format and the exactitude of the corresponding Define.xml.
This whole documentation process can also become a true implementation guide for statistical teams.