...
All database validation and user acceptance documentation should be readily available to the study personnel to ensure that all database functions being performed on a study have been validated for quality and reliability. Additionally, consideration should be given to ensure that project team access to clinical data is sufficient to expedite efficient and high-quality interim reporting, data-metrics evaluation, and safety-reporting requirements (see the Safety Data Management and Reporting chapter and the Measuring Measuring Data Quality chapter chapter).
The need for thorough validation and trial database design is even more critical for trials utilizing electronic data collection. If a software malfunction or unintended loss of information occurs, data collected on paper CRFs can be re-entered. Because electronic data collection eliminates paper documents as an intermediary step between the study observations and the database, it is critical that database validation and reliability issues are resolved on the validated system prior to the entry of any study data. As electronic data entry moves closer to the point of patient care, electronic information more often will be the source data and, as such, require protection and secure warehousing.
...
Database design specifications: Documentation of the table definitions used to build the study database and file structure.
Raw data: The final raw data files preserved within the study database format, and all original data transfers in their raw format.
Audit trail: A complete electronic audit trail documenting all modifications to a database by date, time, and user identification.
Final data: Preserved in a standard file format (e.g., ASCII, SAS transport) that can be easily accessed, reviewed, or migrated to another system.
Original study documents: The original and/or scanned images of all original documents, which may be archived separately in a central records facility if necessary.
Procedural variation documentation: Memos and relevant information about any variations from standard operating procedures or working practices that occurred during the trial.
Database closure: Documentation of each database-lock and unlock, describing the time and conditions surrounding those procedures (for additional information, see the Database Closure chapter)
Site copies of data: May be required for audit purposes; if needed, these copies should be locked, read-only datasets delivered on CD-ROM or a similar storage medium.
...