Added an optional Description field to the Roles (AKA Data Central Reviewer Roles) configuration in Platform Configuration > Configuration > Roles, allowing administrators to define the responsibilities, goals, and expected review focus associated with each custom role. The Description field supports up to 500 characters and is available in the Roles grid, Add / Edit Data Central Role window, and Configuration import / export workflows.

Added predefined descriptions for default system-generated reviewer roles to provide clearer context and improve consistency across studies and environments. These descriptions are also leveraged by AI-driven capabilities, such as Data Advisor, to better tailor insights, recommendations, and guidance based on the intended responsibilities and review focus of each reviewer role.

This enhancement improves role transparency, supports more consistent reviewer configuration, and enables more context-aware AI assistance across the platform.

Resolved an issue where the unique value list in Subject panel column filters could display “No data to display” despite records being present in the panel. This occurred when the Subjects configuration referenced fields that no longer existed in the underlying data. The filter logic has been updated to automatically exclude invalid configured fields, ensuring unique values load correctly and filters function as expected.

Additionally, validation of the Subject field in Graphical Patient Profile (GPP) configuration has been updated to be case-insensitive, preventing valid Subject fields from being incorrectly flagged as invalid due to capitalization differences.

Users can now review, accept, reject, and modify AI-generated mapping recommendations directly within the Variable List using a structured and auditable workflow.

AI recommendations are visually highlighted in the grid, with indicators at both the cell and row level to guide review. Users can take action on individual or multiple fields through bulk selection, enabling efficient acceptance or rejection of suggestions. Rejected recommendations revert to prior values where available, while accepted or manually modified fields automatically remove highlighting to reflect finalization.

State-based controls prevent edits while mapping is in progress or when specifications are in an approved state. When a specification or domain is approved, all remaining AI recommendations are automatically applied and finalized.

Each action, including acceptance, rejection, or modification, is tracked through real-time audit logging at the variable level, ensuring traceability of changes, including affected fields, action type, user, and timestamp.

The Variable Grid now provides enhanced visual indicators to help users identify missing mappings, required actions, and intentionally unmapped variables.

Required fields, including Source Field, Mapping Method, and Mapping Arguments, are automatically highlighted when empty, making it easier to detect incomplete mappings. These indicators update dynamically as values are entered, ensuring the grid reflects the current state of completion.

The grid adapts based on mapping logic. Fields that are not required for selected mapping methods are excluded from validation highlighting, while variables marked as ‘Not Mapped’ are visually distinguished to indicate intentional exclusion without triggering validation errors.

System-generated recommendations from mapping automation and rules are clearly identified, helping users differentiate between automated suggestions and manual inputs.

A built-in legend is available within the grid, allowing users to reference all highlighting states and interpret visual cues.

The Critical to Quality Factor (CTQ) library has been reintroduced in Risk Management Configuration, providing a centralized source for managing CTQs and CTQ Considerations across RACT studies.

Users can now create and maintain CTQs and CTQ Considerations in a universal library, then select and apply them to Assessment Settings as needed. CTQs and CTQ Considerations can be activated or inactivated at the Assessment Setting level without affecting other settings. Additions, edits, and deletions are managed from the CTQ library and apply to the next version of the Assessment Setting, with indicators shown on the current version when applicable. Ongoing and finalized Assessments are not impacted by edits made in the CTQ library.

These updates help ensure CTQs remain unique and consistent across RACT studies and downstream workflows, including RBQM and IDRP.

Resolved an issue where SCE program output files, such as log and listing files, could overwrite previous versions when a program was retired, promoted back to Production, and re-run.

Previously, when a program at a major version, such as 1.0, was retired and later returned to Production, such as 2.0, the corresponding output files did not retain prior versions. This resulted in version gaps and loss of historical outputs.

With this fix, output files are properly versioned and archived during major version transitions. Each run preserves the prior output version, ensuring a complete and continuous version history aligned with program lifecycle changes.

CTMS imports now support alternate delimiter characters for multi-value fields. This allows object names and other values that contain commas to be imported correctly without being split unexpectedly when commas within values are treated as a delimiter, while maintaining backward compatibility by continuing to use commas as the default separator when no alternate value is provided. Users should include an additional row in the input file to specify the alternate delimiter following the metadata of the input metadata and field values. If not specified, the ingestion process will default to commas for backwards compatibility

The recommended non-comma delimiter for multi-value field elements is now ‘|' suggesting a 'pipe-delimited’ file for any CTMS imports that have commas within those multi-value data fields that will be imported.

Introduced Data Advisor, an AI-powered capability that provides in-context analysis directly within Data Listings and Visualizations. Users can quickly understand data through summaries, role-based insights, and recommendations, explore further by asking questions (via text or voice, in any language), and take action by creating prepopulated issues, all without leaving their workflow. This enhancement helps accelerate data interpretation and supports more informed, efficient decision-making within Data Central.

For an overview of Data Advisor, click: Use Data Advisor in Data Central

The Mapper designer was not correctly distinguishing between implicit schema inheritance and explicit user-defined transformations. As a result, even when users did not override data type or length, the system failed to properly propagate schema changes from the source, breaking the expected pass-through behavior.

To resolve that, the logic was updated to explicitly detect whether a user-defined conversion (data type or length override) exists.

If no explicit override is present → schema dynamically inherits from the source.
If explicit override is present → user-defined transformation is consistently enforced regardless of source changes.

In CDA, the SMQ Comparison Analysis sheet has been enhanced to include a new Both option in the SMQ scope selector, allowing users to view Broad and Narrow terms together. The existing Broad and Narrow options continue to behave as before, displaying only their respective term sets.

This enhancement aligns the sheet with MedDRA specifications and provides greater flexibility in SMQ analysis by enabling combined Broad and Narrow term views.

The Issues panel now provides a smarter, more intuitive filtering experience for string-based columns. When filters are applied in one column, opening the filter popout in another column now visually distinguishes which values remain in the filtered dataset.

Values still present in the filtered results appear at the top of the list (white background, sorted A–Z). Selecting these values further refines the results.

Values no longer present in the filtered results appear below, with a gray background. Selecting any of these values resets existing filters and displays only the newly selected data.

This enhancement helps users quickly identify relevant filter values and make more context-aware filtering decisions.

Custom Listings now provide a smarter, more intuitive filtering experience for string-based columns. When filters are applied in one column, opening the filter popout in another column now visually distinguishes which values remain in the filtered dataset.

Values still present in the filtered results appear at the top of the list (white background, sorted A–Z). Selecting these values further refines the results.

Values no longer present in the filtered results appear below, with a gray background. Selecting any of these values resets existing filters and displays only the newly selected data.

This enhancement helps users quickly identify relevant filter values and make more context-aware filtering decisions.

Enhanced validation for bulk query actions in Data Central to address a rare timing condition that could occur during rapid user interaction. In limited scenarios, a bulk query action could be processed without explicitly validating selected query identifiers, resulting in the Query Audit Report displaying the elluminate User ID as “-Not Applicable-” and incorrectly reflecting the action location.

Additional backend safeguards have been implemented to ensure that all bulk query actions explicitly validate selected query identifiers before processing. This update prevents unintended processing of filtered result sets and ensures accurate user attribution and audit reporting of query activity.

All query operations, including create, answer, close, cancel, reopen, and bulk actions, have been verified to function as expected.

Resolved an issue in the Statistical Computing Environment (SCE) where output datasets generated from SAS program executions displayed zero rows when opened from the output panel. Programs executed successfully and datasets contained data, but the dataset view appeared empty.

This update ensures that output datasets display correct row counts and data content when accessed in SCE.

To improve the security and control users have over their blinded data an update to the configuration of each mapping domain to enable users to selectively determine whether the mapping allows any blinded data when being executed. A new option “Blinded data settings” has been added that allows the user to select one of three options for allowing blinded data through the mapping:

Default - standard behavior, allows blinded data depending on the blinded data view access of the user running the mapping

Mask Unblinded Data - restricts any unblinded values from being pushed through the mapping. All columns using blinded source data will be mapped using their masked values irrespective of the user running the mapping.

Include Unblinded Data - explicitly allows unblinded values through the mapping output. All columns using blinded source data will be mapped with their raw unblinded values irrespective of the user running the mapping.

This options allows users to selectively choose what mappings will allow blinded data and which will not without concern of what users is running any mappings or imports that trigger mappings within the study. Improving the security of blinding in the platform on transformed data. 

Multiple SCE jobs that previously failed with the error ERROR: User does not have appropriate authorization level for file WORK.PROFILE.CATALOG have been resolved. SCE programs now run with an explicitly configured separate SASUSER property for each execution, preventing concurrent SAS sessions from stepping on each other when the profile catalog is opened/updated. As a result, dataset caching and SAS program executions are more reliable and intermittent failures related to WORK.PROFILE.CATALOG contention are eliminated.

User impact

Jobs that were intermittently failing during dataset caching or conversion to sas7bdat should now complete successfully.

No changes to user privileges are required; this is a runtime configuration change that reduces lock/write contention when many SAS sessions run concurrently.

Existing cached datasets continue to behave normally; subsequent runs that rely on cached data should run without the previously reported access errors.

CDA 25.1.3,

CDA 25.2.1

Resolved multiple issues on the SMQ Comparison Analysis sheet where Narrow selections did not immediately update aggregate values and subject counts (N) were calculated incorrectly. The chart logic has been updated to properly isolate SMQ scope selections and ensure subject totals, adverse event counts, and proportion calculations remain accurate regardless of Broad or Narrow term selection.

These corrections ensure that aggregates, subject counts, and percentage values now reflect the correct study population and update consistently when SMQ terms are selected.

CDA 25.1.3,

CDA 25.2.1