The challenge: beyond simple validation

While DHIS2 provides built-in tools for data quality —such as validation rules and outlier detection— they often fall short when it comes to governance and long-term oversight.

Standard tools typically offer a “snapshot” of errors at a specific moment but lack a structured way to manage them. Organizations often face the following hurdles:

• Lack of traceability: Native functions list problems but don’t allow you to assign an issue to a specific person, track its progress, or document whether a value is a genuine error or a confirmed outlier (false positive).
• Workflow fragmentation: There is no native bridge between identifying a data quality issue and the administrative process of fixing it. This leads to offline tracking (spreadsheets, emails) and a loss of accountability.
• Communication silos: Without a centralized system to assign issues or leave comments, the feedback loop between data managers and focal points is often broken. There is no easy way to document the status of a data issue, if it is being reviewed or if a manual override was applied based on specific field knowledge, making collaborative cleanup nearly impossible to coordinate at scale.

In short, there was a missing layer of issue management—a way to transform data validation from a one-time check into a collaborative, transparent, and auditable workflow.

Design principles

To solve these challenges, we built Data Quality Extended based on following core pillars:

• Actionable issue management: Every anomaly is converted into a trackable “Issue Object” that can be assigned, commented on, and monitored through to resolution.
• Hybrid analysis engine: The app combines standard DHIS2 checks (Outliers and Validation Rules) with the flexibility to run custom-coded steps for specific departmental needs.
• Unified tracking: Whether an issue is created manually by a user or automatically by an algorithm, it follows a standardized format for seamless reporting and follow-up.

The solution: Data Quality Extended

Data Quality Extended provides a comprehensive management layer that sits on top of DHIS2, turning raw validation results into a functional helpdesk for data managers. While standard systems merely flag errors, our solution facilitates a complete “detect-to-resolve” workflow. By centralizing disparate quality checks into a single interface, it eliminates the need for external spreadsheets and fragmented communication, ensuring that every data anomaly is accounted for and corrected at the source.

How it works:

• Integrated tracking: Provides a dedicated space to manage “issues” as living objects, allowing teams to assign responsibility and track status (e.g., pending, resolved, or false positive).
• Flexible analysis: Supports both generic DHIS2 checks —like Outliers and Validation Rules— and custom-built steps tailored to specific program requirements, such as those used by the NHWA or WHO projects.
• Manual & automated discovery: Beyond automated algorithms, users can manually flag issues, providing a unified repository for all data concerns regardless of how they were identified.

The Tech Stack:

For those interested in the engine under the hood:

• React 18 & TypeScript: For a robust and type-safe user interface.
• @dhis2/app-runtime: To ensure seamless communication with the DHIS2 Web API.
• DHIS2 UI Library: To maintain a consistent look and feel with the native DHIS2 environment.
• DataStore & Tracker: Used for storing app configurations and tracking data quality issues without requiring an external database.

If you want to know more, explore our github repository for this project!

Future work:

The roadmap for Data Quality Extended focuses on deepening its integration within the DHIS2 ecosystem and expanding its flexibility to serve diverse organizational needs. These planned enhancements represent our vision for a more proactive and tailorable data governance toolkit:

• Native Notification & Assignment: Integrating with the DHIS2 native notification system to automatically alert and assign issues to specific users directly within their existing environment.
• Customizable Business Logic: Expanding the analysis engine to allow for configurable issue statuses and actions, enabling organizations to tailor the workflow to their specific departmental protocols.
• Enhanced Automation: Moving toward a more automated “detect-to-notify” pipeline that further reduces the manual effort required for data oversight.