DHIS2 & AI
Discover the possibilities to enhance efficiency, improve data quality and use, leverage cutting-edge predictive modeling approaches—and more—by using innovative Artificial Intelligence and Machine Learning tools with DHIS2
Jump to a section on this page
Enhancing DHIS2 with AI tools
As a well-established platform for data collection, management, and analysis in use in over 100 countries, DHIS2 offers many exciting possibilities for leveraging AI to improve the efficiency and effectiveness of the programs it is used to support, whether in health, education, logistics, or other sectors.
Through the DHIS2 Climate & Health project, we have developed cutting edge tools for machine learning-powered predictive modeling within DHIS2: the DHIS2 Modeling app and Chap Modeling Platform. Beyond this, our approach to designing DHIS2 as an interoperable, extensible platform with reusable components and a fully open API is facilitating external innovation using widely available AI tools. There are numerous AI-enhanced DHIS2 innovations that have been deployed in countries where DHIS2 is used, many of which have been shared with the global DHIS2 community.
On this page, you can learn more about the potential for leveraging AI with DHIS2 and explore a selection of existing AI-enhanced DHIS2 innovations. The field of AI is changing rapidly. Helping countries leverage AI effectively is a central component of HISP UiO’s strategy, and we plan to update this page with new tools and resources as they develop.
Improve forecasting and early warning through machine learning-powered predictive modeling
Through the DHIS2 Modeling App and Chap Modeling Platform, users have access to a flexible suite of predictive modeling functionalities directly within DHIS2. These tools make it possible to access, import, train, tune, run, assess, and share predictive models. Through the Modeling app, you can generate forecasts based on the model(s) you have selected, using both health data within your DHIS2 system and climate, weather, and environmental data that is available through the DHIS2 Climate App. These tools are primarily designed to support analysis and early warning for climate-sensitive diseases, but can also be applied to other use cases, supporting health intelligence approaches. These new tools are currently in an experimental phase, and are undergoing iterative testing and development based on real-world feedback.
Facilitate data exploration with natural language queries
With the AI Insights app for DHIS2, developed by the International Federation of Red Cross and Red Crescent Societies (IFRC), you can analyze DHIS2 health data using natural language queries. Healthcare professionals, data analysts, and decision-makers can ask questions in plain language to identify trends, compare performance across organization units, and generate actionable insights from their data.
The Convo app for DHIS2, developed by BAO systems, allows you to ask questions about your data in natural language and have visualizations, tables and interpretations be created automatically using AI, helping users to make sense of their data and take action.
Enhance data capture through text and speech recognition
With AI Data Entry, developed by SolidLines, you can skip manual data entry in DHIS2. Just upload photos or audio, and let AI extract and validate the data automatically with intelligent processing pipelines.
Address data quality issues with algorithmic outlier detection & statistical correction
The Magic Glasses 2 app, developed by John Painter of the US CDC, is a tool that includes features to improve data quality and support health program evaluation. It provides algorithm-based functionality for addressing data quality issues, such as automatic detection of potential outliers and use of statistical techniques to adjust reported values to control for related variables. It also supports evaluation and use of various time-series models to forecast counterfactual scenarios (i.e. what would a disease outbreak curve have looked like if a programmatic intervention had not been carried out), helping evaluate intervention impact.
Streamline data review with algorithmic process automation
Tanzania’s Ministry of Health has introduced an AI-based Alert Triage tool, developed by the UDSM DHIS2 Lab, to their event-based surveillance system for outbreak-prone diseases, reducing average time to triage for the majority of submitted alerts from 36 hours to almost instantaneous. In the first months of its deployment, the system successfully processed 85% of the backlog of more than 15,000 reports.
Improve metadata mapping, naming, and conflict detection with generative AI
Generative AI can help users structure and manage metadata in DHIS2, improving system usability and interoperability.
- Suggest data mappings: Fast-track semantic mapping for interoperable solutions by using large language models (LLMs) to compare field names, code lists, and descriptions from a source system to DHIS2 metadata and proposes the best match.
- Syntactic/Semantic conflict detections: Resolve conflicts, such as between data elements with the same meaning, indicators mapped twice under slightly different names, etc.
- Naming conventions and metadata maintenance: Keep DHIS2 metadata clean and usable by suggesting data element names based on descriptions and conventions (e.g. health area prefixes, country codes).
Accelerate DHIS2 app development with AI tools
There are numerous AI tools available to help developers create new applications. The DHIS2 Developer Relations team has shared a presentation to help DHIS2 app developers take advantage of these tools.
Share AI questions & innovations with the global DHIS2 community
Join the discussion on AI & DHIS2 on our Community of Practice, where you can ask questions, share ideas and approaches, and learn about innovative work being done in our global community.
Committed to Ethical AI
Development and update of these AI solutions for DHIS2 is happening at a rapid pace. The HISP Centre at the University of Oslo (HISP UiO), which develops and maintains DHIS2, and the global HISP network that supports national DHIS2 systems are actively engaged in helping countries take advantage of these technologies. With this rapid innovation has come concerns about ethical use, transparency and ownership. HISP UiO remains guided by principles of country ownership of information systems and data. We are actively engaged in TRUST: The Norwegian Centre for Trustworthy AI, and are committed to supporting the ethical adoption of AI. emphasizing transparency, governance, risk assessment & mitigation, minimizing bias, and maximizing equity & fairness.