Demonstration of the VODAN Africa Platform - Data Pipeline, FAIR Principles, and FAIR Data Stewardship
The VODAN Africa digital health system was developed with the purpose of creating Findable Accessible Interoperable and Reusable (FAIR) Data at the point of service in order to improve the quality of health care. This was because the data produced were sent to the Ministry of Health, and none were available at the point of production. VODAN Africa developed a digital health system that produces FAIR Data without losing its provenance in a federated manner. This concept was validated with clinical and research data. The VODAN Africa Initiative also developed a data stewardship program that trains experts in handling, processing, analyzing, and visualization of FAIR Data.
FAIR Guiding Framework
The FAIR-OLR (ownership, localization, and regulatory compliance) is focused on these three areas
- The data is owned by the health facilities
- The data produced should be in residence (at the point of production)
- The data should be compliant with local and international regulatory framework
VODAN Architecture
- Uses CEDAR workbench to collect data and create metadata templates
- Data is entered only once and visualized by the internal dashboard and the external dashboard
- The data is stored using Triple Store
Data FAIRification at Health Facilities
- Each facility has a CEDAR workbench for data collection
- Data is FAIRified through bulk upload
- The system interacts with the external dashboard
- Each health facility use controlled vocabulary and can create new vocabulary as the need arises
- Federated query can be performed, and algorithms can be deployed
Experiences and lessons from Mekelle
The team at Mekelle led by Samson Yohannes, Getu Tadele, and Tesfit Gebremeskel has been resilient towards the deployment of the VODAN Africa digital health system at 10 health facilities in the Tigray region, despite the sustained bombardments from the Ethiopian Federal Forces
Due to limited resources as a result of the crises, the Mekelle team developed a lightweight system that served as a Minimal Viable Product (MVP) for the health needs of the region.
- Used CEDAR as template engine
- Built a new system to create FAIR Data
- The same system sends a report to DHIS and triple store
Achievements
- FAIR Infrastructure Built: infrastructure for federated learning, metadata template creation, and FAIR data collection, federated AI-ready data for forecasting pandemic
- Accessibility: secured accessibility and backup, SPARQL, and other API endpoints
- Standardization and FAIRification of Existing data: prospect for integration of FHIR resources, platform to ensure standards and provide services,
FAIRification of bulk existing data - Capacity Building: Data Stewardship Program, GAIC, DISH, UNA, FAIR Connect
FAIR Data Stewardship Program
The purpose of the program is to develop capacity building for data clerks, data stewards, and domain experts on data FAIRification. Under the Nuffic DISH Computer Science Level Courses, VODAN Africa has a course specifically for FAIR Data Stewards.
They are trained in the following areas
• Metadata templates
• FAIR Data
• Vocabularies
• Access, control, and security
• Data visiting (querying)
• Internet of FAIR Data and Service
• Understand VODAN Africa Minimal Viable Product (MVP)
In the next phase, VODAN-Africa will now be setting up the International FAIR Data Stewardship Center for FAIR Data creation in Africa, under the leadership of Hon Julia Duncan Cassell from Liberia and Dr. Sakinat Folorunsa, from Nigeria.