Global Health: The Age of Big Data
As mentioned in a previous post: internet coverage is increasing in developing countries and universal internet access is no longer just a dream. So, what does this mean for global health? What does this mean for health data systems?
In the past decade, countries have monitored their health programs based on irregular surveys and health ministries have struggled to build reliable routine data systems. However, the rise of the internet and mobile technologies has led to a boom in data sources in developing countries — especially in Sub-Saharan Africa.
Vertical programs (e.g. HIV, malaria), NGOs, national health management information systems (HMIS) are digitizing their data collection process. In addition to these domestic data sources, there is increasing data that can be used for better health system stewardship, including mobile phone data (call records), demographic data and high resolution satellite imagery.
Yesterday the challenge was to build data sources, but today countries have to organize this “ocean of data”.
Is global health entering the age of big data?
Big data is often defined by 3 Vs: the (extreme) volume of data, the wide variety of data types and the velocity at which data is processed. By this definition, it is safe to say that global health is moving towards the era of big data. And while we’re not there yet, we can learn from other fields how to prepare health data systems for the upcoming challenges.
Three of these new challenges can be defined as: access, integration, and use.
- ACCESS — Data should be freely available for everyone to use without copyright restrictions (except for private/personal data). Both the private and public sectors are creating new data sources which should be open to everyone: civil society, governments, the aid sector etc. At BlueSquare we are committed to open data, and have joined the Advisory Group for Open Data for Developing Economies, an initiative from The GovLab, Web Foundation, USAID and FHI 360.
- INTEGRATION — To make a large variety of data sources available, we have to: map existing data sources and limit fragmentation. Today, health data systems in developing countries are fragmented, which leads to multiple data sources. These “data islands”, mean that existing data systems do not allow for information to be exchanged or cross-checked with other systems. We are currently working with national HMIS teams to build data warehouses, whereby stakeholders can find relevant and high-quality data for program monitoring.
- USE — Don’t collect data if you’re not going to use it. We build data systems to improve decision-making and to better support strategic purchasing of health services. So, we need to make sure that people can actually use the data collected through those systems. Health staff need to have the right tools and expertise. Countries should therefore invest in the development of their staff’s data culture. Because, as the variety of data grows larger and wider, the challenge of using the data in a meaningful way only increases.
We’re living in exciting times. We have an opportunity here to (re)invent public health analytics in the light of these changes, and have a real impact on health data systems.
In my next post, I’ll explain how we’re working to concretize this vision in Benin with multiple stakeholders. Stay tuned!