Africa COVID-19 Response Toolkit: Dashboard
This post is the second of a series on the Africa COVID-19 Response Toolkit, an open-source digital reporting platform developed by the Ethiopia COVID-19 Response Team, highlighting its purpose, key features, and instructions for deployment.
What is the goal?
Like in many countries, the Ethiopian Ministry of Health has an urgent need for up to date information on the number and location of both potential and recorded cases across the country. As they collect more surveillance information from a range of sources on those who may be at risk and those who have COVID-19, there is an increasing need to analyze and display this data in an easily accessible, digestible and usable format.
The goal of this project was to develop a dashboard that would collect, analyze, and display key information from across the country in real-time in order to inform the decisions of public health officials. This information is displayed in a number of graphs, charts, and other data visualizations to help illustrate the current status of confirmed and potential COVID-19 cases tracked by the Ministry of Health. The main focus of this analysis is on survey forms collected by the Ministry of Health to track potential cases. This project was pursued at the request of the Ministry of Health. Given the early stage of this project measuring its exact success is ambiguous however it will primarily depend on whether it is consistently used to make actionable decisions by the Ministry of Health.
How does it work?
Dashboard front page
The front page of the dashboard provides high-level statistics on the current state of COVID-19 cases in Ethiopia. This data is pulled daily from the Prime Minister’s Office (PMO). This data includes essential statistics, such as: the total number of cases, hospitalization status, travel history, and a day/week/month breakdown of case occurrences.
Survey Heatmap
We are also developing a heatmap from data collected from the suspected cases from publicly collected surveys. We are developing features where we will be able to toggle and filter these surveys on the heatmap so that they can analyze where more or less surveying is needed or the distribution of specific symptoms.
Medical Resource Tracker
There is also a view to analyzing the current level of medical resources across different regions to track resource needs in real-time.
What has been the progress so far?
In the first few weeks a number of key steps have been completed, which have resulted in an initial product being deployed. This is now being refined based on feedback and further features added.
- Gathering and Cleaning the data: The first step involved finding data sources to collect our real-time data and then manipulating that data into a manageable and easy format to use for our charts and visualizations. This took place primarily over a weekend hackathon.
- Gathering requirements: During Sprint #2 the team met with a group of epidemiologists to conduct a review of the dashboard and to suggest changes to the given visualizations. These suggestions were then turned into tasks for our development team.
- Building the visualizations: Following this, we then worked to architect how we would construct these visualizations and wrote the code to complete the suggestions laid out from our requirements.
- Expanding the feature set: Beyond this, we began to include additional features such as multiple language support and incorporated new data sources such as medical resourcing tracking.
- Deployment: The Dashboard was then released, available for initial use and review for testing.
- Refinement: The exact use cases for the dashboard are still evolving and therefore we will continue to refine it over time, depending on how people are using it and what data they deem is most essential.
Two key types of people have been essential to this project: software developers and epidemiologists. While the core of the team has been developers tasked with creating an initial product, input from epidemiologists has been vital to understanding the types of data visualizations which will be most useful to public health professionals and officials using the dashboard. The dashboard is currently open-source with anyone able to make suggestions for how to improve it and build out additional features.
Modules and materials
Below is a list of the component pieces of the Dashboard, including descriptions and links to the underlying materials.
Dashboard
Here is a link to the live dashboard
Example Surveys & API
Below are some examples of the survey forms used to collect the data imported into the dashboard in the heatmap/surveillance visualization. The API documentation is also shared below to give you an idea of all the available models and data you can work with/visualize.
Medical Resource Tracker
This is an example of the data used to populate the medical resource tracking visualization. It contains information on important details of each facility including the number of PPE, tests, and available beds.
What are the key lessons learned?
- Focus on getting the product management right early on: Our biggest learning was the need to ensure there were effective product management structures in places from the beginning, this includes setting up operational practices such as a ticketing system/project management dashboard for completing work
- Get input early on from scientific people and health care workers: We learned to not only focus on building the technical solution but also to get input from the end-users early on in order to ensure the solution met their needs
The Africa COVID-19 Response Toolkit is a blog published by the Ethiopia COVID-19 Response Team (ECRT).
Sign-up here to join our mailing list to receive updates and more insights from the Ethiopia COVID-19 Response Team.
If you’d like to submit a story to the Africa COVID-19 Response Toolkit, comment below or connect with us on Twitter, Instagram, Facebook, and LinkedIn.