Wits, GCRO, IBM and Gauteng Province opens Covid-19 dashboard to public
- Solomon Assefa and Zeblon Vilakazi
The Gauteng Province has been using data and cloud technologies to monitor and respond to Covid-19, and now they are sharing access with the public.
As of 20 August 2020 the Gauteng Province in South Africa has 33% of the national cases for COVID-19 with 202,000 confirmed cases — and the numbers continue to rise.
To address this challenge, the Gauteng Province has been using a visual analytics platform developed by IBM Research Africa in collaboration with the University of the Witwatersrand (Wits University) and the Gauteng City-Region Observatory (GCRO). The dashboard visualizes anonymous data from multiple sources, including the University of Pretoria and GCRO, to help officials create policy based on data-driven decisions and to help the public understand their risk when they leave home.
The dashboard was designed to address three key questions for policy makers:
- Where are the current hot spots of infection?
- What are the predictions for the spread of the virus?
- What are the risk factors that make certain communities more susceptible than others?
For example, using the dashboard we can see that the initial locations of the pandemic hot spots were in and around Sandton as business travellers returned from overseas. In April the virus began to spread to other districts in the province, and thereafter we can see that the hotspots have appeared in the Joburg CBD, Tembisa and Soweto. Such insights can help officials decide where health workers are needed and the type of intervention strategy.
Mduduzi Mbada, Head: Policy Research & Advisory Services Unit, Office of The Premier, Gauteng Province recorded a video for the Think Summit stating: “We believe this pioneering initiative will help us win the fight on flattening the curve, but also to build a strong public health system and bring back the economy because of the impact of COVID-19.”
In addition to data visualization, the dashboard also has a “what-if” prediction engine developed by epidemiologists and data scientists at Wits University. The engine provides predictions for different lockdown levels for non-pharmaceutical intervention control strategies. Highly sophisticated data analytics such as these can give officials the appropriate instruments of estimating how many beds and ventilators will be needed at the peak of the wave based on level of alert.
Policy makers can see the expected usage of hospital beds and the different kinds of health care setting using data from the Gauteng Province. Thanks to inputs from the Gauteng Province Department of Health, the dashboard gives a detailed account of the needs of the province as a function of time and location.
Starting today the platform is now available for free to the public via the IBM Cloud. Using the latest Chrome or Firefox web-browsers anyone can access the dashboard to learn:
- How many active cases are there in my neighbourhood?
- When is the predicted peak for my neighbourhood?
- Are the number of cases rising or falling in the area of my local supermarket?
“I am thrilled to see a U.S. company like IBM collaborate with Wits University to assist with the public-health response in the battle to defeat the COVID-19 pandemic. This is a best practice on how society benefits when government, academic and business collaborate. The All-of-America response reflects the generosity of American spirit and the importance of our partnership with South Africa in the long-run,” said Lana Marks, U.S. Ambassador to South Africa.
The research behind the predictive engine has been submitted for peer-review and is currently available in the medRxiv offers the following conclusion:
While lockdown measures have been successful in curbing the spread, our study indicates that removing them too swiftly will result in the resurgence of the spread within one to two months. Reducing the stringency index by 10 will delay reaching the apex by about 6 months, where reducing it by 20 will delay by only four. This indicates that post-lockdown measures should be staged and the reduction of the stringency index should be slow.