- Many local experts have agreed on the potential for a third wave of Covid-19 infections in South Africa
- According to some of them, a spike in case numbers could be seen within the next few weeks
- But a new AI-based algorithm shows that the risk of a third wave is small
Experts have warned that a third wave of Covid-19 infections could hit South Africa very soon. According to a new artificial intelligence (AI)-based algorithm, however, the risk in most provinces of South Africa is low.
The model was designed by the University of the Witwatersrand (Wits), in partnership with iThemba LABS, the Provincial Government of Gauteng, and York University in Canada.
Still, we must not let our guards down, cautioned Professor Bruce Mellado, Director of the Institute for Collider Particle Physics at Wits University, in a news release.
“The current data show us the risk for a third infection wave of Covid-19 is small across most of the provinces in South Africa, but we still remain highly vulnerable.”
The AI-powered early detection system functions by predicting future daily confirmed cases, based on historical data from South Africa’s past infection history, that include features such as mobility indices, stringency indices and epidemiological parameters.
“These parameters are consistent with clinical public health measures that can contain, control and mitigate against the Covid-19 pandemic,” said Dr James Orbinski, Director of the York University Dahdaleh Institute for Global Health Research, adding that South Africans need to continue to adhere to Covid regulations and take the necessary precautions to prevent the spread of the pandemic.
How does the model work?
The AI-based algorithm supports the data of an already-existing algorithm based on more classical analytics, explained the university. Both of these algorithms work independently and are updated on a daily basis.
It has an early detection system that can predict future daily confirmed cases, based on historical data from SA’s past infection history. According to Orbinski, the “parameters are consistent with clinical public health measures that can contain, control and mitigate against the Covid-19 pandemic”.
Since the advent of infection waves are driven by circumstances that are difficult to predict and therefore to control, these kinds of algorithms can provide an early warning to policymakers and the population. When the data display a significant change that is consistent with the advent of a new wave, early detection algorithms are then able to issue an alert.
AI predictions never entirely accurate
Although algorithm-based predictions can never be 100% accurate, Professor Mellado is confident that the new model offers a very strong prediction over at least a two-week period. Predictions can be made over longer-term periods, but will become less accurate, he explained.
“Our team’s development of an early detection algorithm for the third wave speaks to the power of AI to generate data-based solutions to highly complex problems,” said Mellado.
What other experts have said
If South Africa is to experience a third wave of infections, it will likely reach the country in late April or mid-May. Concerning when we might see a spike in post-Easter Covid infections, experts recently told News24 that it could happen as early as next week, or within a number of weeks.
Professor Alex Welte, from the South African Centre for Epidemiological Modelling and Analysis (Sacema), said a spike in cases is dependent on whether super-spreader events took place over the Easter weekend, and how big they were.
"When we have 5 000 or 10 000 cases, we should start to become worried, but now we are averaging around 1 000 cases," he said.
KwaZulu-Natal Research Innovation and Sequencing Platform (Krisp) clinical researcher, Dr Richard Lessells, also said although there was a "clear downward trend" in infections in SA, the virus was still spreading.
"We cannot let our guard down," he said.