(Published - 9 July 2020)
Updated COVID-19 epidemiological models from the University of the Western Cape (UWC) and international collaborators suggest the Western Cape has already reached its pandemic peak.
Professor Nico Orce - from UWC’s Department of Nuclear Physics and Nuclear Astrophysics - as well as national and international collaborators have been modelling and monitoring the evolution of the COVID-19 pandemic for months.
Their simplest “D” (for Deaths) model - originally developed by Prof Orce’s very own lecturer in Quantum Mechanics, Jose Enrique Amaro from the University of Granada in Spain - indicates that the COVID-19 death rate in the Western Cape has already reached its peak towards the end of June. Furthermore they found that it should fall close to zero by the end of August, with a total mortality at less than 3 000.
“Our model fits well to available data, suggesting that we've already gone through the COVID-19 pandemic peak in the Western Cape,” Prof Orce said. “This is also consistent with the decreasing WC/SA death ratio, as the pandemic evolution seems to be delayed for the rest of the country. We are currently climbing down the peak.”
“Unfortunately, the rest of the country isn’t quite at the same point - and that’s especially true in Gauteng, where they seem to be peaking very steeply right now and driving the curve up. It’s a very problematic situation, and unless there are decisions and interventions made for the province, it’s not looking good.”
But there’s some hope though that this scary trend may be due to anomalous or delayed counting.
“For instance, the number of deaths in Gauteng surprisingly stayed at 282 during July 2, 3, and 4, and then there was a sudden jump for the next three days. Upcoming data in the next few days will be crucial.”
Prof Orce and his collaborators have been following how their different models match or diverge from the reported virus patterns in other countries, including China, Italy, Spain, UK, France, Belgium, Sweden, Iran and the US.
“Deviations from a model prediction are a good thing when you know the limitations of your model. They can tell us more about aspects originally ignored by that particular model: the lockdown or spatial migration effects, for example.” Prof Orce noted.
Spatial migration has been more of a concern since South Africa moved into lockdown level three, with many of the “harder” lockdown measures relaxed to varying degrees.
“When people move from areas that are more badly affected by COVID-19 - like the townships in Cape Town, for instance, where people are less easily able to self-isolate while doing what’s necessary to provide for their households - that broadens the possibilities for viral spreading and secondary waves. We urgently need to reach that level of transformation (housing, public transport, jobs) if we don’t want more devastating effects during future pandemics.”
Now is the perfect time to test slightly more complicated models that they also developed to reproduce the asymmetric curve typically observed on the way down the pandemic peak - and Prof Orce is keeping a close eye on the official statistics.
“It takes about 30 days, on average, between infection and death, for those who don’t recover,” Prof Orce explains. “And it’s been a little over a month since lockdown level 3 started - so now we’re getting a very clear idea of the effects of lockdown on the model, and on the epidemic itself.”
Modelling, Sharing And Collaborating
As a nuclear physicist, Prof Orce is no stranger to modelling. And the global models considered in his work, outlined in a recent paper on Global analysis of the COVID-19 pandemic using simple epidemiological models, provide an elegant way to distinguish between the scary exponential phase and the normal phase, where the flattening of the curve is observed.
Prof Orce and his collaborators’ model has been very good at predicting deaths all the way to the peak for many countries.
“You don't see many models out there where incoming data are used to characterize the evolution of the pandemic,” Prof Orce said. “But the more data we have, the more accurate our prediction, and the more we are able to see where it needs to be modified. It’s an important lesson for science in general: The more we observe, the more we learn.”
Graphs generated from the model are updated every day and shared on social media. People all over the world have viewed, commented and contributed to making graphs more user-friendly and exploring possible impacts.
“This has been a very pleasant surprise to me. Many people - including me - sometimes have the feeling that social media is full of rubbish. But we have been using it more often lately to motivate people during these hard times, and soon many nuclear physicists from all over the world were contributing to the development of the model, and other physicists were using it as computer practicals and science outreach.”
Still, to beat the pandemic, we’ll need more than models.
“Independently, we must all do our bit to help stop or prevent the spread during Level Three using masks, keeping distance and avoiding large gatherings,” Prof Orce says. “If we all make an effort, however small, together we can make a big difference. Education provides the most powerful weapon to achieve this transformation, Madiba told us.”