At a modelling symposium chaired by Health Minister Zweli Mhize on Thursday, estimates for South Africa’s Covid-19 curve may have varied, but one point was constantly stressed: no epidemiologist or actuary has a crystal ball to predict exactly how the pandemic will play out here.
Presenting on behalf of the Actuarial Society of South Africa (ASA), Barry Childs highlighted the range of opinion on the matter even just within his own industry.
“Some actuaries are saying we’re way too high, as much as four times too high; some are saying we are way too low, as much as four times too low,” Childs said.
Indeed, some of the highest projections presented during Thursday’s symposium strain credulity for laypeople – as journalists pointed out during question time.
A report from health modelling experts at the South African Covid-19 Modelling Consortium (SACMC), for instance, projects as many as 13 million South African coronavirus infections to have occurred by November – a figure which is more than twice as high as the current cumulative recorded coronavirus figures worldwide.
Of these, however, the vast majority of infections are expected to go undetected – with around 3.7 million potentially diagnosed.
That figure is not far off the projections presented by Deloitte’s Ashleigh Theophanides, which predict around three million Covid-19 cases by December.
When it comes to total South African deaths from Covid-19, the number which several different groups are citing is around 40,000.
Deaths in the 40,000-range have been projected by the ASA, Deloitte and the SACMC.
The South African death toll as of 20 May 2020 was 339 – which means that if those estimates are valid, the country is about to enter some very rough months.
Projections of when the country’s infection curve will peak vary. The SACMC is estimating the peak to come between July and mid-August. Deloitte projects the infection curve to peak in July, but the death toll to peak around September.
There is wide consensus that the country’s health facilities are in danger of being overburdened relatively soon.
The SACMC’s Juliet Pulliam said that some projections suggested that the country’s ICU beds may be full as soon as June.
Pulliam said it is estimated that at the peak, between 20,000 and 35,000 ICU beds may be needed, while the country’s current ICU beds number about 3,000.
When it comes to drivers of Covid-19, South Africa’s high population density was identified in two presentations as a major concern.
UCT’s Co-Pierre Georg told the symposium that the data suggests so far that what “really drives” transmission in South Africa is “how many people you are in close contact with”. He said it was already evident from Cape Town data that the dynamic was very different from ward to ward, based on population density.
Said Georg: “In Cape Town, what is driving transmission correlates with density per square kilometre”.
A quite different tone was struck by actuary Nick Hudson, who co-ordinates a multidisciplinary initiative called Pandemic Data and Analytics (Panda), which has previously argued that stringent lockdown measures may ultimately lead to greater loss of life than the pandemic itself.
Hudson says Panda’s own modelling expects a much lower number of deaths in South Africa, largely due to the country’s youthful population, and a peak in daily deaths much earlier than is generally predicted.
Hudson said his research demonstrated that of 66 countries worldwide whose infection curves are believed to have already peaked, 56 peaked within 80 days of the first infection in that country.
Noted Hudson: “The timing of that peak is completely independent of the stringency of lockdown. There is no correlation between lockdown stringency and the number of days to peak.”
The actuary also suggested that on the basis of international examples, there was no reason to believe the lifting of lockdown would cause a “resurgence” of infection. DM
"Censorship of anything at any time in any place on whatever pretence has always been and always will be the last resort of the boob and the bigot." ~ Eugene O'Neill