/ 20 May 2020

South Africa could see 40 000 Covid-19 deaths by November

Lethal Virus Military Exercise In Italy
Covid-19 could claim up to 40 000 lives in South Africa, according to experts. (Fabrizio Villa /Getty Images).

South Africa is likely to see about 40 000 deaths from Covid-19 by November, according to mathematical modelling released late on Tuesday. And the country will quickly run out of hospital beds to treat patients.

Scientists now predict that the country will reach the 30 000 mark in diagnosed Covid-19 cases by the end of the month.

South Africa had recorded its first case of the new coronavirus on March 5. It was only a matter of days before journalists and the public alike began to clamour for the mathematical modelling the government was using to prepare for an uncertain future with a virus that globally was just months’ old. 

On Tuesday evening, Health Minister Zweli Mkhize — together with researchers — finally released the sobering projections compiled by teams from the University of Cape Town (UCT), University of the Witwatersrand and Stellenbosch University. 

Health economists and epidemiologists have partnered to produce predictions about the epidemic’s course and cost in the past six weeks. Modellers have projected best-case and worst-case scenarios that are now being used by national and provincial government departments for planning, including budgeting and the procurement of essential medical supplies. 

Even in an optimistic course of events, the country faces tens of thousands of deaths in the coming months as the epidemic peaks in July or August, scientists revealed. Pessimistic projections, meanwhile, place the death toll closer to 50 000.

South Africa does not have enough hospital beds, ICU or otherwise

But, regardless of whether the country follows the best-case or worst-case projections, one thing is sure: South Africa’s roughly 3 300 hospital beds in intensive care units (ICUs) in the public and private sector combined are not going to be enough. 

“Whether we are going to follow the optimistic trajectory of the epidemic or the pessimistic trajectory of the epidemic, the threshold of ICU beds [that we have] is going to be exceeded,” UCT senior lecturer in statistical sciences and modeller Sheetal Silal said. 

Our ICU beds could be overwhelmed as soon as early June — and that this is particularly likely in the Western Cape. 

When this happens, patients in need of ICU beds will have to be accommodated in other wards, and Silal warned this surge in needed beds was likely to outstrip the country’s capacity.

Two-thirds of the country’s ICU beds are in the private sector, explained health economist Gesine Meyer-Rath from the Health Economics and Epidemiology Research Office at Wits.

The government has already been in talks with private hospital groups to be able to pool hospital-bed capacity during the outbreak. This would allow it to place patients without medical aid in private beds if needs be. However, details about what the private sector is expected to charge the government for this plan have not been made public.

In March, several major hospital groups in KwaZulu-Natal, including Netcare, Mediclinic and Akeso, pledged to donate beds to the national response free of charge.

But, because the need for beds outstrips supply, healthcare workers are likely to be forced into making tough triage decisions about allocating beds and other essential equipment. The South African Medical Association and other professional bodies have already issued guidelines about how to triage patients. Wary of leaving doctors and nurses to shoulder the burden of making life or death choices about the rationing of care during the outbreak, private and public hospitals are working to revitalise the hospital ethics boards that should be making these kinds of calls. 

Caring for Covid-19 patients likely to cost more than R25-billion

ICU care — with a price tag of up to almost R9.5-million — is expected to be the single largest cost in treating Covid-19 patients between now and at least November, according to the modelling released on Tuesday evening, which also includes projected costs of items such as additional hospital linen, said health economist Meyer-Rath.

The cost for equipment such as ventilators, masks and oxygen, as well as the staff needed to respond to the Covid-19 epidemic, is estimated to come to between R25-billion and R31-billion. 

Meyer-Rath said modelling projections are already informing how the government is responding to fill the gap both in equipment and beds.

But Silal cautions projections could change. Models are being updated with new information weekly and are likely to shift as the country learns more about how the new virus is playing out locally. Initial models, Mkhize explained, were based largely on information from other countries.

Models don’t predict the future exactly, but they can help us plan for it

To build mathematical models, scientists plug in a set of assumptions based on available data. For instance, some of the data used to build the models presented on Tuesday is based on how patients in Wuhan, China fared, based on their ages. As South Africa gathers more of this kind of information locally, this data will be plugged into models. 

The current models also have other limitations. For instance, researchers cautioned, there is still a lot they don’t know — this includes  how the virus is spreading at very local levels in some areas. Silal added that teams also haven’t been able to account for how South Africans will change the way they live and work to protect themselves — a factor that helped to shape the trajectories of past Ebola outbreaks

“Models are adaptable… [they] have a set of assumptions about the world, and we update those assumptions as new information becomes available,” said modeller Juliet Pulliam, who is the director of the South African Centre for Epidemiological Modelling and Analysis at Stellenbosch University. 

“Models are really tools that can help you think about the future,” she said. “But they don’t tell you the future in the sense that someone with a crystal ball might tell you the future.”

Mkhize said the government will be sharing modelling information publicly more regularly from now on, after criticism  about a lack of transparency.

“I think it’s a requirement as part of democracy and transparency that the public is able to engage and see and critique and criticise the models on the information that are produced,” said Wits University Professor Harry Moultrie. “That’s part of a robust scientific and academic engagement.”

This story was produced by the Bhekisisa Centre for Health Journalism. Subscribe to the newsletter here