SA’s Covid-19 models were ‘flawed’, says former NICD expert
And no-one is modelling the ‘collateral damage’ of the lockdown, and the impact of the economic meltdown on disease and death
The modelling of how many people would contract Covid-19 and die was “flawed and illogical and made wild assumptions”, says Prof Shabhir Madhi, the former head of SA’s National Institute for Communicable Diseases (NICD).
It’s an important claim not just because of Madhi’s stature (he also co-ran the SA Medical Research Council), but also because the Covid-19 model, which suggested between 87,000 and 350,000 deaths, was instrumental in the government’s decision to implement a lockdown
The government’s initial model also predicted that 600 Covid-19 patients would need treatment in intensive care units (ICU) in SA by April 1. But by April 18, the last publicly released figures showed there were 32 Covid-19 patients in ICU.
While the lockdown would have slowed the spread of the disease, it had only been in place for four days by April 1, so this was likely not the reason why there were so few ICU patients.
Madhi says the initial modelling and fatality estimates were “back-of-envelope calculations”.
The models were prepared by the SA Centre for Epidemiological Modelling & Analysis (Sacema) at Stellenbosch University, as well as the NICD. They used data from Wuhan, the Chinese city where the outbreak began.
The Sacema model modelled its fatality projections around different scenarios in which 10%, 20% and 40% of the population contracted the virus.
Madhi says he raised problems with the model at the time. Nowhere in the world – not even in Wuhan – had 10% of the population contracted the virus. “I told them that at the start, this is implausible,” he tells the FM.
He claims the model was not sufficiently flexible to adjust for infection rates, depending on how many other people in the population had contracted the virus. As it is, the more people who are infected, the fewer new people a contagious person can potentially affect.
This week, the FM asked Prof Juliet Pulliam, the head of Sacema, why their model was based on a minimum 10% infection rate, and whether they still believed deaths would be as high as initially predicted.
Pulliam responds: “The model you refer to was a preliminary assessment that was based on the best available information at the time. A new modelling report will be released by the NICD [this week].”
However, Sacema’s model, which was instrumental in shaping public policy, could have been distributed for review by other experts in real time. This is an approach adopted by some other countries.
In its defence, Sacema’s model isn’t the only one to be adjusted downwards. The UK’s Imperial College London initially predicted huge numbers of deaths, including 2.2-million in the US, but this has since been toned down.
However, its initial figure of 350,000 deaths was frightening – 10 times more than the number of people who die annually of tuberculosis, SA’s biggest killer. Madhi says those mortality rates were based on “wild assumptions”.
Madhi himself estimates that over the next two years, there may be 43,000 Covid-19 fatalities in SA, as the virus waxes and wanes. This year, he estimates there will be 25,000 deaths.
Wanted: a smarter strategic response
That 25,000 number sounds high, but Madhi says that’s probably equal to the number of people who die of TB in SA each year, which would never lead to a national lockdown. In part, he says, this is because TB is out of sight – a disease of the poor which predominantly kills in the rural areas.
Stats SA figures show that in 2017, 25,336 people died of diabetes and 22,259 died of cerebrovascular disease, which includes strokes and aneurysms. That year, 28,678 people died of TB.
Madhi says the Covid-19 deaths, however, are likely to happen within a short period, during winter when hospitals typically have high numbers of flu cases too. Typically, hospital wards reach 90% capacity with peak flu cases, and Covid-19 could make this a lot worse.
This appears to have been the rationale behind locking down the country and slowing transmission.
However, public opposition to the lockdown has grown steadily as the economic situation worsens: experts estimate that more than 1-million people will lose their jobs and the economy will shrink by more than 6%.
Madhi argues that the initial lockdown was necessary to increase hospital preparations and testing and to slow transmission, so that the hospital system is not overwhelmed. The fact that fewer people used taxis and the affluent stayed at home has definitely slowed the spread of the virus, he says.
The problem is, he says, that there isn’t enough testing happening, nor enough isolation beds for people with Covid-19 who live in high-density areas. Which doesn’t mean the government isn’t trying: 2,000 beds are being prepared for quarantine at the Nasrec expo centre.
The NICD, an agency of the department of health, has previously told the FM that it cannot make its models public due to a confidentiality agreement with the department. But this means it’s not possible for epidemiologists to review and critique the assumptions.
Madhi also warns that no-one is modelling the “collateral damage” of the lockdown, and the impact of the economic meltdown on disease and death. He says the government’s focus is on decreasing infections and fatalities rather than the broader picture.
At this point, the economic impact is unclear. Neither economists nor business leaders have provided any estimates of “life years lost” because of the increase in unemployment.
However, as Madhi says, poverty can make things worse. It increases the risk of contracting TB in high-density settings like informal settings and fuels HIV infections as women trade sex for money, to cite just two examples.
A cautionary tale from the DRC
However, a bad response to a disease can be disastrous.
In the Democratic Republic of Congo (DRC), a strict Ebola response decreased movement and led to a halt in the country’s wider disease vaccine programme. The Ebola outbreak killed 3,500 people, but the country then had to deal with the world’s worst measles outbreak, which left more than 6,000 children dead.
On Thursday, the World Health Organisation (WHO) warned that lockdowns should not stop the immunisation of babies. It cited the measles deaths in the DRC as an example of why this is necessary.
WHO director-general Dr Tedros Adhanom Ghebreyesus said: “While the world strives to develop a new vaccine for Covid-19 at record speed, we must not risk losing the fight to protect everyone, everywhere, against vaccine-preventable diseases. These diseases will come roaring back if we do not vaccinate.”
Madhi, who himself is a global expert in vaccine science, is particularly worried about how few vaccinations for measles have been given to children in SA over the past month.
To subscribe, for free, to the FM's lockdown newsletter click here.