On Monday 23 March, President Cyril Ramaphosa announced that from midnight on Thursday, the country would go into lockdown for 21 days. This was done in an effort to halt the spread of the runaway SARS-CoV-2 virus which causes the respiratory disease known as Covid-19.

Essentially, during that period, the whole country, other than essential services, will come to a stop. The cost of the decision is impossible to calculate and the damage to an already weak economy will take decades to recover from, if the country makes the right decisions.

Hundreds, if not thousands of firms, the good and bad, will go out of business and with it an already dire unemployment rate will go up. Whole swathes of the economy, including those that had enormous job creation potential, like tourism, might never fully recover.

In making this decision, South Africa joins a raft of other countries, mostly in Europe, doing some version of the same thing. The US is presently under a 15-day lockdown. 

South Africa is different in one respect: it implemented its lockdown far earlier into the trajectory of infection than other countries did. Indeed, Boris Johnson, the UK prime minister, announced his government’s measures just hours after Ramaphosa’s announcement, but did so after 6,650 people had tested positive for the virus and after 335 patients, who had also tested positive, had died.

UPDATE on coronavirus (#COVID19) testing in the UK:

As of 9am on 23 March, a total of 83,945 have been tested:

77,295 negative.
6,650 positive.

As of 1pm, 335 patients who tested positive for coronavirus have sadly died.

The digital dashboard will be updated later today. pic.twitter.com/NwiiPlZzKZ

— Department of Health and Social Care (@DHSCgovuk) March 23, 2020

In South Africa, with 402 testing positive as of 23 May, (see: South Africa’s confirmed coronavirus cases jump by 128 to 402) no deaths have yet been recorded.

The UK government’s decision is a departure from its earlier policy of attempting to slow the spread of the virus, do more to isolate the vulnerable (older people or those with compromised immune systems), but then to allow the majority (around 60%) of the population to become infected, get sick, recover and by doing so become immune and thereby halt the further spread of the disease (so-called “herd immunity”). That approach would have also allowed the UK economy to largely continue as before.

The UK and the US approach was radically changed following the publication of the so-called Imperial College paper Impact of non-pharmaceutical interventions (NPIs) to reduce COVID- 19 mortality and healthcare demand. It pointed to the danger that mere mitigation would result in health systems being overrun at around the same time by serious and critically ill patients, particularly those requiring intensive care unit (ICU) treatment. 

The paper predicted that the result of mere mitigation efforts there would be as many as 250,000 deaths in the UK and 1.1-1.2 million in the US. In addition, it is not clear whether having been infected does provide immunity. Clearly, this horrifying prediction was completely unpalatable, and it forced a reversal of existing policy in both countries. The impact of the Imperial College paper has gone far beyond just the UK and US. It would almost certainly have been a key factor behind the Ramaphosa government’s decision on Monday.

Something about the SARS-CoV-2 virus: It is a coronavirus – a relatively simple RNA virus (as opposed to a DNA virus). Chinese scientists were able to publish the virus’s genetic sequence less than a week after they isolated it. (Read: The race to produce a vaccine for the latest coronavirus) Just six other coronaviruses are known to infect humans, causing normal colds, but two of them, the Middle East Respiratory Syndrome (Mers) and Severe Acute Respiratory Syndrome (Sars), cause severe diseases. (See: What We Know So Far About SARS-CoV-2.) 

It is not clear why some coronaviruses are relatively harmless and others dangerous. An up-to-date and accessible summary of what is known about the SARS-CoV-2 and Covid-19 is hosted by Our World In Data here: Coronavirus Disease (Covid-19) – Statistics and Research which, in turn, uses data from the European Centre for Disease Prevention and Control here: Covid-19. At present, gigantic amounts of research and data are being generated all over the world, from appropriate policies to prevent the spread of the virus, to potential drug treatments and of course, the potential of a vaccine. (Novel Coronavirus Information Center). The World Health Organisation (WHO) has also put together an excellent video of the disease: The Coronavirus Explained & What You Should Do, and its public health implications.

In the absence of proven drug treatments or a vaccine, the following related issues are important: how and how quickly does it spread; and how sick do people get (and what is the mortality rate) across different classification criteria? It is these that drive policy or, as we see, dramatic interventions of nationwide lockdowns.

We know that SARS-CoV-2 is highly infectious and more so because carriers are asymptomatic (they don’t show symptoms) for about 5-6 days before they get sick [Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (Covid-19)], but this can be as long as 14 days or perhaps longer. To be safe on this metric, the South African government has given itself 21 days. It is possible that some carriers don’t get sick at all. Key measures used here are related measures, namely the “doubling rate” and the basic reproduction rate (represented as R0).

The doubling rate or exponential growth is introduced to children by the (mythical) story of the man who once got the emperor of India to pay him in rice with the formula that he would start with one grain of rice and double the number for each block on a chess board. An online maths game for kids demonstrates the point: The Rice And Chessboard Story — Learning How Doubling Makes Numbers Grow (useful for parents trying to get their children to learn something during the lockdown). 

Without intervention, the virus’s reproduction rate (ie, how many people are infected by one carrier) is very high. An R0 below one means that an infected person infects less than one other and the disease, therefore, dies out. Anything above one means that the infection rate grows. Using data from China, the Imperial College paper used a base assumption of 2.4 (range 2-2.6).

Regarding how sick people get, an assumption based on the Chinese experience and especially the city of Wuhan, 81% of those diagnosed with Covid-19 have mild cases that can be managed at home. About 14% have serious cases needing hospital treatment and 5% were critical cases of patients who suffered from respiratory failure, septic shock, and/or multiple organ dysfunction or failure. Almost half of the critical cases (2.3%) died.

It is the 19% of serious and critical cases combined with the high reproductive rate that looks to overwhelm public health systems.

Unfortunately, the above numbers are not very useful. There are wide discrepancies within China (Estimating Risk for Death from 2019 Novel Coronavirus Disease, China, January–February 2020), between China and the epicentre of the current outbreak, northern Italy, and between these and the data from the Covid-19 afflicted cruise ship, the Diamond Princess (Cruise ship outbreak helps pin down how deadly the new coronavirus is). More information here.

There are significant differences between age groups, and gender, but especially where there is any other underlying disease. The recent data from Italy is especially interesting. The fatality rate was very strongly weighted towards those with underlying health conditions. (Characteristics of Covid-19 patients dying in Italy Report based on available data on March 20th, 2020)

In Italy, the median age of those who succumbed was 80. About 41% of all those who died were aged between 80-89, with the 70-79 age group accounting for a further 35%.  The other significant detail from Italy was the presence of pre-existing conditions. Approximately 75% of the dead had two or more pre-existing conditions, 50% had three or more pre-existing conditions, in particular heart disease, diabetes and cancer. Of the five who died who were between 31 and 39 years old, each of them had serious pre-existing health conditions. But whether it is age or the underlying conditions which in Italy are associated with older people that have driven these outcomes is not yet clear.

For Ramaphosa then, the decision to implement the lockdown must have been extremely difficult. We rightly expect our governments to make evidence-based policy, but in this case, there is not sufficient evidence. The highly respected Stanford University Professor John Ioannidis (In the coronavirus pandemic, we’re making decisions without reliable data)  in a 17 March piece entitled A fiasco in the making? argues that as the coronavirus pandemic takes hold, we are making decisions without reliable data.

However, the absence of evidence is certainly not the same as the evidence of absence. The government would have taken note that the fatality rate of Covid-19 is strongly correlated to people with existing diseases of the type that are particularly prevalent in South Africa, including tuberculosis (TB), diabetes and those with weakened immune systems from the very high HIV infection rate. (Graphs that tell the story of HIV in South Africa’s provinces)

While some countries are focused on flattening the curve (How to flatten the curve of coronavirus, a mathematician explains) to give health systems breathing space, South Africa already operates beyond the curve. There is no scope for additional patients needing life-saving ventilators.

In 2015, 460,236 South Africans died. More than half (55.5%) of deaths were attributed to the group of non-communicable diseases, and communicable diseases accounted for 33.4% of deaths, while injuries were responsible for 11.1% of deaths. Other research shows that a total of 63,000 people died of tuberculosis in 2018 and two-thirds of those were HIV-positive. However, as many as 400,000 fell ill with TB in that year. TB, being a respiratory disease, could very well make sufferers particularly susceptible to Covid-19.

Even with just a 1% fatality rate and a R0 of 2.5, Covid-19 would rip through South Africa and reach the 60% infection rate predicted by Health Minister Mhkize within a year. Based on a population of around 57 million, this could amount to as many as 342,000 dead South Africans. This is just a fraction of the numbers that would need hospitalisation. Here is an epidemic calculator allowing for anyone to input their own data. None of this is even a prospect worth contemplating and this is the best case. A 4% fatality rate increases the number of dead to around 1.4 million.

In a well-argued piece, Harry Crane of Rutgers University’s Department of Statistics and Biostatistics  makes the point that Professor Ioannidis sought to treat the Covid-19 pandemic as an academic exercise and not a global crisis. Crane argues that for dynamic and complex problems like the pandemic, we cannot avoid uncertainty and we can’t delay action waiting for more evidence. Once the evidence arrives, it will be far too late to do anything about it.

This is exactly the approach taken by Professor Yaneer Bar-Yam of the New England Complexity Institute who has led a global initiative, End Coronavirus, that seeks to minimise the impact of Covid-19 by providing useful data and guidelines for action. They make the point that if everyone got tested for Covid-19, we could temporarily separate the infected from the uninfected and then help reduce the spread of the virus and return society to a semblance of normality as soon as possible.

As far as evidence is concerned, we know that lockdowns do work. (See Coronavirus: The Hammer and the Dance – Tomas Pueyo) In Wuhan, R0 moved from 3.9 before the lockdown there to around 0.32 immediately afterwards. (See: Evolving Epidemiology and Impact of Non-pharmaceutical Interventions on the Outbreak of Coronavirus Disease 2019 in Wuhan, China.) As of 19 March, there have been no new cases of coronavirus in the entire region of the province of Hubei where Wuhan is situated. Italy too, after its belated lockdown, is seeing the beginnings of a fall-off in new cases and deaths. (See Italy Coronavirus: 69,176 Cases and 6,820 Deaths)

There is, of course, the risk that after the lockdown we simply continue as before, which would see another exponential increase in new cases. Obviously, we cannot even think about another lockdown. We must insist that the lockdown is followed up with the identification of all possible infections and widespread but focused testing of all likely infections. Because there will be a relatively small number (compared to Europe and the US), this should not be that difficult to do. The lockdown gives the country some breathing space, including to develop testing protocols and to see whether any prospective drug treatments are effective.

What we can say is that each one of us is an important data point in the largest and most expensive experiment of all time. Closing the country down for 21 days to locate a few thousand people among 57 million of us might seem like overkill but we all have a direct stake, even just as data points, in making sure that the experiment runs smoothly and renders useful information. Think about this before you ask whether whatever you want to do in the 21 days might make the data less certain.

If we make the effort a success, whatever your own views of it are, there is an additional free bonus for us all. South Africa re-establishes itself at the top table of countries that can make the right decisions, a country where things can get done. We haven’t had that spirit since 1994.

As a final point, the lesson we need to learn from this pandemic is that the next one is already evolving. It might emerge tomorrow, a hundred years hence or any point in-between. Next time, we need to be much better prepared. (See: Bill Gates: The next outbreak? We’re not ready) DM