By its very nature, our argument requires a dispassionate view, one that is data-centric rather than people-centric. There are critical facts in the data that are being disregarded in the people-centric view currently guiding policy. We are careful to postulate the lockdown response hypothesis, but deliberately avoid speculation on other elements.

It is indisputable that we faced “unknown unknowns” at the outset of this pandemic. That position has changed; there is now enough data to require a policy rethink. We can draw certain conclusions and postulate certain outcomes from “known knowns”, while requesting that as more data is collected, policy reviews must be ongoing. At the outset of this pandemic, various models were relied on to formulate policy. These models (Imperial College, Oxford, IHME, SACEMA etc.) have been proven demonstrably wrong. The minutiae of how they were so wrong is well recorded and does not warrant repetition. What we do know with certainty is that Covid-19 is neither as fatal as initially predicted, nor has the R0 ever reached the levels assumed.

The spread of Covid-19 has been more rapid in some countries, but where that holds true, we also see that the infection rate reaches an inflection point and then moves into a decremental curve. It does not keep extrapolating (R0 increases for a time then starts to decrease).

How these initial models have been relied on to drive the policy of lockdown will no doubt be recorded in the annals. It seems almost redundant, but entirely necessary, to point out that since the 1950s, more than 300 contagious diseases have emerged or re-emerged in populations that had never been exposed to them.  These include such dread examples as Lassa Fever, Marburg Fever, Ebola, HIV/Aids, SARS, Zika, MERS and Swine Flu. In not one of these outbreaks was the world economy shut down.  Yet we can recall the dread and fear related to each of these events.  Furthermore, we still do not have vaccines, despite decades of research, for some.

How the current event has triggered worldwide suppression/lockdown is hard to understand.

Let us not discount the viral diseases that we are, and have been, prepared to live with. Government has not and is not considering draconian measures for viral hepatitis, HIV/Aids or influenza.

Yet, viral hepatitis caused 1.34 million deaths worldwide in 2015 and the mortality rate for hepatitis has been increasing. Approximately 325 million people, or 4.4% of the world’s population, have viral hepatitis. And 1.75 million new infections of hepatitis C alone occur each year. At the end of 2018, approximately 37.9 million people worldwide were living with HIV. In the same year, 770,000 people died from HIV-related causes and 1.7 million people were newly infected. Worldwide, flu causes about three to five million cases of severe illness and about 290,000 to 650,000 deaths each year, despite there being a vaccine for influenza.

As surely as the politicians repeat the, “We will be guided by science” mantra, they will also deflect all blame for flawed decision-making to science post-pandemic. Science will be identified as the vector for economic destruction rather than the ally in dealing with the pandemic. In some critical respects, science does have a case to answer. Every day we are regaled with how the suppression/lockdown has succeeded in “flattening the curve”. But the evidence to support this contention grows thinner and thinner. Let us not get bogged down in comparisons founded on speculation and (currently) unanswerable cross-country comparisons.

These are the facts as they stand. While these graphs are dense, we share them to represent key metrics.

Orange Line – the severity of the suppression / lockdown as defined by the Oxford University Covid Tracking Project. Purple Line – Tests per 50K Population, regardless of the various testing dynamics. Our World in Data. Light Blue Columns – Active cases per day per 1K Population (2018 World Bank Data), cases 20 days and older are deemed resolved (The shape of the curve changes marginally if you substitute Confirmed cases per day). European Centre for Disease Control Dark Blue Columns – Deaths per 1K Population (2018 World Bank Data). European Centre for Disease Control The left Y-Axis is number of people (we have scaled the ratio differently for Tests, Active and Deaths to ensure everything is graphed). The Right Y-Axis is the Severity of lockdown scale. The X-Axis is dates in the last 50 days to 11 May 2020.

Hopefully, this dispels some current myths, offered in no particular order and without speculation:

• Covid-19 has spread at different velocities and has taken hold at different dates in different populations, regardless of lockdown severity or implementation date.
• Covid-19 is not a runaway train. In every instance described, we see a familiar pattern common to viral outbreaks. Undoubtedly Russia is at the start of their outbreak, while Spain, Portugal, Italy and Belgium are over the first surge.
• The severity or earliness of the suppression/lockdown does not have a predictable flattening of the curve. Countries with aggressive R0s turn out to be Spain, Belgium and Ireland. Sweden and Japan (with their heavily criticised lowkey implementation strategies) are conspicuous by their low R0s.

If you assume suppression/lockdown to have any efficacy, then the USA is coming out of lockdown earlier than the data suggests is practical. Regarding America as one country for this purpose ignores the date differences, population density and other dynamics of initial outbreaks in various locales and is therefore deeply flawed. A cursory look at the Johns Hopkins dashboard will attest to this.

Or, view the confirmed cases per sub-region (R0):

Orange Column – 3 Day Average
Blue Column – 7 Day Average
The Y-Axis is scaled per the data for the region.

The problem with absorbing the compelling message these graphs portray is that they are snapshots of a disease at different trajectories in the cycle. By regarding 1,000 confirmed cases as Day 1 of the outbreak, we can track the trajectory of the curve for each country from the same relative start position. Indeed, there are some uncannily similar trajectories for countries – what is

profound is that the trajectory of the infection is determined by population size (a proxy for geographical size or population density).

By this expedient, every possible variable is removed from the equation (weather, health capacity, lockdown stringency, ad infinitum). The results are nothing less than astonishing:

There are 89 countries with >1K confirmed, there are 33 countries that have reached the 40-day marker described.

While there is great divergence in the active cases at Day 1, the disease spreads through the population at a trajectory that culminates in a very tight band of active cases at Day 40. Be wary of looking to the right of this date as the sample size starts to thin. But it can’t escape notice that the active rate is in decline.

Then there is the question of the demographics. Medical science was gifted a “closed population” experiment (in unfortunate circumstances) as early as February in the case of the Diamond Princess cruise ship and late March for the USS Theodore Roosevelt. We choose these two examples for the obvious disparity in their age demographics.

The Princess, having an average passenger age of 68, and the Naval craft presumably having a young healthy population.

Testing ratios in both cases were orders of magnitude higher than the numbers being recorded for countries for obvious reasons. Several things stand out – the rate of infection was not dissimilar, 18.9% and 15.1%. Most tellingly, although the population sizes were similar, 3,711 to 4,500, the mortality rate was 0.32% (12 deaths) compared to 0.02% (one death) on the Naval craft and hospitalisation rates of 20% from the Diamond Princess, compared to 1,03% from the Roosevelt.

It was reported early on that 80% of patients  on the Diamond Princess showed mild or no signs of the disease. This was an early and forceful indication that the elderly were at significantly greater risk from the virus. Further, that even in the elderly vulnerable age group, severe illness occurs in a minority.

Subsequently there have been many confirmations, not limited to:

Italy as at 4 May 2020. Of the total deaths, 25,635 (95.3%) were 60 years or older.

USA as at 1 May 2020. Of the total deaths, 34,343 (92.05%) were 55 years or older.

It is blindingly obvious that the people most at risk from Covid-19 are not in the population group that drives the economy, and even less so in the learning group. Even with the limited data available in South Africa, old people with comorbidities are clearly the most affected in our population.

Pitting health and economic costs against each other, as if there is a disjoint between the two, is disingenuous at best and negligent of history at worst.

The simultaneous destruction of both the demand and supply side of the economy for any period is a high-risk gamble; for an extended period the risk extrapolates (this is critical, it is not linear).

While we claim no economic background, there can be little if any doubt that the decision to lock down (for 21 days) and subsequent hesitation in lifting the lockdown will have disastrous effects on our country.

The reasoning currently being offered by the government in defence of their position, and as justification for the overreach of their powers, does not stand up to even the laxest scrutiny. By the government’s own admission, we have been tipped over an abyss.

SARS predicted a R285-billion shortfall in tax revenue. Treasury has declared that in the best-case scenario (where the economy bounces back, an extraordinary tenet in a government-strangled economy) we will have lost three million jobs – seven million in a worst-case scenario (the one we currently choose to be in). For perspective, the 2009 financial crisis “only” cost less than one million jobs, and even so, the consequences were severe.

The economy must reopen, not in the staggered manner currently expounded. It must reopen as rapidly as possible. Decision-making must be restored to the system, not the government. Government interventions of this magnitude have a chequered history (at best). Let society establish the new norm. Let companies (directors and employees) resolve the challenges of physical distancing, hygiene, testing etc. Now, more than ever, the sense of a need for cooperation between science, industry, government, the populace and medicine is at an unprecedented peak: let us leverage that sentiment for positive outcomes.

A vaccine may or may not be developed in the next year. Seroprevalence testing will give us a better understanding of the characteristics of this particular virus regarding infectivity and immunity. At the end of the day, Covid-19 is not going away. We will have many more infections and more deaths. We will probably face a surge in winter (as with many respiratory viral infections).

While the loss of a single life to Covid-19 is tragic, actively increasing the concomitantly caused loss of life and impoverishing millions is avoidable. There can be little doubt, if any, that the loss of life caused by the lockdown will far exceed the loss of life from the virus. This has been the case elsewhere historically; for example, the measles outbreak (a direct consequence of an imbalanced health system focused predominantly on a single disease), following the west African Ebola outbreak, claimed double the number of lives attributed to Ebola. Jolting the South African unemployment rate from 27% to over 50%, has death writ large.

What we need now is decisive and reasoned action. It might not be popular in the short term, but this is not the time for political point-scoring. Leadership is not just about making the right decisions; it is getting civil society to embrace those decisions. While evidence for lockdown is evaporating, so too is support for that same strategy waning fast. We can no longer be silent and compliant. We need bold and decisive leadership; we need to demonstrate flexibility in the light of emerging evidence. We need to continue sanitizing. We need to protect the elderly and those with co-morbidity, but not at any price. We need to continue to upscale testing and health service preparedness. We need to abandon strategies doing more harm than good, and we need to rescue our economy and the 2020 academic calendar for all levels of education, acknowledging that there will be a price to pay. We need to act now.

The South African government’s current stance on suppression/lockdown should be brought sharply into focus with this graph (expanded from above with the Y-Axis in proportion). Lockdown has not “flattened the curve”, despite being in place for six weeks

Orange Line – the severity of the suppression / lockdown as defined by the Oxford University Covid Tracking Project. Purple Line – Tests per 50K Population, regardless of the various testing dynamics. Our World in Data. Light Blue Columns – Active cases per day per 1K Population (2018 World Bank Data), cases 20 days and older are deemed resolved (The shape of the curve changes marginally if you substitute Confirmed cases per day). European Centre for Disease Control Dark Blue Columns – Deaths per 1K Population (2018 World Bank Data). European Centre for Disease Control The left Y-Axis is number of people (we have scaled the ratio differently for Tests, Active and Deaths to ensure everything is graphed). The Right Y-Axis is the Severity of lockdown scale. The X-Axis is dated in the last 50 days to 11 May 2020.

The infection rate has increased substantially and in fact accelerated after suppression/lockdown. And consider that South African testing only up-ticked significantly weeks after lockdown had been decided on as a strategy.

Is the above graph sufficient to justify the current lockdown, or, is any graph, from any country in the world, sufficient to satisfy our government’s position? We say an emphatic no.

You are welcome to test our hypotheses against any country in the world. The dashboard is available at coronavirus.africa.com. DM

Professor John Gear was previously head of Public Health at the University of the Witwatersrand and Founding Director of the University of Witwatersrand’s Rural Facility and, currently in his retirement, Medical Director of the Tshemba Foundation which recruits volunteer health professionals to rural northern Mpumalanga

Ian McGorian, “Datanaut” provided the mathematical and statistical expertise underpinning the article.