When the news of the novel coronavirus spread around the world, it was not known how infectious or lethal the disease might be. Faced with the prospect of a potentially devastating spread, an appropriate application of the precautionary principle was required. A few countries—notably including Taiwan—grabbed the opportunity to contain the spread by banning travel to and from Wuhan. But most didn’t.
To their eternal shame, the Chinese Communist Party banned travel between Wuhan and other Chinese provinces but let international flights continue. The disease proliferated. Quite quickly, harrowing stories began emerging from northern Italy and Madrid—of overwhelmed hospitals and dying health workers. The WHO estimated a dramatic case fatality rate of 3.4%. Epidemiologists, plugging similar estimates into models, projected tens of millions of deaths worldwide.
For a few weeks, some hospitals in New York City looked just like northern Italy. This was covered graphically and at fever pitch in mainstream and social media. With their leaders under pressure to do something, most countries responded by assuming that COVID-19 would wreak havoc for them too and commenced with lockdowns of varying severity.
Lockdowns are a subset of social distancing measures that aim to “flatten the curve”, so that preparatory measures can be taken and the rate at which hospital cases enter the health system can be kept manageable. Provided that this rate is managed, flattening the curve in an epidemiological model does not change the number of viral deaths it predicts—it merely spreads them over time. Curve-flattening only saves lives to the extent that we avoid having some treatable cases of COVID-19 and other afflictions end in death. Because ICU death rates have been extremely high (up to 80%), this beneficial effect on the case fatality rate may not be as material as originally hoped.
Though this salutary effect counts to the credit of lockdowns, there are also debits. Whether through direct impacts, such as prohibition of elective surgeries or cancer screening, or through indirect economic impacts, such as poorer nutrition or higher rates of suicide and domestic abuse, these debits are substantial and cannot be ignored. Policymakers face the unenviable task of striking a balance between these credits and debits when determining the severity and duration of social distancing policies. All of them risk committing accidental homicide on a ghastly scale if they ignore any element of a terribly complex situation.
Amid this drama, many scientists began conducting studies that would permit analysis of the infectiousness and lethality of the disease. While such on-the-fly work is always prone to error, a dozen or more studies have emerged. They have their flaws and limitations, but, strikingly, they all deliver a remarkably consistent message. The message is that COVID-19 is nowhere near as lethal as was initially feared.
The coronavirus presents negligible risk to young people and to healthy middle-aged people; modest risk even to healthy old people; and material risk to unhealthy old people. The question of who is healthy in this context seems to revolve mainly around whether you have type 2 diabetes, high blood pressure, chronic obstructive pulmonary disease, cardiovascular diseases or chronic kidney disease. In European countries that make their data publicly available, about 90% of deaths occur in people over the age of 70. About 90% of those deaths involve one or more of these conditions. Even among very old people, many who are exposed to the virus appear not to get infected and the majority who are infected don’t develop symptoms. Of those unfortunate enough to develop symptoms, around half get ill enough to be hospitalized and about 5% die. Multiplication of those factors makes for a low mortality rate even among those most at risk.
Coronavirus in South Africa
Quite notably, neither TB nor HIV have been observed as material comorbidities. Countries that have high incidence rates of these two conditions do not appear to be doing any worse than countries that don’t. Furthermore, South Africa does not have many old people. Just 3.2% of our population is over 70, compared to 17.2% in Italy. This is good news for our country—although the bad news is that we have above-average obesity and diabetes rates, and both are strongly associated with the comorbidities listed earlier.
On balance, though, we are compelled to revise our models of this disease in South Africa, and therefore our policy decisions. It does not seem likely that the total deaths that will arise from COVID-19 this year will exceed 20,000, with or without lockdown. This needs to be evaluated in light of the 600,000 deaths that occur in South Africa each year, 130,000 of which occur among the over 70s. We must also consider that some substantial portion of these deaths would have occurred anyway.
The consequences of lockdown
The economic contraction caused by lockdown hit an economy already in recession and has been estimated to be in the range of 8 to 20%, depending on how long its duration is. Some of this rapidly mounting damage—perhaps 5 to 10%—will be as good as permanent. No amount of stimulus will solve it. A million formal sector jobs appear to have been lost and half as many again in the informal sector, plunging several million South Africans into immediate poverty.
Many who have raised this point have been accused of trading lives for money. But this accusation is desperately naïve. The effects of reduced means in terms of death and sickness are well known—so well that they are even modelled by insurance companies when they set their premiums. They are particularly grim for middle income countries such as our own. When incomes fall, deaths of despair (from suicide, domestic abuse, murder, substance abuse and the like) rise. Hunger causes long-term illnesses, retarding development and condemning children to lives of misery and dependence. Outbreaks of cholera, typhoid and dysentery lurk in shadows, waiting to pounce as sanitation and immune systems decline. Vaccination programme slippage threatens eruptions of measles and other communicable diseases in future years.
Using such insurance models as a guide, it would not be unreasonable to expect that several million South Africans will be so profoundly affected by this lockdown as to have their lives shortened by, say, five years. And those shortened lives will be far more miserable. This insight was heavily promoted recently by those who bemoaned austerity responses to the global financial crisis. It is a disgrace that those same voices are silent now.
Actuarial models indicate the life-years for the entire population lost to lockdown will be at least two orders of magnitude greater than the life-years lost to coronavirus. We’re talking about tens of millions of years in total, versus, at worst, tens of thousands of years.
Early restrictions on travel and large gatherings were morally defensible as an application of the precautionary principle. For distant island nations that had plenty of warning, eradication was plausible, but for others it has never been an option. While developed nations enjoy great safety nets and can absorb the impact of lockdowns, developing countries clearly needed to be more careful when contemplating harsh measures. Their large middle classes are a paycheque or two away from ruin.
Had the prospect that coronavirus could kill tens of millions of people been increasingly confirmed, lockdowns may have been justified. That possibility no longer exists. It now seems unlikely that more than a million will die worldwide. The humanitarian crisis provoked by lockdown, however, is a matter of sheer certainty, threatening to rip through the very fabric of our society. At this point, continuing lockdown in countries such as ours—in any form whatsoever, and whether “risk-based” or not—is indefensible.
We should wear masks, we should cocoon our old and infirm, and we should be cautious in our interactions. We should not incarcerate the masses in their homes, limiting their hard-earned freedoms and decimating the economic system that furnishes their livelihoods. We should be doing everything we possibly can to allow businesses to operate. If we don’t, the outcome will be dire.
Nick Hudson – Fellow Actuarial Society of South Africa (FASSA)
Peter Castleden – Fellow Actuarial Society of South Africa (FASSA)
PANDA (Pandemic ~ Data Analysis) is a multidisciplinary initiative seeking to inform policy choice in the face of COVID-19. Panda’s technical team brings to bear knowledge from the fields of actuarial mathematics, economics and medicine and is continually recruiting.