A health worker screens a resident at Diepsloot Covid-19 screening and testing site at Diepsloot Sarafina Park on May 08, 2020 in Johannesburg, South Africa. It is reported that more than 12 000 people have been screened and over 1000 people tested in Diepsloot. The Premier urged the people of Diepsloot to continue practicing safety measures including social distancing and wearing cloth masks when leaving home. (Photo by Sharon Seretlo/Gallo Images via Getty Images)
With the latest adaptations to the South African government’s level one Covid-19 restrictions and the growing feeling that the Omicron variant is a game changer, it is worth reviewing where we are.
Do we carry on as before, or, like the Swedes. in the early days, take the whole thing very lightly? (At great cost: they had a much higher death rate than their Nordic neighbours.)
First, I look at some of the earlier science and how that panned out, followed by an overview of applicable virology and epidemiology. Then I look at counter-measures that have been applied since the early days and evaluate whether they still make sense.
A review like this is useful because of the errors made in the past when the pandemic hit a phase change.
A key early claim was that Covid-19 does not mutate as fast as the flu virus, so an effective vaccine would stop it. Estimates I have seen are that it mutates two to four times slower than flu. But it also spreads a lot faster and the probability of a mutation has to be multiplied by the number of virus replications to get an accurate picture, which has turned out to be far higher than for flu.
An oddity of the way Covid-19 evolution has progressed is that, instead of variations from a single source, variants have emerged unrelated to other variants of concern. Omicron, for example, according to genomic evidence, is not derived from Delta. That has also confounded the view that a relatively simple strategy would thwart it.
Despite these setbacks, vaccines have remained effective against serious illness and death. To understand this, it is useful to know how the immune system works. This is a big, complex topic, so I focus on two aspects: antibodies and T cells.
The role of antibodies
An antibody is a Y-shaped protein that binds to another protein. The “V” part of the “Y” has a pattern specific to the protein that the antibody targets. Think of it like a lock and key. When the body detects a foreign protein, it produces antibodies specific to that protein. Once the antibody binds to the protein, it interferes with its function. A virus does not only have one protein and there can also be antibodies that target different parts of a given protein. As a virus mutates, parts of the protein that an antibody targets may vary enough that the antibody is no longer effective.
In the case of the SARS-CoV-2 virus that causes Covid-19, antibodies tested to detect illness target its nucleocapsid (N) protein, which plays an important role in the lifecycle of the virus. Since the spike protein is most strongly implicated in disease, it is the target of vaccines.
Antibodies to the N protein are more sensitive than antibodies to the spike protein for early detection of infection. However, the vaccines are not based on the SARS-Cov-2 virus so antibodies to the spike protein (caused by a vaccine) won’t be detected on a standard Covid-19 antibody test.
Why do vaccines still work — if not as well — against new variants?
The role of T cells
Even with big changes in the spike protein, some antibodies will still be effective — they may target parts of the protein that didn’t change. Also, there is another layer of defence, T cells. T cells attack infected cells. T cells are trained to attack a particular invader and are less specific than antibodies and, hence, less likely to be confounded by new variants.
Another factor in longer-term immunity is that antibody production subsides over time, whereas the T cell response can last for more than a decade (T cells specific to Severe Acute Respiratory Syndrome disease, caused by another coronavirus, SARS-CoV, were still effective 17 years after infection).
This is a very complex subject, but the simple message is: antibodies are the first line of defence. If they work well, they prevent the infection from taking hold; if they don’t, T cells can still stop disease from progressing. This is one reason that despite Omicron spreading like wildfire, the rate of serious illness and death was much lower than before — previous infection and vaccination have trained T-cells well enough to react even if antibodies were not up to the task.
Each new variant may result in different comparative effectiveness of vaccines versus infection-derived immunity, but vaccines have two key advantages. They don’t carry a significant risk of serious illness and death, and any immunity you gain is not at the cost of infecting others. A vaccine that induces the wrong mix of antibodies for a specific variant may still reduce contagiousness by reducing the time you are ill, but the biggest win over disease-induced immunity is reducing serious illness and death.
Vaccine mandates
So are vaccine mandates justified? On the whole, yes — although we could consider proof of previous infection as an alternative. However, in a well-informed society without much influence of fake news, why should any form of coercion be necessary? Ironically those people most opposed to vaccine mandates are at the forefront of spreading anti-vax misinformation.
Epidemiology remains one of the most misunderstood topics. When a novel disease breaks out, there are three factors in determining R0, or the basic reproduction number, which is the average number of people an infected person passes the virus on to: how infectious the virus is, how long someone is infectious and degree of social mixing.
Once the disease spreads, the current R number becomes lower for various reasons. Those who now have resistance do not become reinfected. Interventions to slow the spread can also — in the now infamous phrase — flatten the curve. The curve most often spoken of represents the number of active cases, an important metric for understanding pressure on health systems.
With no interventions, the curve goes up exponentially until an appreciable fraction of the population has resistance, then flattens out and drops. The result is a bell-shaped curve (technically, a logistic curve — similar to, but not the same as a normal distribution curve).
It’s important to distinguish between natural limiting factors like growing community resistance, and temporary counter-measures like distancing and masking. The latter fake the effect of a less contagious disease but if you relax them, R bounces back up.
Different types of immunity
What makes the whole picture more complex is the mix of different types of immunity — infection derived from more than one variant and various vaccines.
Nonetheless the basics stay the same — distancing et cetera fakes the effect of a less contagious disease and, therefore, slows and lowers the peak.
Masking is a much-misunderstood topic. A figure as revered for straight-talking truth-telling as Professor Shabir Madhi talks about masking as if the sole purpose is protection against infection. It is not: a significant factor is containing the spread (source control). Any air escaping from an infected person contains infected droplets. A mask doesn’t have to catch them all; all it needs to do is stop them from spewing out widely. Masking doesn’t have to be perfect. All it needs to do is reduce R. A similar thing applies to hand hygiene: infecting yourself through your hands may be a rare occurrence, but applying all these measures, including distancing and ventilation, adds up to a cocktail of measures, each imperfect on its own but adding to reducing R.
Understanding that these things work together should be clearly explained; I am bemused to see people talking loudly in each other’s faces unmasked, while handing around the sanitiser. You don’t breathe through your hands!
Has any of this changed since the start? R today is much lower — despite the fact that the original estimate for R0 was 2.5 and that for Omicron is about 10. Firstly, much of the population has immunity in some form even if the vaccine roll-out is disappointing — up to 70% could already have been infected. Secondly, we are maintaining measures like distancing and masking, if imperfectly.
If we dropped all these measures, we would have another spike in infections, though probably not as bad as before. So why not drop them all now — including masking, as proposed by Madhi?
The Czech experience
The Czechs were the poster child for informal masking. In March 2020, a social movement promoting masking took off and everyone with a sewing machine, it seemed, was making masks. Their pandemic was one of the mildest in the world, with few deaths. In June 2020, restrictions — including masking — were lifted. It took a while for the virus to make a comeback and it did with a vengeance, giving the Czechs some of the worst numbers in Europe. Despite this, masking was never reintroduced at the same level as before.
The risk with lifting restrictions that are easy to implement and not especially irksome — yet which require widespread individual compliance — is that it is hard to bring them back.
So which can we safely lift now?
Temperature screening is mostly pointless. Someone with a high fever should know they are ill and many thermometers are inaccurate or poorly used. At one location, my temperature was regularly recorded as 26°C. If that was correct, I would be dead or a reptile.
Hand sanitising could be done less often — at entry to every shop, it is overkill. However, the message of hand hygiene is worth maintaining.
Distancing could be emphasised less, with ventilation a stronger factor. Omicron is so contagious that even a distance of 1.5m may be too little, but strong ventilation can remove the virus before it reaches others.
There is no need to close a workplace and fumigate it for every case detected. The evidence for surface contagion is weak and, if hand hygiene is maintained, is unlikely to be a big factor.
The Swedish experience
Back to Sweden. Comparing societies requires a close understanding of R0. 2.5 for the original strain was a central estimate. In societies with more mixing, it was higher. A country like Sweden is not noted for crowded public spaces. This is why its pandemic should be compared with similar societies.
South Africa is particularly difficult to compare with other countries because we have a unique mix of rich and poor, top science and incompetent public health. We were the first to report Omicron (previously discovered in Botswana) as a concern, yet our testing catches only 10% of infections and excess deaths are about triple official deaths.
One thing is for sure: we aren’t Sweden. Much of our population relies on crammed taxis for transport and lives in crowded circumstances. So, if we are going to be as relaxed as Sweden was back then, we need to be sure that this is justified. I am not sure that we are quite there yet. Despite being “milder”, Omicron can still kill and its rapid spread in some of the countries stressed their health systems.
And this may not be the last variant of concern. The light may be visible at the end of the tunnel but we aren’t there yet. And we could still go backwards.
Asymptomatic and pre-symptomatic
The government’s latest relaxation of rules is based on accepting that testing and contact tracing in the public sector is ineffectual. To say that people who tested positive who are asymptomatic shouldn’t isolate because 90% of infections remain undetected is to admit defeat. Asymptomatic transmission is no longer a thing? Tell the virus.
Asymptomatic doesn’t mean you have no ill effects. It just means that you don’t have any you know about. An example: silent hypoxia. Poor lung function makes you feel ill because carbon dioxide accumulates in the blood. On the other hand, if hypoxia is caused because the blood is doing a poor job of taking up oxygen, you may feel fine, even euphoric (a trait of the early stage of altitude sickness). You could be suffering undetected organ damage.
Asymptomatic should also not be confused with pre-symptomatic and this is where the new policy is really confused. Viral shedding can peak as much as two to three days before symptoms show. If symptoms do show and you have in the meantime not been isolating because you believe you are asymptomatic, you will have passed the peak of contagiousness before you isolate. What kind of sense does that make?
The trick now is to be adaptive — to start getting the message out that we need to fine-tune our strategy to circumstances. The government’s one-size-fits-all approach is broken. We need to acknowledge that and be more flexible. But we also should not throw away hard-won lessons. There is no magic to this — just a very hard problem that would be most easily solved if all of society worked together and was well informed.