摘要：As the world passes 3 million deaths from covid-19, New Scientist explores the pandemic's trajectory and discovers many reasons for hope that the worst will soon be over

COVID-19 has now claimed the lives of a reported 3 million people around the world and the true figure is probably far higher. Globally, the number of new cases per week has nearly doubled over the past two months, as new variants cause a surge in cases in many countries.

But as we mark this sobering milestone and explore the shape of things to come, there is much reason for hope, too. Last week, the world produced its one-billionth dose of covid-19 vaccine, and the few countries that have already managed to vaccinate a large proportion of their population are seeing the benefits.

No one can say for sure what will happen next, but the hope is that the worst will soon be over. “It’s in our hands,” says Andrew Noymer at the University of California, Irvine. “It depends on vaccination. And variant roulette.”

Many countries may not be able to get hold of enough vaccines to inoculate most of their population this year, though. “The rich countries have been very clear that they will continue to ensure their needs are met first,” says Andrea Taylor at Duke University in North Carolina. “There are distinct haves and have-nots.”

The beacon of hope for the world is Israel, which was the first to vaccinate the majority of its population and where case numbers are falling sharply. “In Israel, vaccination seems to have worked a treat,” says Noymer. “They are leading the way.” However, with lockdown measures still to be completely eased there, it isn’t clear if the country has managed to pass the herd immunity threshold beyond which the virus can’t cause major outbreaks.

“That is the big question we don’t know the answer to yet, whether any or all of these vaccines are good enough to get us past the herd immunity threshold,” says Mark Woolhouse at the University of Edinburgh, UK.

After Israel, Bhutan, the UK, Chile and the US are the countries that have vaccinated the highest percentage of people – but not enough to get close to herd immunity. In some US states, such as Michigan, case numbers are surging because of the B.1.1.7 variant first seen in the UK.

Case numbers have also shot up in Chile even as the vaccine is being rolled out. While this has led to some alarming suggestions that vaccination isn’t working, it isn’t surprising.

At least 70 per cent of a population, and possibly as much as 90 per cent, needs to be immune – in the sense of not passing on the virus even if infected – to get over the herd immunity threshold. In Chile, less than half the population had been vaccinated when cases surged, and most of those people had received just one dose. The main vaccine used in Chile so far is also Sinovac’s CoronaVac, which is 67 per cent effective at preventing symptomatic infections according to a study in Chile, lower than many other vaccines.

In the UK, the fall in case numbers is mostly due to a lockdown fully imposed in January, and there remains a risk of a resurgence if restrictions are lifted too fast. Nonetheless, the benefits of vaccines in reducing the number of cases and deaths are becoming increasingly clear, with a government analysis suggesting vaccination had saved 10,000 lives by the end of March.

In fact, most vaccines are proving to be far more effective than researchers dared to hope last year. In the UK at least, vaccine uptake is also much higher than polls suggested it would be, with 95 per cent of people offered a vaccine accepting it.

Globally, though, just 6 per cent of people have received at least one dose so far. In Africa, itis less than 1 per cent.

In many countries, case numbers and deaths are once again rising fast. India is reporting the highest numbers – more than 200,000 cases per day and climbing – and these are likely to greatly underestimate the true total. Relative to the population, Turkey, Brazil, Italy, Germany, Canada and the US have some of the highest case rates.

While there are many factors involved, new variants are playing a big part. The B.1.1.7 variant has rapidly come to dominate in Europe and North America. It now seems to cause about half of all cases worldwide, but we don’t have a complete picture because many countries do little virus sequencing.

“The variant from the UK now appears to cause around half of all covid-19 cases worldwide”

The B.1.351 variant that evolved in South Africa and the P.1 variant causing the outbreak in Brazil are spreading globally too, but not to the same extent as B.1.1.7. They are each responsible for about 5 per cent of global cases.

Despite this, some countries, such as South Korea, Japan and Thailand, continue to keep case numbers low. South Korea, for instance, is battling a small surge in cases, and so far just 1.3 million of its 60 million inhabitants have had at least one vaccine dose. Yet it has reported just 110,000 cases and 1782 deaths since the pandemic started. By contrast, the UK, with a similar population, has reported more than 4 million cases and 150,000 deaths – there have been fewer cases in South Korea than deaths in the UK.

Meanwhile, the disease remains effectively eliminated in a few countries, such as Australia and New Zealand.

What happens next?

What we all want to know, of course, is what is next. The short answer is that no one knows. Many predictions have already been wrong, says Oliver Pybus at the University of Oxford. “None of us have a crystal ball.”

Broadly speaking, the best-case scenario is that countries manage to dampen down the global resurgence and effectively eliminate the virus locally as vaccination proceeds.

The nightmare scenario would be a variant that is both more transmissible than B.1.1.7 and also completely evades vaccines. This could cause an even bigger wave of cases and deaths than any so far. Fortunately, there is reasonable confidence that the vaccines will continue to protect against severe disease even if they fail to stop transmission, says Pybus.

Between these extremes there are many possibilities. The key factors include: how fast we can roll out vaccination worldwide, how long its protection lasts, whether dangerous new variants evolve and how people behave.

How fast can we vaccinate the world?

Vaccine production is ramping up fast. According to science analytics company Airfinity, the billionth vaccine dose was produced on 12 April and the second billionth will be produced by 27 May.

Taylor’s team has added up all projected production figures from vaccine manufacturers, and the total equals 12 billion by the end of 2021. That is enough for two doses for all adults in the world, which would be a wonderful achievement. But Taylor doesn’t expect them to get close.

“Twelve billion is very optimistic. It is very unlikely to be produced in 2021,” she says. “The manufacturing process is incredibly complex, and there are just a million places where it can get held up.”

Manufacturers overestimated production in 2020, according to Airfinity, producing 30 million doses rather than the projected 800 million. Most haven’t met their targets for the first quarter of 2021 either, says Taylor.

She thinks the most likely scenario is that rich countries will get vaccine-makers to stop making existing vaccines and switch to producing booster shots against new variants to be given to people who have already been inoculated.

“That removes capacity that could be serving lower and middle-income markets,” she says. “It just prolongs the pandemic, which puts everyone in a worse place.”

How long will vaccine protection last?

One big worry is that the protection conferred by vaccination and previous infections will fade rapidly, which could lead to more waves of infection. However, the expectation is that prior immunity will protect against severe illness and death even if it doesn’t prevent infection, so waves should be far less deadly.

We don’t yet know how long vaccine protection lasts, but it is definitely longer than six months for the Pfizer/BioNTech and Moderna vaccines at least. Many researchers are optimistic that protection will last much longer, but we are going to have to wait to find out for sure.

Will the virus evolve to evade vaccines?

None of the vaccines provides 100 per cent protection and some new variants such as B.1.351 and P.1 can evade vaccines to some extent. Most recently, a small study in Israel found that people infected a week or more after their second dose were eight times more likely to have B.1.351 than would be expected given its prevalence.

The good news, however, is that the incidence of B.1.351 in Israel remains low. It doesn’t seem to evade the Pfizer vaccine used in Israel to the extent that the R number rises above 1, that is, that B.1.351 can keep spreading.

Yet not all vaccines are as effective as the Pfizer one. And the worry remains that variants could evolve further to become even better at evading vaccines. “We must expect that, at some point, the vaccines will fail to protect against transmission because that is pretty much what happens with similar vaccines [against similar viruses],” says Pybus.

However, this might take years to happen and, even if it does, the variants may not make those with prior immunity seriously ill. It is much harder for viruses to dodge our T-cell response, which helps prevent serious disease, than for them to dodge antibodies, which prevent infection. In other words, SARS-CoV-2 could become like existing human coronaviruses, causing only mild illnesses.

“It’s not a doom and gloom story,” says Pybus. “I think it’s very realistic that the vaccines are going to have a massive effect. I also think it’s very realistic that they will need to be updated.”

How often they will have to be updated depends on the number of cases, he says. The more viruses there are replicating, the faster they will evolve. It is especially dangerous to have lots of cases in populations that have been partially vaccinated. That is why it is in the interest of rich countries to help the rest of the world get vaccinated as soon as possible.

How we behave

It might turn out that vaccination alone isn’t enough to get us past the herd immunity threshold. “If it doesn’t miss by much, getting the R number down even a little bit will get us across the threshold,” says Woolhouse. “I can easily see us being in that space, where the vaccines by themselves aren’t quite enough, but with some additional measures we don’t get large epidemics.”

That might mean maintaining measures such as physical distancing and wearing masks in public spaces. The danger is that many people will abandon all precautions even before vaccination is anywhere near complete. “We need to be careful still,” says Noymer.

“We can see the path ahead,” says Pybus. “But there are still miles to walk.”

What is the true death toll?

The real number of global coronavirus deaths is likely to be much higher than the official count of 3 million.

“My guess is probably double that,” says Andrew Noymer, an epidemiologist at the University of California, Irvine. It could be even more, he says, but is almost certainly under 40 million.

The reason we don’t have a clear answer is that many countries do little testing, so they don’t know which deaths are due to covid-19.

In some places, especially in rural areas, many deaths aren’t recorded. Then there are countries, such as the Central African Republic, from which we have no data at all. In others, such as Russia and China, official figures are questionable. “They have good statistics, I believe, but they are not being transparent about them,” says Noymer.

In a number of countries, however, it is clear that the number of extra deaths that have happened during the pandemic is much higher than can be accounted for by reported covid-19 deaths, suggesting massive undercounting.

Noymer thinks researchers will be arguing about the global covid-19 death toll for many decades, just as they are still debating the toll of the 1918 flu pandemic.

“We don’t know how many people died of the 1918 flu, and we’ve had 100 years to sort that out,” he says. “Once you have a situation in which there is no consensus, then there never will be, in some sense. We just don’t have a lot of data and everything to do with covid has become politicised.”