VIENNA — “I have never seen an emergency room so crowded with very, very sick patients,” recalls Annalisa Malara, a doctor at Codogno Hospital in Lombardy, Italy. “We were literally overwhelmed by the number.”
That was late February, when Malara diagnosed Italy’s first case of locally transmitted Covid-19. An emotionally wrenching marathon of hospital shifts followed. Malara felt like she was constantly scrambling — to get enough oxygen to keep patients alive, to arrange transfers to other hospitals, to try and sleep so she could keep going. “We had to watch patients die,” she said. “It’s something that I think I can’t forget — I will never forget.”
More recently, Codogno Hospital has been Covid-19-free. But with case counts rising across the country and the continent again, Malara worries about a return to tragedy. Last week, she spent hours on the phone with colleagues to check on the status of their intensive care units. The situation was stable — though no one was sure how long that would last. “Everyone is very scared,” she confessed.
Only six months after Italy’s coronavirus crisis became a warning to the West about how quickly the virus could strain even the best-resourced health systems in the world, the World Health Organization warned Thursday of a “very serious situation” unfolding again across the agency’s European region, as weekly cases surpassed those reported during the first peak of the pandemic in March. This “should serve as a wake-up call for all of us,” Dr. Hans Kluge, the WHO regional director for Europe, said.
At the country level, the situation is no more reassuring. Italian officials are once again reporting around 1,500 new infections each day. That’s not yet anywhere near the last peak of 6,500 — but it’s a significant rise from 200 in early July. France and Spain, currently the hardest-hit countries in the region, are tracking an onslaught of cases even worse than their springtime peaks. On September 7, Spain became the first European country to count half a million total cases — more than 100,000 of them diagnosed in the two weeks prior. Days later, France recorded a jump of 10,000 new cases in one day. In Austria, between late April and mid-June, cases stayed below 100 for weeks, then gradually rose, with 768 reported on September 16.
Even Germany, an oft-cited example of coronavirus response excellence in Europe, has slowly seen its daily case count edge up, with nearly 2,000 infections — a doubling from August 1.
Yet one can almost forget there’s a pandemic going on in many parts of the continent.
Here in Vienna, where Covid-19 infections are rising so fast Germany just declared the city a high-risk hot spot, restaurants in my neighborhood are full of maskless people practically sitting elbow-to-elbow, kids are back in school and on the playgrounds, and opera season has reopened.
This disjuncture — between the rapidly rising case counts and relatively relaxed social life — has left some confused, and others complacent. When I showed up at a small garden brunch on Saturday, the hosts said they’d wondered whether they should cancel — worrying what their neighbors would think of the gathering — after they heard the news that morning that Austria had just recorded more than 800 new daily infections. Meanwhile, a 20-something university student told me that the coronavirus now feels like old news to her peers.
Part of the confusion has to do with the fact that the current phase in Europe — despite how bad it looks according to case numbers alone — has a different dynamic from the first. And that’s not only because Covid-19-related hospitalizations and deaths aren’t rising nearly as quickly as they did in the spring. This new dynamic is hard to see when you consider case counts or even hospitalizations alone.
“Even more than in March and April, we need to use [all the data] to paint a picture that’s complex and balanced,” said Edouard Mathieu, the Paris-based data manager of Oxford University’s Our World in Data project. Indeed, Europe’s latest coronavirus chapter is a much more nuanced story than the previous one — but it ultimately ends in the same place: a looming and very real risk of exponentially growing cases, leading to thousands of unnecessary deaths and the threat of new lockdowns.
Why cases alone can’t tell us about Europe’s new Covid-19 surge
There are a few important caveats to keep in mind when examining Europe’s surging coronavirus case count. The biggest one: You can’t “take the numbers today and place them on top of the earlier curve and assume it’s the same thing,” Flavia Riccardo, a researcher at the Italian National Institute of Health, told Vox.
Official case numbers are always an artifact of how testing is being done — and how testing is done has changed dramatically over the course of the pandemic. “Most cases we were seeing at the beginning were clearly symptomatic, because the indication was to test only symptomatic people,” Riccardo said. “Generally people were asked to stay home until they had more severe symptoms. That’s the opposite of what’s happening now.”
That broadening of the testing criteria for the coronavirus, along with an increased capacity to swab, has meant many more tests are being done, and more cases documented. In mid-March, German officials performed roughly 20,000 tests per day, according to Our World in Data. Now the number is a staggering 150,000. By the end of May, France and Spain were doing 37,000 and 44,000 daily tests, respectively; France now does 144,000 tests per day and Spain, 89,000.
This suggests two things: During the first wave of the pandemic, health officials only captured a fraction of the coronavirus infections in the population, so the real peak in the spring was much higher than the official graphs suggested. “We probably measured something like less than 10 percent of what happened,” Mathieu said. Second, the recent surge looks relatively large compared to the spring spike — but in reality, it’s probably smaller.
Still, the expansion in testing alone can’t fully explain the current case uptick, at least not in every country. That’s where another metric — test per case — becomes useful.
The test-per-case ratio tracks exactly what it sounds like: the number of tests being done divided by the confirmed cases. When the number drops too low it means an epidemic is likely out of control, since officials can no longer keep up with the demand for testing and see where new pockets of disease are spreading.
That’s what’s happening now in Europe — most notably Spain, Italy, and France — where cases are rising faster than can be explained by the increase in testing alone.
For example, back in June, Spanish officials did 130 tests for every confirmed case. As of September 12, that number had dropped to only nine tests per case. In Austria, the number of tests per confirmed case has fallen from 250 in early June to 20. A similar decline in the UK is proving to be worrisome. There, reports of an explosion in testing wait times and delays in people getting notified of positive results have sparked rumors of another lockdown.
Hospitalizations and deaths in Europe are rising — just at a slower rate than before
But there’s another aspect to the data we have to consider in order to understand the trends in Europe: how quickly infections, hospitalizations, and deaths are rising. Today, they are generally increasing at a much slower rate than in the spring (you can see that more clearly when you plot the data on a logarithmic scale). While this is good news, the trends are still worrisome.
Let’s focus on hospitalizations for a moment to understand why. Hospitalizations are often an intermediate step between rising cases and increased mortality. They can also signal the extent to which a health care system is strained and, when they start rising, provide a warning that more resources need to be marshaled quickly.
Earlier in the pandemic, there was a two-week lag between an uptick in cases and a rise in hospitalizations, said Maria DeJoseph Van Kerkhove, the WHO’s Covid-19 technical lead. And when hospitals became crowded with Covid-19 patients, that was followed by an increase in Covid-19 deaths.
Yet, for most of the summer, the surging Covid-19 caseload in European countries wasn’t accompanied by that growth in hospitalizations or deaths — a trend many attributed to the shift in transmission from mainly older to younger groups.
“Across Europe, there were a number of outbreaks in social settings — nightclubs, restaurants, and social gatherings,” Van Kerkhove told Vox. The median age of cases dropped. So even though cases overall grew, hospitalizations and deaths were down — and that makes sense because younger people are more likely to experience mild or symptom-free infections.
Covid-19 demographics are changing again — shifting back into older populations. In France, for example, 4 percent of tests performed on people age 60 and older are now positive — a doubling from two weeks ago. By September 10, there was a 44 percent increase in the proportion of people over age 75 who have been diagnosed with the virus compared to the previous week.
In Italy, the median age of cases went down from nearly 60 in February to 30 at the end of August. It’s back up to 40 — and Riccardo thinks it’ll rise again as outbreaks move from social settings like nightclubs into households again.
[Updated data for September 14] Ongoing hospitalizations for COVID-19 in France's 10 largest urban areas— Edouard Mathieu (@redouad) September 14, 2020
• Nice: 38% of its first-wave peak
• Bordeaux: 36%
• Marseille: 34%
• Toulouse: 31%
• Rennes: 29% pic.twitter.com/ThAFQnLxX6
[Updated data for September 15]— Edouard Mathieu (@redouad) September 15, 2020
Latest COVID-19 figures for French hospitals:
• 2,713 new hospitalizations in the last week (+30% week-on-week)
• 479 new ICU admissions (+34%)
• 183 new deaths (+26%) pic.twitter.com/xlpAxJJqwx
Mathieu walked me through the situation in France, which he’s been tracking. In July, cases started increasing in a way that couldn’t be explained by testing alone — albeit slowly, doubling every two weeks instead of every 3.5 days, like in March. A rise in hospitalizations didn’t follow immediately.
It’s become clear that was because younger people were catching the virus. By mid-August, “the virus started to affect older people, and then a few weeks later, hospitalizations have started to increase,” said Mathieu. By September 10, the French public health ministry reported that new Covid-19 hospitalizations were growing in all but one region of the country.
“Now we are starting to see deaths increase,” Mathieu added. In late July, there were 10 Covid-19 deaths per day. Now, there are 30. “This whole process took almost two months instead of [several] weeks.”
The exact same trend is playing out now in Spain, he added. “In March, the number of deaths was doubling every two to three days in Spain,” he said. “The current rate is much slower — deaths are doubling on average every two weeks.” But they’re still doubling.
Even if it’s happening more slowly this time, it’s still exponential growth that could require more lockdowns
There’s one last factor to consider in the slowing Covid-19 death rate — and it’s good news. Doctors are better at diagnosing and treating the disease than they were at the start of the pandemic, so patients are more likely to survive.
“Governments like to congratulate themselves because they are not seeing the level of hospitalizations and deaths as we saw at peak — and the reason for that is not because Covid isn’t still burning in the same way,” said Lawrence Gostin, a Georgetown University global expert. “We’re catching it earlier and treating it better. So you see lower death rates because of earlier detection, better treatment, and elderly and vulnerable were either exposed and died or they learned the lesson and are staying away.”
While that’s certainly comforting, when cases start to rise and resources to deal with patients — drugs, personal protective equipment, beds, staff — don’t grow fast enough to meet demand, “your fatality rate will increase again,” Devi Sridhar, professor and chair of Global Public Health at the University of Edinburgh, warned.
And that’s when the specter of lockdowns starts reappearing. “If you see hospitals filling up and ICU beds full, [politicians will] have no choice but to implement some kind of lockdown unless you want your health system to collapse,” Sridhar added. “You can’t have people dying in hospital doorways because they can’t get access to oxygen.”
Sridhar wasn’t at all surprised that Israel just imposed a new lockdown, or that Madrid — where doctors are calling the situation “March in slow motion” — is also resorting to lockdown measures. She also predicts there will be more to come for Europe — that we’ll “pay for summer holidays with winter lockdowns.”
“Every country was under pressure to lift restrictions as soon as possible,” she said. “What’s happened is a muddling of economic and health objectives — we’re not doing either properly.”
Mathieu finds this frustrating since it was so predictable. It’s “exactly what happened the first time: People failed to think in terms of exponential growth.”
During phase one of the pandemic, coronavirus outbreaks proliferated exponentially, meaning an uptick in cases that appeared slow at first rapidly accelerated, outstripping countries’ abilities to manage them. Governments were forced to use the only instrument they had to deal with unchecked Covid-19 spread — one that didn’t require an exponential increase in doctors or hospital beds, Mathieu pointed out: the lockdown.
This time, the growth is happening even more slowly — we have an even better warning. Yet, Mathieu said, “There’s a weird benchmark thing going on where some people think there’s a threshold of terribleness and as long as we’re under that threshold, we shouldn’t worry about it” — even though it’s still exponential growth.
Take France, for example. While the country’s death rate is nowhere near the last peak, when 1,000 people were dying each day from Covid-19, “30 could soon be 50 or 100. And that’s dangerous for a mathematical reason ... exponential growth.”
If hospitalizations in France keep increasing exponentially at the current rate of 30 percent per week, for example, it’ll take only eight weeks to reach April levels again. “For now, every line is going up exponentially in France — hospitalizations, ICU admissions, death — even though for now the absolute numbers are still very low.”
“We need to strike a balance,” he added, “between telling people it’s not the same thing [as the first coronavirus peak], it’s not as bad, it’s slower — but we should do something about it.”
Oliver Johnson, a professor of information theory and the director of the Institute for Statistical Science at the University of Bristol, is also worried that people have already forgotten about exponential growth. “Two, three weeks ago, [people were saying] it’s just cases. Now, people are saying it’s just hospitalizations. And it’s like okay, well — what’s the next stage after that?”
Now we’re heading into winter, when social distancing is more challenging and people are more likely to gather indoors, where the virus has a much better chance of spreading. “We have reason to believe these kinds of viruses spread better when it’s cold. And if you start counting even to Christmas, it starts to look quite scary,” Johnson added. “My worry is that it’s a long winter.”