While I have been doing routine COVID-19 model runs for the United States and Canada, I have not been forecasting in Europe. Unfortunately, I can’t say for sure whether the model’s stellar performance will translate to accurate COVID-19 forecasts in Europe. Let’s have a look at the best and worst case scenarios for the hardest-hit countries.

Identifying the Most Impacted Countries in Europe

How do we identify which countries we want to model? We have several tools to identify the most troubling trends in the existing data. First, let’s plot the new daily COVID-19 cases for all countries in Europe to identify surges. We are just looking for trends, so don’t worry about which line represents which country.

I’ve identified 2 clusters of COVID-19 surges in Europe. I’ve circled them in pink in the above plot. Countries in Group 1 are currently the hardest hit. On the other hand, significant imminent outbreaks of COVID-19 are likely in Group 2 countries.

Now that we’ve identified the patterns, let’s look at which countries make up each group. The hardest hit countries in Group 1 include France, Spain, and the United Kingdom.

Meanwhile, seven countries in Europe should expect an imminent and significant surge of COVID-19 cases. Group 2 includes Belgium, the Czech Republic, Germany, Italy, the Netherlands, Poland, and Switzerland. Did you notice that all seven countries have eclipsed their new daily case peak back in April? Some have done so significantly.

Confirm These Plots with a Map

Before we jump into the models, let’s confirm these plots so we don’t waste time modeling countries that are not experiencing a COVID-19 spike. My COVID-19 dashboard includes a map that will confirm the above plots.

Group 1 countries are currently the hardest hit. If we plot the new daily cases on the map, France, Spain, and the UK should show the highest numbers. The darkest blues indicate the highest number of new daily cases.

Group 2 countries are a little trickier. New daily case counts in many Group 2 countries are similar to other countries that are at risk for imminent outbreaks. Instead of looking at daily new cases, let’s look at the percent change in daily new cases over the past two weeks. Group 2 countries will have the largest percent change. Look for the darkest reds on the map.

Recent Model Performance

Following a strong June performance, my COVID-19 model has been on fire. Since 1 July, it has been correct on 90-100% of both two-week and one-month predictions. It did not get a single two-week prediction wrong in August.

The model’s recent performance gives me plenty of confidence in its predictions for Europe. However, models can be fickle, and its success in the US and Canada does not necessarily mean it will do as well predicting a COVID-19 surge in Europe.

Back in the spring, I wrote about how there are actually two best-fit lines for each country. We’ll use those as a baseline for the best and worst-case scenarios in our model runs.

While we hope for the best-case scenario, what actually happens will likely fall between the best and worst-case scenarios. The model projections for the worst-case scenarios are sobering, but remember the chance of the worst-case scenario playing out is extremely low.

Please note that I have run the model through next spring. The accuracy of the model predictions drops significantly for forecasts more than 1 month into the future. We’ll start our analysis with Group 1 in alphabetical order, followed by Group 2.

France

France is the most concerning country in Europe right now. It is registering a 7-day average of 20,000 new COVID-19 cases per day. There is no sign of it slowing down. In the best case scenario, cases will peak in late October or early November at about 30,000 to 35,000 new cases per day.

In the worst case scenario, cases will continue to spike into the winter. They will peak in late February with about 120,000 new cases per day. France will have over 15 million total cases by next summer.

Spain

The good news is that the best-case scenario is playing out for Spain…for now. New cases are showing signs of leveling off. However, they can restart an upward trend at any time. Heading into the winter with new daily cases elevated, the risk of a resurgence is high.

In Spain’s best case scenario, cases are peaking as we speak at about 14,000 new cases per day. While the model shows we may be past this peak, the actual new case counts are still rising, albeit slowly.

On the other hand, this could, and likely will, be just a temporary lull. While I highly doubt the worst-case scenario will play out, I think it’s equally unlikely that the current leveling off of new cases is actually the winter peak. Don’t forget, there’s still two months until the start of winter.

United Kingdom

As the UK approaches 20,000 new daily cases, it has shattered its April peak of about 5,500 new daily cases. As a result, the government has reinstated COVID-19 restrictions, including a three-tier lockdown. Several British media outlets estimate the new restrictions could impact over half of England’s population.

When you first look at the UK’s best-case scenario, you may find yourself a bit confused. Mathematical models do output confusing results from time to time. My best guess is that new cases will peak somewhere between the green and orange lines, or at about 20,000 to 25,000 new cases per day.

The time frame will closely follow the green line, but will be time-shifted about a month later. That means that while the green line shows the peak occurring in early October, the actual peak will likely occur in early November.

As with Spain, any leveling off of new cases within the next month or so will likely be just a temporary lull. Let’s shift gears and look at the worst-case scenario.

As eye-popping as the new case predictions are in the worst-case scenario, the chance of it actually playing out are next to zero. With the new lockdown rules in effect, there is no possible way COVID-19 cases will surge this much in Britain.

Let’s now pivot and look at Group 2 countries. These countries appear to be at the very beginning of a possible surge. As a result, model predictions will not be as accurate as those for France, Spain, and the UK.

Belgium

Belgium has seen an alarming spike in cases, going from 1,600 to 10,200 new daily cases in just the past 3 weeks. Even in the best-case scenario, the model has certainly reacted to the surge.

Like we saw in the projections for the UK, I think the green line is the most likely timeline for Belgium. Just how high cases will spike is much less clear. In a best-case scenario, they will not peak higher than the orange line. My best guess is that it will peak between 15,000 and 20,000 daily new cases.

Amazingly, the worst-case scenario is not much different from the best case scenario. However, in a worst-case scenario, the outbreak is longer in duration. Depending on your point of view, that’s either really good news or really bad news. In the worst-case scenario, 4.2 million Belgians will contract COVID-19 by Spring, 2021.

Czech Republic

In the context of COVID-19, you don’t hear much about the Czech Republic. It did an amazingly good job suppressing the virus in March and April, with daily new cases peaking at about 250 per day. However, cases are now exploding across the Czech Republic. The country reported 11,100 new cases yesterday.

Because new cases in the Czech Republic are surging so rapidly, let’s try a different tactic. The best and worst-case scenarios are so similar, let’s just consider a single case that is most likely to play out. It falls between the best and worst-case scenarios.

New cases in the Czech Republic will continue to climb throughout the rest of 2020. They’ll peak around the first of the year at 20,000 to 25,000 new cases per day. By next spring, COVID-19 will infect just over 3 million Czechs.

Germany

Germany’s current COVID-19 trajectory looks very similar to France’s. Germany has been more successful than France at suppressing the virus, but has suffered a recent resurgence in cases. In the current surge, Germany’s new daily case counts are a little over 1 month behind France’s.

Now, just because we say that Germany is a month behind France, it doesn’t mean that Germany will suffer the same fate. Indeed, the model shows that Germany will peak at about 12,000 new daily cases, compared to over 30,000 in France. However, the duration of the outbreaks in Germany and France will be similar.

Germany also fares much better than France in the worst-case scenario. Though highly unlikely, cases could spike to 120,000 per day in France. Meanwhile, Germany is not expected to see more than 33,000 cases per day, even in the worst-case scenario.

Italy

While COVID-19 ravaged Italy back in the spring, there is no indication from the model that Italy will be the epicenter of Europe during the surge this fall and winter. I expect many of the countries we’ve discussed today will peak higher than Italy.

The green line represents the most likely projection for Italy’s best case scenario. While I expect the timeline to be accurate, the number of new daily cases will peak higher than what the green line indicates, but lower than the orange line’s apex.

Italy’s worst-case scenario puts it on par with Belgium, the Czech Republic, and Germany. The model projects Italy will peak at 35,000 new daily cases around 1 January. I hope that Italians are especially sensitive to the seriousness of COVID-19 after being hit so hard in the spring. If they take COVID-19 seriously, there is no doubt in my mind that Italy will trend much closer to its best-case scenario.

The Netherlands

Like neighboring Belgium, The Netherlands successfully held COVID-19 at bay back in the spring. It’s now also facing a significant surge in cases. New daily cases have risen from 2,800 to 8,000 in just the past three weeks.

The model projects that cases will peak at 10,000 new cases per day around 1 December. However, I believe that the Netherlands will see closer to 15,000 new daily cases at its peak. It could be even higher than that.

The worst-case scenario paints a start contrast. Without action, new daily cases in the Netherlands could spike to 40,000 per day by early January. By next summer, nearly 6 million people could contract COVID-19.

Poland

Poland is another country that has not appeared in the COVID-19 conversation very much. Like the Czech Republic, Poland successfully suppressed the coronavirus back in the spring, but is now seeing a surge in cases. It continues to set new records every day.

In a best-case scenario, new daily cases will peak in mid-December. New daily cases at the peak will likely be higher than what the green line indicates. I don’t expect the peak to exceed what the orange line depicts. I think an apex of 15,000 to 20,000 new daily cases is a safe bet.

Poland’s worst-case scenario follows a very similar pattern. I expect the peak number of new daily cases would fall between what the green and orange lines forecast. Should that happen, over 6 million Poles could contract COVID-19 by next summer.

Switzerland

Much like the rest of west-central Europe, Switzerland is currently in the midst of a huge spike in COVID-19 cases. Its small size makes it a bit tricky to forecast, but here’s what the model’s best guesses are.

There is a chance that Switzerland could see what happened in Québec, Canada last week. A little over a week ago, Québec’s projected case counts looked almost identical to Switzerland’s best-case scenario. As quickly as cases shot up in Québec, they’ve peaked and have started coming back down. Don’t pin your hopes on it, though. Worst-case scenario forecasts for Québec were apocalyptic. I expect multiple waves will hit both Québec and Europe this winter.

Because Switzerland is starting at a lower baseline of new cases, its worst-case scenario looks far better than Québec’s. In this scenario, cases could peak as much as a month after what the model predicts. Should the worst-case scenario play out, about 2.7 million Swiss residents could contract COVID-19 by next summer.

Conclusion

The COVID-19 surge across Europe is especially concerning for other parts of the world. The European Union had for the most part suppressed the virus over the summer. Most countries reported less than 500 new cases per day this summer, and some have since spiked to over 30,000.

Countries that are letting COVID-19 run rampant should be particularly on high alert. These countries include Brazil, India, and the United States. The current explosion in cases across Europe looks to at least mirror the outbreak in Melbourne, Australia during their winter. Cases in Australia went up by two orders of magnitude between late June and early August.

France has gone from 400 new cases a day to about 30,000 cases per day. Much of the rest of central and western Europe has seen a COVID-19 surge of at least one order of magnitude, or ten-fold. They are showing no signs of slowing down.

Now imagine what those same calculations look like for a country like the United States, which is starting from a baseline of about 50,000 daily new cases. Buckle up. It’s gonna be a wild ride this winter.

The post COVID-19 Surge in Europe: How Bad will it Get? appeared first on Matthew Gove Blog.