Watching the numbers

Despite the reopening of most borders in Europe, making travel plans for the summer causes most of us to feel uneasy. Many of us follow the development of COVID-19 infections across the world that are offered in real time, such as those reported by Johns Hopkins University. We can look at the number of infections per capita and see whether the infection curves are headed upwards or downwards. We’re concerned when we see a sudden rise in cases from countries that were thought to be relatively safe and wonder whether we should still take the risk to travel there. Previously booked flights and organised tours can still be cancelled at any moment, causing uncertainty about whether we will be able to take our carefully planned trips. I have certainly decided to cancel my summer trip to the US, where part of my family lives, as the pandemic is raging there at the moment.

To hold or to fold?

When organising a vacation, most of us don’t want to take this gamble and many of my friends and colleagues have decided to stay here for the summer. But this is often decided more out of frustration than real enthusiasm after we’ve already had to spend several months at home with the family. Of course, we can always take a vacation within the Netherlands—but that might be a crowded experience this year, despite the social distancing measures that are in place.

Time delay

Clearly, the current vacation season can be characterised as yet another pandemic-induced change from our normal routines. What makes travel planning so difficult is that the new SARS virus is hitting us with a time delay, and we only know that there is a problem weeks after certain measures (such as opening vacation resorts across Europe) are taken. Once we see a spike in infection rates or a rise in hospital admissions, it is often too late to contain the spread. As a result, severe measures such as lockdowns and border closings will most likely be implemented suddenly, which could affect the mobility of travellers.

Predictions using big data

The early prediction of infection hot spots could help us overcome these obstacles, and researchers have increasingly been using big data approaches from online resources to do just that. A paper published this week by a group of researchers from Harvard University applies an algorithm that includes a variety of data points that are readily available online to predict areas where the infection rates are most likely to rise. First, they looked at a multitude of data points such as corona-focussed searches from physicians in medical databases, tweets about COVID-19 cases, available data from thermometer apps and many others. They then linked these to geographical locations to develop an algorithm (a calculatory model) that can predict where infection rates are likely to rise.

Early prediction would allow governments to implement local measures, such as focussed testing and isolating those who are at greatest risk. This would give them a better chance of containing the spread. The researchers showed, for example, that this model was able to predict the development of current hotspots in the United States three weeks before the spike in infections became visible. By adding additional parameters to their model, they are striving to make even earlier predictions which could in turn help officials contain the spread of the virus more efficiently.

Early stages

Although this big data model is still in an early stage and needs to be validated in larger study, it is an interesting approach. Previous reports have strongly suggested that the virus was present in Europe weeks or even months before it had become visible in the disease statistics, and effective prediction models that use big data analysis could have led to significant improvements in mitigating infection rates. In the future, these models could help officials to direct resources and implement targeted measures in areas where they are needed most—without locking down entire countries and affecting the population at large.

And, who knows, maybe we will plan our vacations in the future not only based on predictions of the highest number of sunny days but also on the risk calculations from a corona travel app. This year, however, we still very much have to depend on our gut feeling when we make our travel plans. I therefore wish you not only a good decision but also a nice and restful summer, whether you spend it in your garden, at a Dutch beach or traveling to another country.

 This is the last edition in the corona blog series by prof. Martin Paul. After summer recess he will of course continue blogging, but on a wider variation of topics.