Miguel Angel Lozano, a computer scientist at the University of Alicante in southern Spain, saw one of Oliver’s first emails asking for help. For two decades, Lozano had researched and lectured in computer science, with a recent emphasis on delineating patterns in complex systems like public transport networks. âI think everyone wants to help in this situation,â Lozano recalls. He agreed to participate and quickly saw where his experience in mapping transport data could come in handy.
Based on his previous work with Telefonica and Vodafone, Oliver knew that measuring and understanding mobility could be crucial in preventing the spread of the virus. âIf we don’t move, there is no pandemic,â she said. Luckily, the Spanish National Institute of Statistics had recently started examining anonymized telecommunications data from three major mobile phone companies, for economic purposes and to monitor the movement of the country’s workforce during peak hours. , among others. After some negotiation, Oliver and his team were granted unique access to this same mobility data and, says Lozano, “applied some of the methods we were working on to this information.”
Francisco Escolano, another computer scientist from the University of Alicante who had joined Oliver’s team, found the mobility data spreadsheets provided by statisticians to be important, but also relatively basic: they simply showed that a specific number of people were moving from one area to another within a specified period of time. To make this data more practical, Escolano helped develop a system that would translate statisticians’ spreadsheets into the Python coding language. This not only allowed them to study the data more closely, but also to create clear visuals of it that could be understood by policymakers.
Oliver scientists collected data from a variety of other sources. They scoured public Facebook and Google posts, looking for additional data that would cover the gaps. They also developed a digital survey for the people of Valencia, whose questions have changed over time as the pandemic has evolved.
The surveys were surprisingly well received by residents, with over 140,000 responses in the first 40 hours. Over the next few months, they generated hundreds of thousands of additional responses – data points – that helped the Generalitat see what its people thought and felt almost in real time. Responses included where, when and how social mixing was done, what personal protective measures residents were taking, how they perceived the relative safety of different activities, from groceries to dining out, and whether individuals were felt secure enough financially to isolate themselves if necessary. âWe were able to answer questions that we wouldn’t have been able to answer otherwise,â says Oliver.
The survey results proved particularly effective in cabinet meetings, Berenguer says, as they showed politicians how behaviors changed in ways they could easily understand. If people seemed to let their guard down and socialize too freely, for example, the government could launch a new public awareness campaign promoting respect for mask wearing or social distancing. But self-reported increases in social interactions could also help improve the accuracy of estimates of the number of cases.
The team also developed other predictive models based on machine learning. One allowed them to predict the prevalence of Covid-19 in a given area at a given time; another helped them analyze wastewater from baths, sinks, washing machines and showers, and look for abnormalities that might reflect changes in local infection rates. A third allowed them to predict future hospitalization rates, recognizing when intensive care units might reach capacity. This has proven to be extremely useful for local health authorities as the pandemic has progressed, allowing them to move staff and equipment across the region to meet expected demand.
During the first and second waves, Valencia spared the worst of the pandemic; in the first week of April 2020, for example, residents of Madrid were dying in four times the rate from that of Valence. At the start of November 2020, Valencia had the lowest total number of cumulative infections by population size of any region in Spain.
But that changed in December 2020, when a third wave – fueled by the more contagious Alpha variant that had emerged in England – caught regional authorities off guard. Although Valencia later became the only part of Spain to ban inter-regional travel around Christmas time, the transmissibility of the Alpha variant had already meant that the virus was circulating more than what was understood by testing. Visitors from elsewhere in Spain and overseas had traveled in and out of this popular tourist destination during months of relaxed social distancing restrictions – with disastrous results. The average number of infections recorded daily in the region rose from 1,450 at the end of December to more than 8,000 a month later, and during the same period hospitalizations have more than tripled and daily deaths have multiplied by six. In addition to a handful of areas in Portugal, Spain’s neighbor, the region of Valencia over 14 days had the highest cumulative incidence of infection in all of Europe.