Avian flu, Covid and other pandemics. The epidemiologist Vespignani: «This is how AI and Big Data help us predict the next ones»

«As we often reiterate in our field, it’s not a question of “if” but “when” the next one will be pandemic. Recently, in the USA, the expansion of H5N1 avian influenza to cattle has occurred, but the list of potential threats is long. Pandemics have marked the history of humanity and will continue to do so. We will also find them in the future. Better to be prepared.” It’s the advice of Alessandro Vespignani among the leading experts in computational epidemiology at an international level (see opposite).

His research focuses on development of models to predict the spread of infectious diseases and understand its dynamics in society. These models are used to simulate disease spread scenarios on a global or national scale by integrating demographic and mobility data to predict the how, where and when of epidemiological threats.

«Professor Vespignani, the general public has learned to know you during the Covid-19 pandemic: preparation and response to pandemics have improved»?
«During the pandemic we have made significant progress: from large-scale sequencing of viruses and the mapping of population mobility, up to more advanced forecasting algorithms and valuable experiences on how to communicate and make decisions. However, numerous challenges remain to be addressed. On the one hand we can access data on the mobility of millions of people in real time, on the other the data from local health situations often arrive late and are not easily accessible. It is necessary to develop methodologies to better understand the often asymmetric effects of diseases and interventions on different social strata, highlighting the observed inequalities.

«We must transform the initiatives launched during the pandemic into permanent structures, such as the National Centers for Forecasting and Analysis. In the USA, for example, the federal government has launched the national network of Insight labs which, with huge funding, multiplies the country’s analytical capabilities through the use of machine learning, artificial intelligence and the best technologies available.”

How do you use computational models to predict the spread of infections?
«Our work as scientists it does not replace that of doctors, healthcare workers and volunteers, true heroes of every epidemic. We are fighting a different battle, based on numbers and information to provide decision makers and the healthcare system with the best strategies to anticipate and fight the pathogen. Computational epidemiology is an additional weapon which, from behind the lines, generates the crucial information to understand and anticipate an often invisible and unknown enemy, which spreads through our behaviors and interactions”.

How do AI and Big Data come into play?
«The integration of AI and Big Data in epidemiology improves the analysis of large volumes of data in real time, refining predictions and understanding of disease transmission dynamics. Tools such as epidemiological surveillance on participatory platforms, Influweb in Italy e Influenzanet in Europe, coordinated by Isi Foundation researchers, the use of data from social networks and large-scale genomic surveillance are revolutionizing research in this field, allowing new ways of collecting data and transforming them into knowledge.”

What is the role of research in the formulation of public policies and in the management of health emergencies?
«Using numbers, data and models, we analyze what happens at the epicenter of the epidemic, the effects of mitigation interventions and signs of possible spread. As we get to know the pathogen better, we focus on designing strategies to combat it, developing forecasts and scenarios. These not only sound the alarm for decision makers, but produce information on the effectiveness of containment measures and estimate the impact of our behaviors on the trajectory of the epidemic. All fundamental elements in the decision-making process to combat health emergencies.”

What technical and organizational skills are needed to effectively deal with the next pandemics?
«We must introduce a multidisciplinary approach through the integration of technology within public health paradigms and practices – says Vespignani -. Each epidemic is a story in itself. The best response to a health emergency is one capable of quickly adapting to the enemy on duty. This can only happen through a process of constant planning and exercise that must be done before emergencies. Through the definition of pandemic response centers, infrastructures, laboratories capable of being activated quickly. And a lot of scientific research. We cannot only remember the need for science when we have a problem. The “search system” is not created in a day. Knowledge is not transferred and operationalized on command.”

The government has launched the National Pandemic Plan 2024-2028: will it work?
«The new pandemic plan broadens the spectrum of possible threats with pandemic potential, outlining not only the possible intervention tools, but also the structures and ways for emergency coordination at national and local level. This plan reinforces already tested methods and strategies, and highlights the importance of coordination and implementation timing. In this context, it becomes central to equip yourself with those centers for the prediction and analysis of epidemiological risk, distributed throughout the national territory and supported by the most advanced computational technologies and artificial intelligence that can make the difference. Pandemic plans are documents, but it is up to governments to translate them into resources and infrastructure that are ready in the event of an emergency. Otherwise it’s just papers in a drawer».