Most of us take the weather forecast for granted. It may not be perfect every time, but we still check the weather app on our phone when deciding if we should take a coat in case it rains later. Or perhaps the forecast for the weekend will help us to decide whether to plan a hike – or to book cinema tickets instead.
Policymakers need this kind of information too. In the short term, it’s useful to know the risk of storms and heatwaves. For longer term planning, the insights of climate scientists could inform decisions about where to build houses or flood defences; and whether solar or wind energy might be a better bet.
But what about forecasting disease outbreaks? ‘Is there an app for that?’
Well, mathematical modellers at the European Centre for Disease Prevention & Control (ECDC) have been working on new tools which offer short-term forecasts of respiratory illnesses, as well as insights that can help planners to ensure health systems are ready for surges in demand.
RespiCast was developed during the pandemic to help forecast COVID-19 outbreaks and has since expanded to look at other respiratory diseases, notably influenza.
To complement this, last month (December 2024) saw the launch of RespiCompass. It will be used to examine the impact of public health interventions, such as vaccination, on disease burden in Europe. For example, it could help to anticipate the demand for hospitalisation and to calculate what proportion of these cases could be averted through vaccination.
RespiCompass is new and will need several months to begin delivering answers to new health policy questions. However, RespiCast is already producing short-term forecasts for this winter virus season. Modelling teams from around the world can participate by submitting forecasts each week. These are combined into a ‘model ensemble’ – a kind of aggregate forecast. This is seen as a more accurate approach to forecasting than relying on a single model.
The project was initially called the COVID-19 Forecasting Hub, but now collects a wider range of indicators from a growing number of experts, according to Dr Eva Bons, ECDC Expert on mathematical modelling.
‘Last season, seven different modelling teams submitted forecasts to RespiCast using 15 different models,’ she says. ‘This year we have even more teams participating with even more models. So far this season, we have received 25 different models from 10 different teams. This growing participation is helping to build a community of modellers, which we hope will drive innovation and improvements in model performance – one of our core objectives.’
In a survey recently of health services in EU/EEA member states, around 80% of respondents said that ECDC collaborative modelling platforms like RespiCast were likely to inform public health actions or decisions at the national level.
‘We have even had staff from public health agencies reach out directly to say they find our forecasts helpful. It’s feedback like this that keeps us enthusiastic and committed to the project,’ Dr Bons says.
The data generated by these new tools are available online, although most of the users are experts or policymakers. ‘Additionally, by bringing various modelling teams together on a regular basis, we aim to strengthen modelling capacity across Europe, ensuring that our forecasting capabilities continue to improve and can support public health efforts effectively.’
Disease forecasting is not yet a rival to the nightly TV weather forecast or the weather app on your smartphone, but it has taken off since the COVID-19 pandemic and looks likely to become a feature of health system preparedness.
It could help hospitals to manage bed capacity, health systems to encourage vaccine uptake, and may even allow vulnerable individuals to anticipate periods of high risk.
Whatever the winter has in store, it’s always best to have your recommended vaccines. Otherwise you risk being under the weather.
