Posted21 avril 2022, 12:06
A study by the Swiss National Fund carried out at the CHUV makes it possible to predict possible waves of flu thanks to weather data. The same approach for the Covid is being studied.
A team of scientists, supported by the Swiss National Science Foundation (SNSF), has just developed a mathematical model that predicts the risk of hospital congestion linked to influenza depending on weather conditions. His work has been published in the Journal of the Royal Statistical Societysaid the SNSF on Thursday in a press release.
We learn that, as the flu is a seasonal virus mainly present in winter at our latitudes, the scientists compared certain meteorological data – precipitation, humidity, temperature and sunshine – and the cases of flu recorded daily for three years at the CHUV in Lausanne. But, contrary to what is usually done, the research team was not interested in the daily average of cases over these three years.
Extreme values
It focused on the extreme values recorded because these are the values that can indicate a risk of congestion for hospitals. Data that is useful for resource planning, emphasizes the SNSF. Scientists have developed a model that uses weather data to predict the risk of congestion three days later, the incubation time for the flu.
“Instead of telling hospitals an average value of expected cases, we tell them the probability that a number of cases exceeding their capacities will be reached, which is more relevant,” explains Valérie Chavez, statistician at the University of Lausanne. and co-author of the study, cited in the press release. By following the evolution of this probability each year from the fall, hospital managers might therefore anticipate a peak in flu cases and therefore a possible overload.
Applicable to seasonal viruses other than influenza – in particular coronaviruses and respiratory syncytial virus (RSV), responsible for respiratory infections in young children – the prediction model set up by the National Research Fund presents for the There is now some uncertainty in the risk estimate, because only three years of CHUV data might be analyzed. Moreover, still due to the lack of data, it is not yet applicable to the monitoring of Covid-19, explain the researchers. On the other hand, scientists are already working on models which, alongside weather data, would also exploit the processes of propagation of viruses in order to follow the phenomena of contagion even better.
(Comm/jba)
