Go back

Researchers fight back in row over Covid-19 modelling

Image: Brian Smithson [CC BY 2.0] via Flickr

Lead author of government-commissioned study says he regrets agreeing to secrecy

The Italian Mathematical Union has come to the defence of the team that created models of the coronavirus pandemic used by the Italian government to plan the gradual reopening of the country.

The lead author of the models, however, has told Research Professional News that he would not do such work again due to the secrecy imposed by the government.

A row erupted after sections of the Italian media declared the models a “failure”, based mainly on figures extracted from the worst-case scenario, in which as many as 150,000 Italians would have required intensive care in hospitals. Instead the figure was a few hundred, critics said.

According to the latest figures, Italy has reported 237,828 Covid-19 cases in total and 34,448 deaths.

An open letter in support of the work was released on 10 June and co-signed by Andrea Pugliese, who studies mathematical models of epidemics at the Università di Trento, and Piermarco Cannarsa, president of the Italian Mathematical Union and a professor at the Università di Roma Tor Vergata.

They highlight that the scenario involving 150,000 intensive care patients was one of 49 and was the worst case examined. “From what we could see, it was good work”, Pugliese added.

The report was commissioned by the government in March from a research group of the Bruno Kessler Foundation in Trento. But it has remained secret, and Pugliese and his colleagues could only see an extract that was leaked to the press in April, stirring the controversy on government choices.

“It was never made public and not even shared with the scientific community,” Pugliese told Research Professional News.

Stefano Merler, the main author of the report, has defended his work. He says government decisions on locking down versus reopening the country were not based on those extreme figures. “What we did was the best possible model based on the limited data available, and we explicitly decided to err on the side of caution,” he told Research Professional News.

Merler, who was involved in a similar modelling exercise in collaboration with the Higher Institute of Health in 2009, during the crisis caused by H1N1, said his group was at first “disoriented” by the government request of confidentiality on the content of the report, and grew more and more dissatisfied after being attacked in the media along with the Higher Institute of Health.

“In the future I will not accept work for the government if a similar confidentiality is imposed,” he said. “I am in favour of full transparency.”