Research Press Release

Epidemiology: A new model predicts COVID-19 outbreak dynamics in Italy

Nature Medicine

April 23, 2020

A new model that predicts the course of the COVID-19 pandemic in Italy using data from the outbreak is reported in a paper in Nature Medicine. This model — which considers eight stages of infection and differentiates between diagnosed people and non-diagnosed people — could provide policymakers in Italy and elsewhere a tool with which to assess the consequences of possible strategies, including lockdown and social distancing, as well as testing and contact tracing. The study shows that the adopted social-distancing measures are necessary and effective and should be promptly enforced at the earliest stage. Lockdown measures can be relieved safely only in the presence of widespread testing and contact tracing, the findings suggest.

Ending the global COVID-19 pandemic requires the implementation of multiple population-wide strategies, but the effectiveness of such strategies and their ability to ‘flatten the curve’ remains uncertain.

Giulia Giordano and colleagues describe a new epidemiological model for the COVID-19 pandemic, named ‘SIDARTHE’, which distinguishes between detected (diagnosed) cases and undetected (undiagnosed) cases and among different severities of illness. They divided the population into eight stages of the disease: susceptible (uninfected); infected (asymptomatic or presenting few symptoms, infected, undetected); diagnosed (asymptomatic infected, detected); ailing (symptomatic infected, undetected); recognized (symptomatic infected, detected); threatened (infected with life-threatening symptoms, detected); healed (recovered); and extinct (dead).

The authors used data from Italy from 20 February 2020 (day 1) to 5 April 2020 (day 46) to show how the progressive restrictions, including the most recent lockdown enforced since 9 March 2020, have affected the spread of the pandemic in Italy. The authors also modeled possible longer-term scenarios of the effects of various countermeasures, including social distancing, contact tracing and population-wide testing. The model predicted that the peak number of actual concurrently infected people would occur around day 50, with 0.19% of the population infected; however, the peak number of detected concurrently infected people would occur about one week later.

The findings confirm the hypothesis that diagnosis campaigns can reduce the infection peak and could help end the pandemic faster. The model does not consider reduced availability of medical care due to the healthcare system’s reaching or even surpassing its capacity, but the authors note that these analyses can be done indirectly. For example, when the number of seriously affected people is high, the mortality ratio will be increased due to an insufficient number of intensive care units (ICUs).

The authors also found that partial implementation of lockdown measures would result in a delay in the peak of concurrently infected people and patients admitted to the ICU, but with an only moderate decrease in the total number of infected people and ICU admission. Conversely, the implementation of very strong social-distancing strategies would result in an anticipated lower peak of concurrently infected people and patients admitted to the ICU, with a marked decrease in the total number of infected people and ICU admissions due to the disease. The authors estimate that there would be 70,000 deaths in the first year with a less-stringent lockdown, and 25,000 deaths with a more-stringent lockdown.


Return to research highlights

PrivacyMark System