Predicting the ultimate outcome of the COVID-19 outbreak in Italy

TL;DR

This paper presents a data-driven method using logistic growth models to predict the final outcomes of COVID-19 in Italy, demonstrating how early data can inform public health decisions and project epidemic trajectories.

Contribution

It introduces a logistic growth-based approach to predict COVID-19 outcomes in Italy, incorporating data similarity with Hubei and updating predictions as new data emerges.

Findings

Predicted total deaths increased from 6,000 to 17,000 due to healthcare capacity limits.

Estimated epidemic end date shifted from April 15 to May 8, 2020.

Model accuracy improves with ongoing data collection.

Abstract

During the COVID-19 outbreak, it is essential to monitor the effectiveness of measures taken by governments on the course of the epidemic. Here we show that there is already a sufficient amount of data collected in Italy to predict the outcome of the process. We show that using the proper metric, the data from Hubei Province and Italy has striking similarity, which enables us to calculate the expected number of confirmed cases and the number of deaths by the end of the process. Our predictions will improve as new data points are generated day by day, which can help to make further public decisions. The method is based on the data analysis of logistic growth equations describing the process on the macroscopic level. At the time of writing of the first version, the number of fatalities in Italy was expected to be 6000, and the predicted end of the crisis was April 15, 2020. In this new version, we discuss what changed in the two weeks which passed since then. The trend changed drastically on March 17, 2020, when the Italian health system reached its capacity limit. Without this limit, probably 3500 more people would have died. Instead, due to the limitations, 17.000 people are expected to die now, which is a five-fold increase. The predicted end of the crisis now shifted to May 8, 2020.