The underlying data isn’t smoothed, but I had to do it that way to extract it from the webplot (the data isn’t published and the author couldn’t send it because Italians aren’t allowed back in their offices).
Charts were made in R with ggplot2 using this script:
U.S. Flu data is from the CDC, Italian data is from Rosano et. al. 2019, and COVID-19 death total from Johns Hopkins CSSE.
Edit: here is a same-scale version: