Rossi's article of 1996, which is Carrier's source for some data, is a potpurri of different and not clearly stated sources, an avalanche of numbers without much detail. Even Rochat, who in his article "Una ricerca impossibile. Le perdite italiane nella seconda guerra mondiale" (1995) made use of this study, then discarded it in his subsequent book "Le guerre italiane 1945-1943" (2005), especially with regards to the Greek campaign, where Rossi's data are excessive. Rochat believes that deaths were around 18,000, which he then corrects to more than 20,000 with the arbitrary addition of "deaths in hospital", which is without any sense (unless he is counting deaths happened long after the end of the campaign, and he gives not explanation anyway). My own calculations, based on archival sources, put Italian deaths in the Italo-Greek war to less than 17,000.