Both are on a yearly basis
so... 2112 deaths/22 months = 96 deaths per 160,000 soldiers PER MONTH. This is equivalent to about 60 deaths per 100,000 soldiers --> That means that the firearm death rate quoted (and I'm not sure all of these were from firearms anyway, but even if they were) is 60 per 100,000 soldiers PER MONTH. How else can one read the statistic you provided?
Did you see the analysis in post 101?