I should have stuck with it when I gave up before.
Tell you what. You tell me which numbers I have wrong in this table and then we can discuss - hopefully free from the emotional aspect of this.
Code:
Population: 1,000,000
Sex Orientation Activity
Male 50% Het 95% N-A: Never had Anal
Female 50% Hom 5% Top Only: Only ever given anal
Other: Have received anal (I did this at work, I had to be somewhat careful about my labels)
% of
(Sex/Orient) # of Risk HIV
Sex Orient Activity Pop People of HIV Cases
Male Het No-A 100% 475,000 0.30% 1,425.00
Male Het Top Only 0% 0 0.00% 0.00
Male Het Other 0% 0 0.00% 0.00
Male Hom No-A 5% 1,250 2.00% 25.00
Male Hom Top Only 10% 2,500 3.00% 75.00
Male Hom Other 85% 21,250 5.00% 1,062.50
Female Het No-A 60% 285,000 0.30% 855.00
Female Het Top Only 0% 0 0.00% 0.00
Female Het Other 40% 190,000 0.75% 1,425.00
Female Hom No-A 100% 25,000 0.10% 25.00
Female Hom Top Only 0% 0 0.00% 0.00
Female Hom Other 0% 0 0.00% 0.00
If my assumptions are close to being right, then we have the following:
1,000,000 people, 4,892.50 HIV cases, so an infection rate of 0.49%
The highest risk group are Male-Hom-Other, so we remove them and see what happens
We remove 21,250 people and get rid of 1,062.5 HIV cases, so each HIV case we got rid of cost us 20 potential blood donors. I think we can all agree that this is a good trade off.
The next highest is Male-Hom-Top Only, so we remove and see what happens.
We remove 2,500 people and get rid of 75 HIV cases, so each HIV case we get rid of costs us 33.33 potential donors. I think this is a decent trade off. You may disagree.
The next highest is Male-Hom-No-A. We remove
That is 1,250 people and 25 HIV cases, so 50 donors per HIV case reduced. Again, could be argued that this reduction is worth it.
The next highest is Female-Het-Other.
That is 190,000 people and 1,425 cases, for 133.33 donors per case reduced. Starting to get pretty expensive.
Anyway, suggest some changes you would like to see to the assumptions, then we can discuss some actual (although made up) numbers. Maybe that will help.