I provided another link (
http://www.avert.org/usa-race-age.htm ) above, which indicates that 75% of known (M and F) HIV infections in the US are believed to result from MSM and/or intravenous drug use (IDU): this suggests that HIV contamination of blood could be reduced by a factor of four simply by eliminating those two groups as donors -- and since no HIV test is perfect, one might still expect a similar four-fold reduction in HIV contamination
of HIV-tested blood by eliminating those two groups as donors
I also provided above a link (
http://egov.cityofchicago.org/webportal/COCWebPortal/COC_EDITORIAL/STD_HIV_AIDS_Chicago_July09.pdf ) to Chicago's survey last summer, which estimated 17+% HIV status for MSM (half of the individuals being unaware of their HIV status), compared to an overall rate in Chicago of 1.2%. If one adopts (say) 5-6% as the rate of MSM (see:
http://www.kinseyinstitute.org/resources/bib-homoprev.html), this would suggest 75-87% of Chicago male HIV infection is associated with MSM, which is roughly consistent with the 70% estimate one obtains with national data from the avert link above. You are free, of course, to argue that the Chicago survey somehow completely failed to sample a large uninfected MSM subpopulation and that in Chicago MSM really occurs at a much higher rate (say 10%) -- so that the true HIV prevalence among Chicago MSM was actually smaller (say 12%) -- but some evidence would naturally be required that such argument had merit
One obtains similar estimates simply on the basis of new infections: the CDC (
http://www.cdc.gov/nchhstp/newsroom/docs/Fact-Sheet-on-HIV-Estimates.pdf ) estimated that 70% of all new US HIV infections in 2006 were among MSM and/or IDU. Crudely assuming that the epidemic has stabilized (which is, of course, not an entirely accurate assumption), leads to the immediate conclusion that about 70% of the epidemic will be localized in the MSM and/or IDU population -- which is what the data shows
If you want a more sophisticated analysis, you can use the 2007 HIV statistics by age table at the advert link combined with the US census 2006-2008 age and sex table data (from:
http://factfinder.census.gov/servlet/ACSSAFFPeople?_submenuId=people_2&_sse=on). This will enable you to estimate the annual risk (for M and F combined) by age group. The ages 15-49 are the high risk ages, so concentrate on them. From the avert tables, we expect 50% of hiv to be associated with MSM (with or without IDU); from Kinsey, take (say) 6% of the population MSM; adjust the age-risks accordingly; set up a spreadsheet and follow a cohort from age 5 to age 49: at the end, 13% of the population has been infected, not much different from the Chicago study's 17%. The assumptions about converting combined M and F risks to MSM risks only are a bit sloppy, so one should not take the numbers too seriously; in particular, it's only off by a factor of two from the five city 25% estimate
Public health policy should be determined by one's best honest assessment of the facts -- which will always be imperfect. In reality, of course, it is also influenced by other factors -- but for this blood bank donor policy question, one ought to have a robust understanding supported by multiple models and multiple data sets
I had no idea who "Paul Cameron" was, until reading your post, but on investigating I find he's irrelevant to this discussion: his professional society expelled him over 25 years ago, and nobody except lunatics will take him seriously