What percent of gay men are infected with HIV?

[dsc]

Some here have posted an infection rate of 1/6 to 1/4. That sounds pretty horrible and makes seem reasonable that gay men should be banned from giving blood and are probably taking enough anti AIDS drugs to fill the world over. Too bad, or actually thank God, the stats are just plain wrong. First for where the wrong statistics come from.

http://www.cdc.gov/nchhstp/newsroom/docs/FastFacts-MSM-FINAL508COMP.pdf

I can't quote a pdf so I will summarize. The CDC conducted a study of gay men in five cities. I have no idea which cities or how the gay men were selected but it can't have been a truely random sample. You can't just give people HIV tests without their concent. So this had to be a self selected study which means that the people who participated wanted to participate. Those are going to be people who tend to suspect they may be infected due to having engaged in risky behavior. It isn't going to be the married gay couple down the street. Since there is no link to the study I don't know how they got their people. Did they hang out at bars? Did they advertise on the radio? Which cities? These questions matter. We don't know the answers. Ok so here I show it might be wrong. Now I will show it is wrong.

http://www.cdc.gov/hiv/topics/surveillance/basic.htm#aidscases

http://www.census.gov/population/estimates/nation/intfile3-1.txt

The top link gives us the number of HIV/AIDS cases while the bottom gives up the population of the US.

According to the first link there were 487,695 cases from same sex male contact, 71,242 cases from both same sex male contact and IV drug use, and 18,266 other cases which I will assign 6% to being from gay males. That gives us 560033 cases. Multiplying by 4 gives us 2,240,132 gay men in the US while multiplying by 6 gives us 3,360,198 gay males in the US. Thus for the proportion of gay men infected by HIV/AIDS to really be somewhere between 1 in 6 and 1 in 4 then the number of gay men would have to be between 2,240,132 and 3,360,198.

So how many gay men are in the US?

According to the US Census there are approximately 134,979,000 males in the US. Most reputable people state that about 6% of males are gay. That is about 8,097,840 gay males or over 4 million too many for even the 1 in 6 figure to be correct. If you use 3% instead of 6%, which puts you very close to the way too low figure of 2% that conservative anti gay bigots use, you still have 4,048,920 or about 80,000 too may gay people for even the 1 in 6 figure to be correct. For the 1 in 6 figure to be correct you have to have about 2.5% of people being gay. The 1 in 4 figure isn't even correct in the fundy delusion land that Robertson lives in.

So what is the correct percentage?

I can't give a totally definative answer but it should be close to 7% which is 560,033 divided by 8,097,840. High? yes. But no where near the percentage some are quoting.

[Djarun]

1. ?????

What percent of straight women are infected with children?!?!?!?!

[Deja Q]

3. Children are easier to get rid of and aren't lethal...

Well, usually not lethal but I've read stories in the newspapers about some very nasty children...

Apples and oranges - invalid comparison.

It's a scary statistic as well. There are some on DU who claim gay men know of the disease and do things safely. If it is prejudice or a blind statement to say "All gay men are reckless", it is also equally dim to say "All gya men practice safe conditions". (the truth is in the middle. I would and many don't. Thankfully, the chances in getting it from an unwanted french kiss is minimal and only possible if everyone involved had open sores and a bad case of gum disease in their mouths. Yuck.)

[Djarun]

11. invalid comparison?

Is it because *ss babies never survive?

[Starry Messenger]

2. Thank you for your work on this dsc.

Excellent post.

[Deja Q]

5. I also agree, excellent post. Insightful and it raises genuine thought.

Thank you, DSC.

[crazyjoe]

4. what percentage of gay men refuse to use condoms?

[dsc]

6. I haven't a clue

nor is it relevent to this discussion. I would be willing to bet gay men are better about condom use than any other group of people though.

[Name removed]

7. Deleted message

Message removed by moderator. Click here to review the message board rules.

[Name removed]

8. Deleted message

Message removed by moderator. Click here to review the message board rules.

[PeaceNikki]

10. That's some seriously fucked up logic.

Are you saying that because gay men didn't run out and get married the day it became legal in California that they don't "want civil unions and then keep their vows"?

[aquart]

12. Why should gays be required to be more "emotionally ready" than straights?

I'm pretty sure Tiger Woods wasn't "emotionally ready" for marriage.

Or Governor Sanford. Or David Vitter. Or Senator Ensign.

Nobody considered that it might be a public service to stop them.

[Name removed]

14. Deleted sub-thread

Sub-thread removed by moderator. Click here to review the message board rules.

[Name removed]

18. Deleted sub-thread

Sub-thread removed by moderator. Click here to review the message board rules.

[Lyric]

30. And how many gay people in California didn't want to marry until they knew

for sure that their marriage wouldn't be dissolved by the courts or overturned by the voters? Perhaps you're different, but as a general rule, people tend to be very wary of willingly subjecting themselves to extreme emotional pain. Until the final ruling, there was always a risk that even the marriages that happened before PropH8 would be nullified. I know *I* would have waited until I knew for sure that I wouldn't be putting my heart at risk of being shredded by the bigots in charge.

Are gay people ALL supposed to be superheroes, willing to risk our hearts at every turn, or are we allowed to be human now and then?

:eyes:

[donco6]

32. Jesus - there's so much wrong with this.

36,000,000 all of marrying age?

Hardly.

Can you flagellate yourself somewhere else for a change?

[dsc]

35. a couple of things

first, there were 18k marriages which is 36k people. Sheesh you could at least multiply correctly. Second, there are a whole host of reasons people may have decided not to marry. One is that the legal status was shaky at best. Two the feds still wouldn't grant rights. I could go on but you get the point.

[ruggerson]

41. A lot of people didn't get married because they didn't want

the eventuality of other people VOTING on their marriage.

And what's this "keep their vows" bullshit? 50% of hetero marriages end up in divorce and you're making up arguments about gay people "keeping their vows?"

[niceypoo]

50. Probabally the same percent that are infected with AIDS

...or close to it

[HamdenRice]

52. Sorry, but this is just dumb

The idea that gay men "refuse to use condoms" is an urban myth on the level of black women welfare queens driving cadillacs to pick up their welfare checks.

[crazyjoe]

56. I never said gay men refuse to use condoms, I'm saying that if you don't,

you are much more likely to be a gay man with aids. Do you disagree?

[name not needed]

61. Probably the same as the amount of straight men who refuse to.

[Mugweed]

9. Probably not as many as straight men wth herpes

nt

[Djarun]

13. Or as many straight men with children out of "wedlock"

[Unvanguard]

19. Gay sex has yet to cause a single unintended pregnancy. n/t

[Djarun]

34. thank goodness for that

[JonQ]

45. Not really a valid comparison

AIDS = children born out of wedlock.

Sorry, I'm not buying that.

[Unvanguard]

15. The city estimate is almost certainly high. Urban populations are not representative.

I doubt it was meant to be, either. Another example of people misusing science for their ideological agenda.

If I recall correctly, the CDC usually uses around 4% for its number of MSM. Using the numbers you provide, that's 9% or so. Still much lower than 1/6, which is definitely too high.

[struggle4progress]

59. The urban rates reported are somewhat greater than the rural rates, but the difference isn't huge:

... in 2006, 82% of AIDS cases were reported from large metropolitan areas and 11% were reported from medium-size metropolitan areas ... As of 2006, data from the U.S. Census showed 65% of the general population of the United States was living in large metropolitan areas, 19% in medium-size metropolitan areas, and 17% in nonmetropolitan areas ...
Cases of HIV Infection and AIDS in Urban and Rural Areas of the United States, 2006
http://www.cdc.gov/hiv/topics/surveillance/resources/reports/2008supp_vol13no2/commentary.htm

So 93% of AIDS cases reported in 2006 were from large or medium-size metropolitan areas, which contain about84% of the population: 93/84 = 1.107, so AIDS seems about 11% more likely in a large or medium-size metropolitan area, than in the country as a whole -- and similarly about 50% less like in medium-size metropolitan area, than in the country as a whole But might these figures simply represent a greater willingness to be tested in large urban areas?

[Unvanguard]

60. No, the difference is probably larger than that.

There's a huge difference between "living in large metropolitan areas" and living in large cities. "Metropolitan areas" include the suburbs, which almost certainly have a lower AIDS rate than urban populations. The CDC study, as was pointed out elsewhere in the thread, was not even attempting to measure national prevalence rates, but rather to study the epidemic in cities that were already known to have high rates of HIV infections (see pinto's #46).

For an example of the importance of this demographic distinction: if 65% of the US population really lived in large cities (as opposed to large metropolitan areas), then the Democrats would easily win every national election (well, assuming away the party realignment that would certainly ensue). Even though both are technically "urban", they are, sociologically, quite different categories.

[struggle4progress]

63. Sorry, no cigar: look at the definitions in the report I linked:

... this report uses the Office of Management and Budget designations for metropolitan statistical areas. Persons diagnosed and reported with either HIV or AIDS are assigned, based on the place of residence at diagnosis, to a large metropolitan area (population greater than or equal to 500,000), a medium-size metropolitan area (population 50,000 to 499,999), or a nonmetropolitan area (population of less than 50,000) ...
http://www.cdc.gov/hiv/topics/surveillance/resources/reports/2008supp_vol13no2/commentary.htm

The OMB definitions can be obtained from the OMB website:

Metropolitan and Micropolitan Statistical Areas
http://www.census.gov/population/www/metroareas/metroarea.html

The MSAs are not cities: they are metropolitan areas. Download the 2006 updates and examine an MSA at random; you'll find something like this, showing the MSA and its subdivisions:

19100 Dallas-Fort Worth-Arlington, TX Metropolitan Statistical Area
Principal Cities: Dallas, Fort Worth, Arlington, Plano, Irving, Carrollton, Denton, Richardson, McKinney
19124 Dallas-Plano-Irving, TX Metropolitan Division
Collin County, Dallas County, Delta County, Denton County, Ellis County, Hunt County, Kaufman County, Rockwall County
23104 Fort Worth-Arlington, TX Metropolitan Division
Johnson County, Parker County, Tarrant County, Wise County

The DFW-Arlington MSA thus comprises 12 counties (Collin, Dallas, Delta, Denton, Ellis, Hunt, Kaufman, Rockwall, Johnson, Parker, Tarrant, and Wise), in which are located a number of cities (including Dallas, Fort Worth, Arlington, Plano, Irving, Carrollton, Denton, Richardson, and McKinney). That grabs the major suburbs around DFW with the cities

It's still the auto era in the US; most communities aren't walkable anymore; there's limited public transit in this country; and one expects many Americans will use automobiles in order to have sex

[Unvanguard]

65. Yes, I know that's true of the report you linked.

That's the point.

But is it true of the CDC study HamdenRice referenced, tracking HIV infection rates among men who have sex with men in five cities? I don't think so.

[struggle4progress]

66. Here's the MMWR link for the five city study:

HIV Prevalence, Unrecognized Infection, and HIV Testing Among Men Who Have Sex with Men --- Five U.S. Cities, June 2004--April 2005

... CDC analyzed data from five of 17 cities participating in the National HIV Behavioral Surveillance (NHBS) system ... During June 2004--April 2005, participants in five NHBS cities (Baltimore, Maryland; Los Angeles, California; Miami, Florida; New York, New York; and San Francisco, California) were also tested for HIV infection after informed consent ...

Editorial Note:

... The 2004 NHBS system was conducted in 17 MSAs with the highest AIDS prevalence. Although this report focuses on testing results from five selected cities, behavioral data are forthcoming from all participating cities ...

http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5424a2.htm

My reading of this is: for the purposes of discussing such NHBS findings, "city" is used as a synonym for "MSA"

[Unvanguard]

67. But they are not taking a random sample of people in that area.

"Participants were recruited at bars (30%), street locations (20%), dance clubs (19%), cafes/retail stores (10%), Gay Pride events (6%), social organizations (5%), gyms (5%), sex establishments (3%), and parks (1%)."

They said the venues were randomly-selected, but even if they considered venues in the entire metropolitan area, naturally they would still be concentrated in the actual cities. So they probably got some people from the suburbs, but probably not proportionately.

[struggle4progress]

68. Those may be perfectly cogent remarks. But public health policy will be based on

whatever understanding we can obtain from the data we actually have.

[brewens]

16. Even if the FDA allows gay men to donate blood it is likely to..

prohibit any that have had sex with another man within the previous 12 months. How many gay men would be willing to abstain from sex for 12 months to donate blood anyway? I don't know many guys that willingly abstain from sex for even one week.
The 12 month restriction would be the same as for others the FDA considers high risk individuals.

[dsc]

17. some do out of not having a willing partner

but I also think those in a monogmous relationship should be able to give.

[HamdenRice]

20. You've made a few serious statistical errors. Your numbers corrected correspond with the CDC

While your statistical inference is interesting, it's still an inference, which is inferior to actually measuring infection rates of real people.

Also there's one huge error you make right off the bat -- you multiplied the probable percentage of males who are gay against the entire male population. But the ban is on men who have had sex with men -- which means sexually active men, which means men who are at least teenagers. Our country does not have an even demographic distribution across age, so you have to multiply the assumed percentage of gay men against the number of males that leaves out all males under say age 15, as well as all gay men who don't actually have sex with men. (Keep in mind the policy screens out MSMs, not men who are gay.)

The other big statistical error you make is that you are looking at HIV and AIDS cases, when the relevant number you need is HIV infections. Somewhere around 50% of HIV positive MSM in several studies don't know they are infected and percentages in some studies are even higher. By definition, HIV and AIDS cases counts people who know they are infected. So you need to double or maybe triple your number. So your figure of roughly 1/2 million cases is equivalent of between 1 million and 1.5 million infections.

When you make these adjustments, the results are actually quite comprable to the CDC statistics of 1 out of 6 to 1 out of 4 MSM being infected with HIV. For example, taking your estimate of 8 million gay men (including children and the elderly and those who are not sexually active) and an infection rate twice the diagnosed cases, we would get an infection rate of 1 million out of 8 million. If we drop out children (about 12%) from the total number of males the estimate is very close to the CDC numbers -- about 1 in 7.

That said, again I have to emphasize that inferences from multiple demographic statistics are simply less reliable than measuring the infection rates of actual populations. These very high rates are not in dispute in the "real world" and have been steady and consistent for a very long time. Sadly, they are also rising. I find it completely puzzling that on DU lately, so many people seem to believe that well known and well established facts are themselves somehow political. They aren't.

[dsc]

22. No it isn't as I mentioned in the other thread

According to the CDC link I supplied we have: At the end of 2006, an estimated 1,106,400 persons (95% confidence interval 1,056,400-1,156,400) in the United States were living with HIV infection, with 21% undiagnosed.1

The total number of AIDS cases in my chart was 1,009,219(cummulative since 1977) which included many who have died from it. Thus it isn't AIDS cases now which might need to be multiplied by 2 or 3 like you did. So my numbers should be multiplied by about 1.096. That makes my range change to just under 2.5 million to just under 3.7 million. Not exactly a big difference. In short I stand behind my numbers.

Now as to your chart. It is all inner city gays in at least three notorious gay sexual playgrounds (LA, NY, and SF). Still no discussion of where these men came from. Did they wait outside bars, bathouses? Did they run ads asking for men? Did they pay for participation? Offer free medical care?

[HamdenRice]

25. You have to stop thinking about it as a single study. It's based on "National Surveillance" studies

As the footnote says, the data comes from the National HIV Behavioral Surveillance (NHBS). It's not a study of just some bars or bathhouses, but uses a multitude of information gathering techniques from many medical sources, and then using statistics and demographics to come up with the most accurate figures the epidemiological community can come up with.

[muriel_volestrangler]

26. But the figures you quote do come from large cities only

and as such should not be used as representative of a whole country. For instance, a quote from the article that the other thread links to:

The National Aids Trust estimates from a study carried out in 2006 that around 0.12% of the general population is infected with HIV, this is around one in 830 people. In contrast the same study found that 23% (just under 1 in 4) of gay men living in London and 10% (1 in 10) of gay men living outside of London were infected with HIV.

http://thelinc.co.uk/2009/12/time-to-end-ban-on-gay-men-giving-blood/

Social behaviour patterns are different in large cities.

[Unvanguard]

28. 1 in 10 sounds about right.

Using a 4% estimate for MSMs and dsc's HIV infection numbers, and correcting the male population for the 12% of children, I get about that number for the US.

[HamdenRice]

33. I don't see it as behavior -- its the issue of sexual community

I think a focus on behavior tends to moralize the issue. For example, HIV rates are almost twice as high for black MSM as for white MSM, but many studies show their behavior is the same, or at least cannot account for the difference in infection rates. The difference in infection rates is accounted for -- well, by the difference in infection rates. Once the disease reaches a certain prevalence in a certain community from which people choose partners, the rate becomes self-reinforcing.

For that reason, I don't think the London study necessarily explains the American pattern, because we have much higher rates of geographic mobility.

That said, sure, I'll agree that the chart shows urban infection rates. On the other hand the USA is 80% urban.

[dsc]

36. No it isn't

you have added suburban and urban people together.

[muriel_volestrangler]

44. Your table shows infection rates in Baltimore and 4 other cities

Note that the total tested was higher in Baltimore than the other cities, despite it being far smaller that NYC or LA (and it won't have such aare larger proportion of MSM in the population that it could make up for that in any way). You can't take the total number of HIV cases found and imply anything from it for the country as a whole. Just weight the figures for those 5 cities acccording to their total popluation and you'll get a prevalence of under 20%.

In contrast, the CDC figures that dsc linked to, which are in this page too, say:

HIV prevalence in males >= 13 in the whole US: 828,000 = 685.7 per 100,000 (ie total males for that age group = 120 million)
Of the 828,000, 532,000 came from male-to-male sexual contact, and 54,900 from MSM or injection drug use when it can't be determined which it was.

Note that these figures include undiagnosed HIV - the page says " CDC now estimates that 1.1 million adults and adolescents (prevalence rate: 447.8 per 100,000 population) were living with diagnosed or undiagnosed HIV infection in the United States at the end of 2006" and "an estimated 232,700 (21.0%) persons living with HIV infection had not been diagnosed as of the end of 2006".

If we use a 6% figure for MSM out of the total male adolescent/adult population, then it's
(532,000+54,900)/(.06*120,000,000) = 8% of MSM who are HIV positive.

That';s using teh CDC figures for the whole country, which are designed to be used together. The one thing they don't do is estimate the proportion of MSM in the general population; they note a 3.7% figure for ever having male anal sex among 15-44 year olds. Even if we drop the overall proportion of MSM to that, we get 13%.

[HamdenRice]

48. Here's what the CDC says about the overall national numbers

But before that, I just should note that 13% (your number) is still in the same ball park as 16% (1 in 6).

That said, all this is inferring from multiple broad demographic sources, while the CDC simply tests people in a variety of medical settings, adjusts the data and comes up with broad demographic figures, which seems much more accurate than what we can speculate about on a discussion board. The CDC fact says the following (sorry, but this site immediately jumps to a pdf; just click stop on your browser to read the html):

http://www.cdc.gov/hiv/topics/msm/resources/factsheets/msm.htm

In the United States, HIV infection and AIDS have had a tremendous effect on men who have sex with men (MSM). MSM accounted for 71% of all HIV infections among male adults and adolescents in 2005 (based on data from 33 states with long-term, confidential name-based HIV reporting), even though only about 5% to 7% of male adults and adolescents in the United States identify themselves as MSM <1, 2>.

The number of HIV diagnoses for MSM decreased during the 1980s and 1990s, but recent surveillance data show an increase in HIV diagnoses for this group <3, 4>. Additionally, racial disparities exist with regard to HIV diagnoses within the MSM population. A recent study, conducted in 5 large US cities, found that HIV prevalence among black MSM (46%) was more than twice that among white MSM (21%) <5>.
...
HIV/AIDS in 2005
(The following bullets refer to the 33 states with long-term, confidential name-based HIV reporting. See the box, before the References section, for a list of the 33 states.)

* In the 33 states with long-term, confidential name-based HIV reporting, an estimated 19,620 MSM (18,296 MSM and 1,324 MSM who inject drugs) received a diagnosis of HIV/AIDS, accounting for 71% of male adults and adolescents and 53% of all people receiving an HIV/AIDS diagnosis that year <1>.
* The number of HIV/AIDS diagnoses among MSM (including MSM who inject drugs) increased 11% from 2001 through 2005 <1>. It is not known whether this increase is due to an increase in the testing of persons with risk factors, which results in more HIV diagnoses, or due to an increase in cases of HIV infection
...
Unknown HIV Serostatus
Approximately 25% of people in the United States who are infected with HIV do not know they are infected <14>.

* Through its National HIV Behavioral Surveillance system, CDC found that 25% of the MSM surveyed in 5 large US cities were infected with HIV and 48% of those infected were unaware of their infections <5>.
* In a recent CDC study of young MSM, 77% of those who tested HIV-positive mistakenly believed that they were not infected <15>. Young black MSM in this study were more likely to be unaware of their infection―approximately 9 of 10 young black MSM compared with 6 of 10 young white MSM. Of the men who tested positive, most (74%) had previously tested negative for HIV infection, and 59% believed that they were at low or very low risk.

[muriel_volestrangler]

49. I'll reiterate: the 13% figure only comes if you use the 3.7% figure for 'had anal sex'

If you use the midpoint of the "5% to 7% of male adults and adolescents in the United States identify themselves as MSM", you get a figure of 8%. 1 in 12, not 1 in 6; and nowhere near 1 in 4.

Your figures were 2 to 3 times higher than the CDCs.

[HamdenRice]

51. So let's say 10%

It's still 10 times higher than a hard hit demographic group like black women (1 in 100), and not in the same ball park as majority demographic groups like white women (1 in 2000).

Again, we're mostly speculating on broad demographic categories. The CDC and its grantees actually go out and measure infection rates in a variety of medical settings, and adjust using the best demographic and statistical models, and according to them HIV rates are between 16% and 46% for urban MSMs.

[muriel_volestrangler]

53. No, let's say "1 in 12", since that's what the CDC says

when used with the midpoint '5 to 7%' self-identified MSM that the CDC also quotes. You could say under 10%, since that's the upper limit of the CDC estimate.

Yes, it's much higher than the figure for some other groups; but that's not what this thread is about. The 5 city figures are not "adjust(ed) using the best demographic and statistical models"; they are weighted towards Baltimore (and, for that matter, more towards the 30-39 age group than the 40-49 one, and away from the 50+ group), and are just simple totals of whom they tested, so the 46% prevalence among blacks is not a proper statistic for any population, urban or not.

The 'adjusted' figure is about 600,000 total MSM with HIV, diagnosed and undiagnosed, in the USA.

[dsc]

27. No it isn't

your own link called it a study, not a survey of studies. From your own link:

A study of MSM in five U.S. cities found extremely high levels of infection among MSM, and many of those infected
did not know it.
• Overall, one in four MSM participating in the study was infected. Black MSM were twice as likely to be infected with
HIV than other MSM.
• Among all of those who were infected, about half were unaware of their HIV status. Results were particularly
alarming for black MSM and young MSM, with more than two-thirds of infected black MSM, and nearly 80 percent
of infected young MSM (aged 18–24), unaware that they were infected.

This is what your chart shows. It is based on one study per the CDC link you provided.

[HamdenRice]

38. You're confusing several things now

The quoted language mentions a study. The table shows national surveillance data.

What they show together is that the data on infection rates is remarkably stable and consistent -- whether gathered from a single study or deduced from a nation wide surveillance survey.

Just because they report the same rates doesn't somehow make them wrong.

[dsc]

54. your table is totaly unlabled so we haven't a clue where it came from

[HamdenRice]

55. It came from the CDC fact sheet I linked to. Also the table itself has a footnote at the bottom

explaining that the data comes from the national surveillance study.

[Unvanguard]

29. That doesn't answer the question, though.

How are they getting their sample for testing? It has to have been done at some point, in some study. If they're just looking at the infection rate, they can't have only relied on people actually diagnosed with AIDS; they had to have tried to find a representative sample, tested those, and then extrapolated. And dsc is right; it's at least not obvious how they could have done so without introducing selection bias.

[HamdenRice]

37. The national surveillance study looks at a wide variety of sources

There are a few dozen centers of study. So, for example, some teaching hospitals participate. When people come in for all kinds of medical issues, they may be asked to take an HIV test. In other cases, data is gathered in VD clinics, and in other studies in bars. Obviously the latter will be more skewed than the former, but statistics and probability allow the CDC to extrapolate the data to the general population.

[pinto]

46. The NHBS wasn't designed to determine *prevalence* it was developed

in part to help design prevention programs based on *behavioral studies* in cities with an already high level of AIDS.

from the CDC -

Monitoring HIV Risk Behavior

Behaviors are monitored with regard to risk taking, HIV testing, care seeking, and adhering to treatment for HIV in different populations. <snip>

The NHBS (National HIV Behavioral Surveillance System), for populations at high risk, began in 2003. NHBS conducts surveys in cities with high levels of AIDS among MSM, IDUs, and heterosexuals at high risk to determine their risk behavior, testing behavior, and use of prevention services. (emphasis mine ~ pinto)

http://www.cdc.gov/hiv/topics/surveillance/index.htm

[Unvanguard]

23. Which "CDC statistics" are you referring to?

The ones in the table you copy (which refer to the HIV infection rate in five cities, not a national rate), or others?

If there are others, could you link to/cite them?

(Also, whatever the rate, it doesn't address the other angle of the issue, which is whether you can do better with a less blanket ban. This has always been the difficulty with people citing the disproportionate level of HIV infections among gay men: with sufficient controls, say, someone in a monogamous same-sex relationship who always uses condoms might not have a particularly high risk.)

[Gormy Cuss]

24. +1.

Although I applaud dsc for trying to make sense from the very limited CDC fact sheet.

[HamdenRice]

43. I also applaud dsc for trying to look at this rationally and come up with a statistical viewpoint

[ruggerson]

39. The figures you cite are idiotic

as they do not take into account suburban and rural gay people, as others have noted.

What is your political agenda here?

[HamdenRice]

40. So the CDC is idiotic?

They are basically the best the nation has in terms of disease detection and monitoring.

[Touchdown]

58. He noted gay men only. Add bisexual men to the math, and he's still basically correct.

Bisexual men could possibly double the % numbers men having sex with men, depending on the statistical model used.

[nealmhughes]

21. Actually, all stats would be what is termed "self-reported," and therefore suspect.

People lie. People have different ideas of what "gay" constitutes, even men who have sex sometimes or always with other men may or may not identify as "gay."

Any stat would be more or less either a guesstimation or an estimation using charts and SSPS or other stastical software.

By the way, there are those who are assuredly "gay" and are celibate and have always been celibate. They are obviously no risk for HIV if they have never been exposed to blood or body fluids in any form, yet they are gay. I am sure that the guidelines are for ensuring the safety of the public in the most general way and not for discrimination per se. The Red Cross does not care from what class of people they get their blood so long as it is within their guidelines.

By the way, people in the US who have spent X amount of time from a time frame of Y-Z months during a certain time period in the UK are not allowed to give blood in the US but can anywhere in the Eurozone. Make sense? No. A holdover from Madcow hysteria in the UK, even though we had our own here in the US.

I had platelets last week, by the way, and all I could think about was the class of people I saw hanging outside the "We buy blood" center the day before. They did not appear to be hard working folks out of work and looking to make a happy holiday for their family by a long shot. It looked like Meth Central. But I took 2 units with no quals, I figure my HIV is under control and cancer trumps HIV for a short life. I wish I could give blood, as I am
B Negative, but I cannot and I accept it. I am willing to face this "discrimination" of being gay and HIV positive and having lived in the UK and avoid internal Red Cross politics in which I have zero voice and rightly so long as they do not discriminate on who gets the blood!

In short, in public health, there is no easy answer: by and large the establishment errs on the side of safety for the public good. Discrimination? Perhaps, but one that is not aimed directly at one single class of people, rather at the likelihood of spreading disease due to people refusing to get tested and also people lying. Sadly, people are apt to do both; lie and refuse to get tested, and that is simply the way of all flesh.

Frankly, when it comes to HIV or gay issues, this one is low on my totem pole by far.

[RaleighNCDUer]

31. Estimating the rate of HIV infection by the numbers going for testing

is like estimating the rate of marital infidelity by the numbers getting marriage counseling.

A really fucked up, bound to be inflated number.

[TicketyBoo]

42. The whole "study"

is bogus, anyway.

What percentage of gay men won't even admit it in the anonymity of a census?

What percentage of gay men won't even admit it to themselves?

Just another example of your government in action, wasting money to come to illegitimate conclusions for obscure purposes.

[JonQ]

47. I would say this is next to impossible to figure out

either way, considering that it requires the respondants to answer A) honestly and B) actually remember how they got it. If a guy has numerous sexual partners, male and female all the while doing intravenous drugs in a month, then finds he has HIV, who gave it to him?

Also not everyone with HIV is diagnosed.

[struggle4progress]

57. ... During June 2004--April 2005, participants in five NHBS cities (Baltimore, Maryland; Los Angeles

California; Miami, Florida; New York, New York; and San Francisco, California) were also tested for HIV infection after informed consent ... Of the 1,767 MSM, 450 (25%) tested positive for HIV (range by city: 18%--40%) ... Of the 450 HIV-infected MSM, 217 (48%) were unaware of their HIV infections ...
HIV Prevalence, Unrecognized Infection, and HIV Testing Among Men Who Have Sex with Men --- Five U.S. Cities, June 2004--April 2005
MMWR June 24, 2005 / 54(24);597-601
http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5424a2.htm

The Chicago study this last summer similarly indicated 50% of HIV positive MSM were unaware of their HIV status. Their estimate for Chicago HIV infection rate among males was 17%. A natural conclusion might be that it is prudent to double diagnosis-based estimates of HIV incidence among MSM

[muriel_volestrangler]

62. The CDC estimates the number of MSM unaware of being HIV positive at 21%

at the end of 2006; after that Baltimore study. See the sub-thread containing http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=389&topic_id=7166242&mesg_id=7166953

[struggle4progress]

64. I'll certainly agree the CDC methodology (using later diagnoses to estimate currently undiagnosed

cases) is reasonable. I think the CDC means "21% of all cases are undiagnosed," not "21% MSM unaware of HIV status" but
lacking a better proxy, one might take the same number to estimate undiagnosed MSM cases: it would mean the diagnosed cases need to be increased by about 26%. Estimates in this thread seems to run 8 - 13% total MSM incidence. I did a spread sheet calculation earlier today (summarized here: http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=389&topic_id=7163238&mesg_id=7176830 ), which led to 13% infection rate by age 49 of a cohort followed from age 15, using age-specific risk estimates from data (assuming, however, no mortality)