> In the US, 0.4% of adults have HIV [1]. With >20 sexual partners the cumulative exposure becomes meaningful. Naively, 20 repetitions of 0.4% gets 7.89% exposure risk.

Nope. The 0.4% is overwhelmingly concentrated in men who have sex with men. It's approximately an order of magnitude lower in women.

Based on your envelope math, I should have a >90% chance of having HIV, but that's nowhere near the truth (and I don't).

> The 0.4% is overwhelmingly concentrated in men who have sex with men. It's approximately an order of magnitude lower in women.

I added a note to interpret with caution because many factors can significantly affect this figure. Some states have an order of magnitude difference from others, potentially offsetting the MSM factor you mention [1].

> Based on your envelope math, I should have a >90% chance of having HIV

The envelope math was exposure not contraction. If you used or condom, or even just received oral, it's quite possible you didn't get HIV even if your partner did.

[1] Compare WY with GA - https://en.wikipedia.org/wiki/HIV/AIDS_in_the_United_States#...

> The 0.4% is overwhelmingly concentrated in men who have sex with men. It's approximately an order of magnitude lower in women.

True, but it's also true that male-to-female transmission is easier than female-to-male transmission, which makes the odds better for the man in any given encounter.

I'm not sure what the systemic effect is.

Depends on the country as well. They may be true in the US but it is not true in other counties.

The risk of transmitting HIV during unprotected penis-in-vagina sex is less than 1 in 1000. With protection it's even lower. And if the HIV-positive one is on meds it's also very protective.

Kind of but your example isn't great.

It's in some senses more of a concern with things like HPV + HSV, where there's not a lot of testing options and barrier contraception is of much more limited effectiveness. (that said, Gardasil vaccination helps).

For most other STIs - either get tested before every new partner or on a reasonably frequent basis, and demand the same from every new partner.

HIV prevalence + transmission likelihood is wildly skewed to gay/bi men who are active with men - and if you are one who's actively having new partners you probably ought to also talk with your doctor about PrEP.