HIV statistics and do people actually understand them.

[Ts RedVeX]

Since I work in the sex industry and I often hear people saying bad things about barebacking, yand then I also see many other people not being concerned about contracting HIV too much, I find it interesting what makes all those people think so differently. This thread is by any means not meant to encourage anyone to bareback or conclude whether barebacking is good or bad. It is supposed to be more about understanding information provided by media correctly as well as how much the way of looking at the info changes one's personal the feeling about it.



Since nobody actually concluded how to calculate the probability of ending up HIV positive in the other thread..


Let's assume that the probability g of getting infected during intercourse with an HIV positive person is in accordance with the data in the table https://www.aidsmap.com/Estimated-ri...8/#item1324093


g=1/123


then every time I have sex with an unknown person, so I do not know whether or not they are infected, my probability of getting HIV is


P=g*x


where x is the probability of that person being HIV-positive.

To calculate x lets look at https://www.hiv.gov/hiv-basics/overv...bal-statistics and take the worst case scenario - that 51.8M people have HIV. Lets assume global population of 7.6G people in 2017- as wikipedia reads: https://en.wikipedia.org/wiki/World_population.


x=51.8M/7.6G
x=6.76m


therefore:


P=1/123*6.76m
P=54.9μ


which means that the probability of not getting infected during a single intercourse is:


N=1-P


Now we have a series of independent events of tossing a coin only that the coin is more likely to give the face that corresponds to not getting HIV.
The probability N(n) of not getting HIV during a series of n intercourses is


N(n)=N^n


So the probability P(n) of getting HIV in a series of n intercourses with different random people is:


P(n)=1-N(n)


P(n) is practically a linear function untill you get to the point of planning to see thousands of different random people. This means that if you are HIV negative and plan to see n different random people the probability of having caught HIV at the end of your punting can be calculated with



f(n)=0.000054*n


as long as you are not planning to see more than too many different random people.


Lemme know what you think, especially if you reckon that this approach is wrong. Not sure why the old thumbnail remains... Crappy GUI I guess ];>... The error in the thumbnail is that there should be "1 client per day" wherever it says more than 1.

[GroobySteven]

Nobody has sex with random people. We have sex within a very predefined group.
I'm not likely to have sex with people of certain ages, genders, locations, preferences. So that would affect your numbers massively.
Also the type of sex I'm likely to have would affect it. As a top, I've a different risk factor than a bottom. As someone who is overall healthy, then I've a different risk factor.

[Ts RedVeX]

How do you mean? I have all sorts of clients visit me. Married single, homo, hetero... They are mostly males who are UK citizens. Which only lowers the risk further because the UK is said to be a low-risk country.

If you want to say that clients are usually male then that is irrelevant because if you diminish the global population by the number of women, you also need to diminish the number of HIV positive people in the world accordingly and x as well as P remain are not affected. They are also not affected by age looks etc as the stats are not not taking those into consideration. Let's stick to the stats we have here and only discuss their interpretation - at least for now. Otherwise we will get lost again.

The assumption that I always bottom is made at the point of g=1/123

[GroobySteven]

Let's stick to the stats we have here and only discuss their interpretation - at least for now. Otherwise we will get lost again.

That's why they don't work. They're irrelevant.
You've worked out nice numbers but based on a global population, without any other factors. If the whole world was at our fingertips, then they might be relevant but we just don't have those sexual partners available.

So yes, your numbers are right if I was going to have the ability or want to sleep with any random person in the world.

[Ts RedVeX]

That basically means the data provided practically cannot be understood without doing what sounds like a hell of a lot of other research. In that case, I am wondering how many people who say that it is safe or unsafe to bareback would actually be able to give a reasonable explanation for their opinions.

[GroobySteven]

I am wondering how many people who say that it is safe or unsafe to bareback would actually be able to give a reasonable explanation for their opinions.

Yep.

[Devlishsaint]

That basically means the data provided practically cannot be understood without doing what sounds like a hell of a lot of other research. In that case, I am wondering how many people who say that it is safe or unsafe to bareback would actually be able to give a reasonable explanation for their opinions.

But in that case doesn't it make sense to be safe rather than sorry?

[trish]

Were I to make the calculation I’d use the same method; i.e. search the reputable data for the value of g and x and calculate the probability that I won’t contract the infection given a single encounter as N = 1-gx. Then the probability of remaining uninfected after n independent encounters would be N^n. The probability of winding up infected after n independent encounter would then be P(n) = 1-N^n.

One could perhaps take more care in assigning values to g and x so that they better reflect the pool of people with whom you (the reader) expects to have sexual encounters, and the type sexual encounter you expect to be having. Some pools have greater or lesser infection rates, some kinds of sex of greater or lesser probability of spreading HIV.

In your example you take g, the probability of acquiring the HIV virus through one bareback encounter (I assume receptive anal) with an HIV infected person to be g=1/123 = 0.00813 . I didn’t double check this value with the link you gave; I assume you looked it up correctly. You also take the value of x to be x=0.00676. Again, the value one would want to use would depend upon the pool of people one is having encounter with. But lets go with your value. So N = 1-gx = 1-0.0000549 = 0.999945 .

You mention an interest in finding the value of P(n) if you’re seeing one client a day. If if you’re seeing one client a day, and the values you chose for g and x are suitable for your specific practices, then (in agreement with your calculations) we find...

the probability of acquiring HIV by the end of a year = 1-(0.999945)^365 = 0.0198 .

the probability of acquiring HIV by the end of 2 years = 1-(0.999945)^(2*365) = 0.039 .

the probability of acquiring HIV by the end of the decade = 1-(0.999945)^(10*365) = 0.182 .

The reservation that stands out most is the value chosen for x is probably too small for a transgender escort who’s seeing men on a daily basis. The value of x can have a dramatic effect on the rate at which P(n) climbs with n.

For an example, suppose you find out that for the last year you’ve been having weekly unprotected receptive anal intercourse with the same HIV inflected client. To make the calculations simple let’s just suppose he was your only client. Then x = 1. So N = 1-gx = 1-g = 0.99187 . We now find

the probability that you’ve been infected = 1-(0.99187)^52 = 0.346 .

Of course if you’ve been seeing other clients the probability is even higher.

[Ts RedVeX]

Tthe amount of sense it makes to use protection to an individual depends on the individual. I don't do bareback with clients, if that is what you wanted to know. The point of this thread is just to show how many different approaches people may have to statistics they encounter and how various interpretations of that information may be.

As to being picking other g and x, the value of g=1/123 is the value already representing the worst case scenario of being anally fucked bareback by a HIV-positive guy. Using the value already means we engage plan to in this most risky sexual activity every time.

As to x, I may have not mentioned it at the beginning of my original post, but I assume that every customer is a different guy and I stated that in the conclusion the original post's conclusion. This is justified for if an escort has a regular client who wants to penetrate her without protection then she can simply ask him to bring a certificate next time. - Considering that they trust each-other, of course. Still, it seems like having multiple encounters with an HIV-infected person can be easily avoided and is therefore unjustified to consider it here. - Nonetheless, it indeed is a very good example of how a way looking at data can dramatically affect the results of analysis:

P(52)=0.346 (same infected client seen 52 times; no other clients)

vs

P(52)=0.0028 (original conditions)

I don't think that being a transgender escort is much different to being a gay escort, so that difference is negligible too.

[trish]

I wasn't really asking anything about your case in particular. Just demonstrating how a change in the value of g or x can have a dramatic effect on the end result.

If one always uses protection then one can get a more accurate estimate of his or her risk if a more appropriate (in that case a lesser) value for g is used. For an escort in who deals with clients more likely to carry the HIV virus than other client pools, it may be a higher value of x (higher than the 0.00676 cited above) would give a more accurate estimation of the risks of business. The escort would have to know the stats for her particular sort of client base.