I need MATH help! (Probability and Statistics) for AIDS discussion

none | 10/15/03 | self

[DB]

Reading further I see you updated your "lottery" example.

[DB]

In my post #13 I stated the assumed conditions for the given results. I make no claim as to the accuracy of his assumptions as the original poster was asking for "math help".

Your characterization of the numbers is not quite correct. The original poster’s base assumptions are that condoms fail 3% of the time and that an infected person infects an uninfected person 20% of the time when a condom fails. So the assumption was not 100% infection after condom failure. 20% of 3% is 0.6%.

[DB]

The issue was his math not his assumptions (at least that's what the poster asked for help with).

Infection only has to occur once. Turning up heads once is all it takes to lose. The more chances you take the more likely it is you will lose.

The original poster's model assumes a 20% probability that a failed comdom used with an infected partner will result in the other partner being infected. 20% of the 3% condom failure rate is 0.6% overall for each event.

[Maelstrom]

No, 3% of the time condoms have a flaw from the manufacturer.

They fail a great many more times. One frequent problem they have is the tendency to disintegrate in the presence of glycerine (such as Vasaline). They also expire, which teenagers frequently don't consider.

Other times they are used improperly or sized incorrectly or fall off when not at full stand. The users are often confident of invulnerability for wearing one. It fails quite a bit more frequently than merely the 3% of the time they have manufacturer's flaws.

[Nebullis]

You are calculating for an exposure rate. The success of infection by the virus is much less than that, or about 1/200 for bloodcontact with the homegrown virus. The transmission rates for HIV on the African continent is higher.

Second, you are assuming a condome failure rate of 3%. This is, I believe, the number for breakage. There is significant research showing that imperfections in latex will leave permeable regions in otherwise intact condoms.

[wideminded]

1. Transplant patients and hemophiliacs were not the only ones given AIDS by the tainted blood supply before proper testing was instituted. Many others got AIDS from tainted blood in the US and France. There is recent news that improper blood handling is responsible for an explosion in AIDS cases in China.

2. You didn't answer my other question. Perhaps you are not in a position to obtain a shot of pure HIV, but Prof. Duesberg is and he has not yet backed up his statements with this proof despite offering to do so many years ago.

[samtheman]

I don't believe your numbers. Show me some non-political sources for these numbers.

[Maelstrom]

From the people advocating their use:
http://www.hivdent.org/publicp/inter/ppinURACD072003.htm

A draft report for the UN's AIDS agency has found that even when people use condoms consistently, the failure rate for protection against HIV is an estimated 10 percent, making them a larger risk than portrayed by many advocate groups.

http://www.nacjamaica.com/youth/condoms.htm
What Is the Failure Rate?

By itself:
Perfect-use failure rate 2%
Typical failure rate 10%

From the distributers:
http://www.contraception.net/resource_centre/barrier.asp
To be most effective, condoms should be used with a spermicide. Used correctly, the typical failure rate of latex condoms is about 3%. However, application problems and breakage can occur with condoms. When used incorrectly, the failure rate for condom use rises to 12%.

Now these are condom *friendly* groups. Family groups report failure rates between 20% and 30%.

[Maelstrom]

So 3% condom failure rate and 12% operator error.

But remember kids, abstinence works 100%, with no failure rate.

[Poohbah]

Odds of infection in a single sex act = .03

There's the first mistake.

One case of unprotected sex != 100% chance of infection. I think the chance of infection is something like 0.2% for each incident of sex with an infected partner.

But this doesn't take into account things like "risky" practices, which push it much higher (I've heard 2% for anal intercourse--and, BTW, anal intercourse produces greater mechanical stress on the condom, increasing the chances of failure).

Condom failure rate of 3% only applies to mechanical failure. The HIV virus can pass through the condom. A more likely failure rate is 10% (SWAG only).

And let's assume that the infection rate in your subgrouping is 20%, and purely random selection of partners.

We have .02 x.10 x .20, for an infection rate of .0004.

That equals a "survival rate" of 0.9996.

Across 520 encounters: (0.9996)^520, or 81.2% survival rate. Make the guy active twice a week: 65.9% survival rate.

Make him active five days a week (common in the gay community: his survival rate is 35.3%.

If he's "barebacking (i.e., deliberately seeking to get infected with HIV, no condom, partner known to HIV+):

Odds of infection: 0.02. Odds of not being infected: 1-0.02 = 0.98.

Assuming he's active 260 days a year (0.98^260), his odds of getting HIV are 99.5%.

[Lonesome in Massachussets]

Since you spoiled it, I didn't bother to solve it. (Actually, I didn't bother to open the Boston Herald at all yesterday, when I got home it was 4-1 Red Sox, so I figured I'd let them spoil the rest of my day. By the time that was over I didn't feel like looking at the paper at all.) I suspected it was Goethe, he always (imho) speaks in platitudes, so I just did a google search on the quote. I think the popularity of Goethe has something to do with trying to rehabilitate the German image after WW-II, in translation he always seem a little lame to me, compared say, to Shakespeare or Marlow.

Anyway, he got a pass on this one.

[Old Professer]

Sorry I spoiled it; the comment seemed so close to the discussion at hand.

BTW, no pen makes me think harder; I'm glad you suggested it.

Haven't missed a one now in months; today's took all of five minutes.

[usapatriot28]

Your characterization of the numbers is not quite correct. The original poster’s base assumptions are that condoms fail 3% of the time and that an infected person infects an uninfected person 20% of the time when a condom fails. So the assumption was not 100% infection after condom failure. 20% of 3% is 0.6%.

OK, then he is still assuming that you are only having sex with an infected person. Either way a major factor is being left out.

[DB]

No argument there. I was simply addressing the proper math based on his supplied conditions. In my post #13 I restated the conditions so that it was absolutely understood what the math was based on along with making clear its limitations due to the given assumptions.

[Yeti]

Rosencrantz and Guildenstern Are Dead.

[Stone Mountain]

I seriously doubt condoms fail at a rate anywhere near 3%,

See post 114...