Log in

No account? Create an account

Thu, Apr. 15th, 2010, 03:44 am
Another Stupid Legal Response to HIV

Darren Chiacchia, an American equestrian athlete who won an Olympic medal in Athens in 2004, is on trial.

The charge? Exposing his partner to HIV in the state of Florida, where it is illegal for an HIV+ person to have sex with someone without disclosing HIV status. According to Chiacchia's bitter ex-lover, Chiacchia hid his status from his boyfriend for years. Chiacchia said he was afraid to tell his lover about his HIV status because he was afraid it would go public. His lover found out and made it public anyway. Chiacchia probably deserves that much.

I don't need to go into detail to explain that knowingly exposing someone to HIV is a shitty thing to do. I'm not going to talk about it - not because I don't think it's terrible - but because it is already obvious to everyone reading this and nothing needs to be added. There will be plenty who pile on Chiacchia for his actions.

But what I will say is that the situation has already been irredeemably mishandled, and that the law is bogus in the first place.

Laws criminalizing the transmission or HIV have been misused by ignorant prosecutors and public. In 2008, for example, a Dallas homeless man who is HIV+ was sent to prison for 30 years for spitting on a police officer. The jury, either unaware or indifferent to the fact that HIV is not transmitable through saliva, decided that the saliva was a "deadly weapon" before issuing a guilty verdict.

This is like charging someone with attempted murder for shooting somebody with a squirt gun.

It is hard to get HIV. There are two ways for adults outside a medical setting to get it: through sexual intercourse, and through injections. Inject HIV+ blood into your body, and you stand a good chance of getting HIV. Have unprotected sex with an HIV+ and you have chance of having HIV (between 1 in 50 and 1 in 200; have unprotected sex with an HIV+ person 100 times, as people in a new relationship may do over the course of a year, and your chance of having HIV is pretty high.) But getting HIV through tears, saliva, urine, feces or skin is impossible. Oxygen kills HIV. Nobody has ever gotten HIV through a cut or scratch; you can roll in a puddle of HIV+ blood on the floor, even with cuts on your body, and it's not going to infect you.

HIV needs to be injected fresh, without exposure to air, from one body directly into your body. This happens during IV drug usage and through sex, it happens through organ transplants, it can potentially happen when an infant drinks dozen gallons of her or his mother's infected breast milk over the course of a year, and can happen to an infant during childbirth if the mother is infected.

Source: Wikipedia: HIV Transmission

In the Florida case, too, there is much to question. It is more likely here that the two had sex, which can lead to transmission, but prosecutors have not said what risky behavior Chiacchia and his partner engaged in, or if protection was used. Chiacchia could have feasibly insisted on condoms knowing he was infected. Nor can anybody ever prove whether Chiacchia really lied about his HIV status; it is possible for any person in a relationship with an HIV+ person to claim not to have known about his partner's status to satisfy a vendetta. Chiacchia's partner has not become infected with HIV; the charge is that Chiacchia wrongly placed his partner at risk, though his partner has not been harmed.

The problems with criminalizing HIV transmission are numerous, but in my mind, the ultimate reason why I think the laws fail is because they do not prevent the spread of HIV - they actually promote it. I am more concerned with stopping a deadly disease than with some sentimental form of "justice" where people are locked up for selfish behavior. When the goals conflict, I side with the practical: with preserving human life and health is paramount. I do not allow innocent people to suffer to punish guilty people; that is not a good trade.

The reason criminalizing HIV is dangerous is explained:

Imagine you are a person who has engaged in some risky behavior; perhaps drug use, perhaps unprotected sex, and you are unsure of your HIV status. You think (as many hypochondriacs do) you might have it, but your mind waffles. You might wake up in the morning convinced you are going to die and go to bed at night feeling reassured, talking yourself out of it: "I felt great today," you might think; "I had so much energy!" You might panic every time you get a cold or allergies thinking it is an early sign of AIDS. Even without treatment you will probably live for 10 years as an HIV+ person - which is why many people procrastinate on getting tested. You can put it off for a few months. It is February: lets say you resolve to get tested before the end of December.

An added deterrent is that if you do get tested and the result is positive, it will suddenly become illegal for you to have sex, and you know this. Perhaps you are already in a relationship, perhaps there is someone you have your eye on. Sure, if you tell your partner or partners your status and they consent to having sex using protection, it would be safe and not illegal. But you are convinced no one would have sex with you if they find out you have HIV. You are scared and irrational, and still believe that the world ends when you find out you have HIV. You still think that everyone you love turns their back on you - and sometimes they do. You wake up each morning terrified. You talk yourself down during the day; you've had panics before, and in the past you got tested and everything turned out OK. Almost everyone has a false panic sooner or later. You go to bed convincing yourself that all is well.

In the midst of this, you have sex with several people. You go to a bar and get drunk, and in your intoxicated euphoria you don't even think about your risk; when drunk you decide that you are sure you don't have HIV, and you have sex. Your anxiety actually drives you to the bars more often, or to use other drugs that lead to unprotected sex.

Lets say for the sake of the scenario that you actually are unknowingly HIV+, and infect several other people.

This could have been avoided had you been tested, and told, through counseling that is often provided with the test, that you have a wide number of options available to you as an HIV+ person, that you are not likely to die anytime soon because medication works very well now, that it is still safe to have sex if you follow certain procedures, that HIV+ people continue to find and have romantic relationships and this is actually encouraged by their doctors, and that the first step to reducing the spread of the disease is knowledge.

Instead, you have unknowingly infected people with HIV (and this is not illegal, since you never knew you had it), because the law terrified you away from getting tested. Many will say that putting others at risk in this way is still selfish and wrong, and sure, that is true, but many, many many people do it, far more than would knowingly infect others with HIV. Many people alive and healthy and HIV- today have had sex while not knowing their status, and are all equally guilty of putting others at risk though they got lucky.

That is just one way that criminalizing HIV transmission leads to more transmissions. The other problem is that it gives the community a false sense of security; if you are in a high-risk group, you may think it's safe to have unprotected sex with strangers because they are required by law to tell the truth. You ask your partner or your acquaintance if he has HIV, he says no, so you decide not to use a condom. This is a common assumption. HIV activists fight to prevent it all costs and think it's wrong, wrong, wrong, but many people do it anyway.

The problem is, most HIV transmissions come from someone who didn't know her or his status in the first place. Your partner may say "I tested negative two months ago - I'm negative" and still infect you. You are most contagious when newly infected, which is ironically before you know you are contagious; it could be days after a test. The false sense of security encourages people to take more risks. The risks lead to more HIV transmission.

People who work full time to stop the spread of HIV know: the answer is to educate people to get over their fear, to encourage testing, openness, honesty, and to most of all, remove the stigma associated with HIV+ people. The view is that human beings are naturally good. They don't want to spread HIV around. Very, very few want to willingly infect others, especially when they are given positive support and hope. Remove the fear, and they will get tested, live openly and honestly, and take steps to protect those around them. Research has shown that people do this and it works.

I do think that Chiacchia's public trial will have a chilling effect on people considering getting tested in Florida. It will put unnecessary stigma on HIV+ people and it almost definitely will lead to more people getting HIV than would have otherwise. Is that a fair tradeoff in the District Attourney's mind?

That's up to the prosecutors, and apparently their answer is yes.

Thu, Apr. 15th, 2010 06:03 pm (UTC)

Great post as always, Matt. You're a talented writer.

What I liked most, funny enough, was that you just wrote "have unprotected sex with an HIV+ person 100 times and your chance of contraction is pretty high". here you just stated a good qualitative assessment without using faulty math.

I think I mentioned this before, and you must know this now, but Bernoulli trials (doing a thing once and independently from all other trials, the outcome being some binary with a fixed probability each time) are not additive. For example, if I flip a coin twice, my odds of getting at least one heads within two flips is not 1 (= 1/2 + 1/2). It's 3/4.

The Geometric distribution is what accounts for this. If the probability of "success" in a Bernoulli trial is p, then the probability that the k-th trial is the first success is (1-p)^{k-1} * p. The probability that the first success has happened already by the time of the k-th trial is 1 - (1-p)^k. So, to use your example:

Say the odds of a person contracting HIV during unprotected anal sex with a seropositive partner is 1 in 100. This is likely not accurate (nor even a good probabilistic model, since so many varying factors can go into the moment-to-moment likelihood-- but it may approximate the mean well) but let's work with it. Then the probability that you'd have been infected after one encounter is .01 (1%) Two encounters? 1.99% 10 encounters? 9.56%. 50 encounters? 39%. 100? 63%.

So it's easy to see how one's intuition on the math begins as it does-- for a low number of trials, these results are roughly additive. But you can see how it diverges far away from that at higher numbers.

Oh. I should say now that I have no real point here about epidemiology, HIV transmission or anything else. Just felt like teaching some probability to an audience who already knew about it. :)

Thu, Apr. 15th, 2010 06:49 pm (UTC)

Re: the coin flip; yeah, I know that, haha. Obviously you don't want to do probabilistic math that gives you a 100% chance of something happening when you can tell, intuitively, that you will never reach 100% probability.

If the risk of infection after one exposure is 1%, then there would be a 1% chance of infection after first exposure, 1.99% chance after second exposure, 2.97% chance after third exposure, 3.94% chance after fourth exposure, and so forth. The equation I am doing is [(1.00-r)(.01)]+r = R, r being your percent risk after last exposure, R being your new risk after an additional exposure.

I do not know how to do an equation using n, n being the number of exposures. I suppose that's what you just wrote.

The quantity for R is, essentially, your percent risk of being infected at least once by now after this many (n) exposures. Of course, R(n) does not equal 100. The quantity that you "lose" from the equation each and every round **equivalent to 0.01(S), S being the chance that you were NOT infected (or in other words 1.00-R)** represents the percent chance that infection has happened at least twice.

That number is theoretical (since double-infection only counts as an infection for our purposes) and is completely contained within R. Perhaps to visualize it in a pie chart indicating risk, you would fold part of the R section over itself.

If the quantity representing double-infection is D, that means that D+R+S=100.

I don't know how the hell I came up with those numbers, though. Are they correct?

Thu, Apr. 15th, 2010 08:14 pm (UTC)

Oh. So the point is, these things are independent. They're memoryless. It's less like playing Russian roulette with the same gun (in which a barrel advances) and more like grabbing a new gun for each turn. The risk doesn't change on the second time or the third time. It's the same as the first time or as the thousandth time. You could be really lucky or really unlucky.

It's like this: if you throw a nickel in the air, you're equally likely to get heads or tails. Now say you throw a nickel in the air a thousand times and happened to get a thousand heads-- the 1,001st time still has an equal probability between heads and tails. The nickel will not "know" that it's made an amazingly "improbable" streak. Because, of course, that streak it just made was just as probable as any other path it could've taken. It only seems insane to you because you're a human with human faculties for pattern recognition. Faculties that are not intuitive and unrelated to whatever it is a coin does. "1010101010010101111110000101010101010101010011" 'looks' a lot more random to you than "1111111111111111111111111111100000000000000000000000000000," but it needn't be. Read up on "normal numbers," "entropy" and "randomness" if this topic is interesting to you.

If you were actually an epidemiologist, Matt, there might be reasons for you to alter your model so as to increase risk over time (or to negate the assumption of independent and identically distributed Bernoulli trials). Say, a person might have an increased likelihood over time of having contracted other STIs or conditions that would allow for an easier effective transmission of HIV. But if you say "1 in 100 risk," that's standard for the first time as it is for the one-hundredth time. Read up on the Geometric distribution for the question of when the first success finally occurs.

Thu, Apr. 15th, 2010 10:43 pm (UTC)

I know the risk doesn't change, but the probability changes because of the fact that, while any one of the 100 exposures could have infection, whether you are infected or not is binary. So you have to subtract the chance of being already infected from new risk.

Thu, Apr. 15th, 2010 08:20 pm (UTC)

Oh, maybe you were trying to understand the concept of accumulated risk. That's what's referred to as the cumulative distribution function of the geometric distribution.

The geometric distribution has a probability distribution function (PDF) that tells you the odds of success (or in this case, "infection") happening on the k-th trial. It also has a cumulative distribution function (CDF) that tells you the odds of success BY the k-th trial (the sum of the probabilities from trials 1 to k). It's the latter you're likely wondering about.

Thu, Apr. 15th, 2010 10:22 pm (UTC)

Yes, that is what I was saying.

Every incident gives you a 1% chance of infection. Some of that chance overlaps with the presupposition that you were not already infected. Some of that chance overlaps with the chance you were already infected.

Therefore, if you already have a 40% chance of being infected, and you are involved in another incident in which you face another 1% risk, to calculate your new chance of being infected you break the 1% into two. You throw away the first 40% of the 1% (which is 0.4%) and add it to the likelihood of double-infection, which is insignificant. The remaining 0.6% add to your new risk, so after this added exposure you have a 40.6% chance of being infected.

Those are the numbers that make sense to me, explained by [(1.00-r)(.01)]+r = R.

Thu, Apr. 15th, 2010 08:27 pm (UTC)

As to the other content of your post, I completely agree. Socially stigmatizing HIV+ persons has the same practical effect as criminalization, I think. You then end up with people who don't get tested (and thus not treated) and who, even if they know, are much more unlikely to tell folk. Because, let's face it: the average idiot college-aged gay man is way more likely to have random unprotected sex with a stranger whose status he doesn't know than to even have protected sex with someone whose status he knows to be positive.

Though that may seem like hedging his risk, he's actually at way more risk doing the former-- both because that particular partner has probably been in many similar situations himself and if positive, is likely unmedicated and more virulent; and because that self-preserving positive partner in the latter situation is likely medicated and has an effective viral load of zero-- not to mention would be using a condom.